onetap

发表于 2022-07-11 更新于 2022-07-11
本文字数： 863k 阅读时长 ≈ 13:05

https://cn.bing.com/search?q=yolov6&cvid=8e7131079dda48fd90814bb4fdc36878&aqs=edge.0.69i59.4695j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | yolov6 - 搜索
https://cn.bing.com/search?q=%e5%9b%be%e5%83%8f%e5%88%86%e5%89%b2+2022&qs=n&sp=-1&pq=%e5%9b%be%e5%83%8f%e5%88%86%e5%89%b2+2022&sc=1-9&sk=&cvid=9F4C550A5A674CC0AD9959085A70FC2C&ghsh=0&ghacc=0&first=70&FORM=PORE | 图像分割 2022 - 搜索
https://cn.bing.com/search?q=cvpr2022+%E5%88%86%E5%89%B2&qs=n&form=QBRE&sp=-1&pq=cvpr2022+%E5%88%86%E5%89%B2&sc=8-11&sk=&cvid=B7006F8D99834CAE91DF48C999BC1A25&ghsh=0&ghacc=0 | cvpr2022 分割 - 搜索
https://zhuanlan.zhihu.com/p/534493711 | CVPR2022能同时做三个分割任务的模型，性能和效率优于MaskFormer！FAIR&UIUC提出通用分割模型，性能优于任务特定的模型！代码已开源！ - 知乎
https://blog.csdn.net/weixin_46142822/article/details/123969164 | (3条消息) CVPR 2022: 图像分割论文大盘点_大林兄的博客-CSDN博客_图像分割论文
https://zhuanlan.zhihu.com/p/478286484 | CVPR 2022 全面盘点：最新350篇论文分方向汇总 / 代码 / 解读 / 直播 / 项目（更新中） - 知乎
https://zhuanlan.zhihu.com/p/538050231 | 语义分割综述：截止2022，语义分割总结与展望 - 知乎
https://www.eet-china.com/mp/a112890.html | 2022入坑图像分割，我该从哪儿入手？-面包板社区
https://www.zhihu.com/question/533975556 | 2022年了，关于图像分割有什么新进展吗？ - 知乎
https://cn.bing.com/search?q=unet+%E7%B3%BB%E5%88%97&qs=n&form=QBRE&sp=-1&pq=unet+%E7%B3%BB%E5%88%97&sc=3-7&sk=&cvid=E214C468127D4EAFB4A6A1CB338F5EC6&ghsh=0&ghacc=0 | unet 系列 - 搜索
https://blog.csdn.net/Sakura_day/article/details/117855369 | (3条消息) Unet演化系列大整理_星画天的博客-CSDN博客_unet变种
https://zhuanlan.zhihu.com/p/355320561 | 2021 Review: U-Net Family | U-Net 家族新的一些进展 - 知乎
https://zhuanlan.zhihu.com/p/57530767 | 图像分割的U-Net系列方法 - 知乎
https://www.cnblogs.com/PythonLearner/p/14041874.html | 图像分割必备知识点 | Unet详解理论+ 代码 - 忽逢桃林 - 博客园
https://cn.bing.com/search?q=2D+2.5D+3d+%E5%88%86%E5%89%B2&qs=n&form=QBRE&sp=-1&pq=2d+2.5d+3d+%E5%88%86%E5%89%B2&sc=8-13&sk=&cvid=D45F47FC950246D3A36D2F59577AB635&ghsh=0&ghacc=0 | 2D 2.5D 3d 分割 - 搜索
https://www.zhihu.com/question/322077321 | 大佬们，我刚开始接触图像分割，对于图像分割深度学习这块一头雾水，可以给出一个从零开始学习的路线吗？ - 知乎
https://cn.bing.com/search?q=hourglass%E7%BD%91%E7%BB%9C&cvid=d2a31183887f466fa1081cfd729bd971&aqs=edge.2.69i57j0l8.2686j0j4&FORM=ANAB01&PC=HCTS | hourglass网络 - 搜索
https://zhuanlan.zhihu.com/p/57631965 | 姿态估计： Hourglass 网络 - 知乎

https://cn.bing.com/search?q=A+transformer+based+Network+for+def&qs=n&form=QBRE&sp=-1&pq=a+transformer+based+network+for+def&sc=1-35&sk=&cvid=9A2E0249215449B09CE07BF4B5694CB3&ghsh=0&ghacc=0 | A transformer based Network for def - 搜索
https://zhuanlan.zhihu.com/p/473818704 | Transformer-UNet：基于Transformer的可变形医学图像配准网络 - 知乎
https://cn.bing.com/search?q=paddlepaddle&cvid=7b62940303f644b889851470215df908&aqs=edge.0.0l8j69i60.11372j0j4&FORM=ANAB01&PC=HCTS | paddlepaddle - 搜索
https://aistudio.baidu.com/aistudio/projectoverview/public/2?kw=%25E7%25A3%2581%25E5%2585%25B1%25E6%258C%25AF%25E9%2580%25A0%25E5%25BD%25B1%25E5%259B%25BE%25E5%2583%258F%25E4%25B8%25AD%25E8%2584%2591%25E8%25A1%2580%25E7%25AE%25A1%25E7%25BB%2593%25E6%259E%2584%25E7%259A%2584%25E5%2588%2586%25E5%2589%25B2%25E6%258C%2591%25E6%2588%2598%25E8%25B5%259B | 开源项目 - 飞桨AI Studio - 人工智能学习实训社区
https://gitee.com/coggle/competition-baseline/tree/master/paddlepaddle | paddlepaddle · Coggle数据科学/competition-baseline - 码云 - 开源中国

https://zhuanlan.zhihu.com/p/92928820 | 使用CMake构建C++项目 - 知乎
https://cmake.org/documentation/ | Documentation | CMake
https://zhuanlan.zhihu.com/p/119426899 | CMake教程（一） - 知乎
https://cn.bing.com/search?q=qt5+%E5%AE%89%E8%A3%85%E5%8C%85+%E7%BD%91%E7%9B%98&qs=n&form=QBRE&sp=-1&pq=qt5+%E5%AE%89%E8%A3%85%E5%8C%85+w%27pan&sc=0-13&sk=&cvid=A466DAA67CA34E08A480A7FEEC1F8416&ghsh=0&ghacc=0 | qt5 安装包网盘 - 搜索
https://zhuanlan.zhihu.com/p/385486784 | Qt 5 下载与安装详解 - 知乎
https://download.qt.io/archive/qt/5.12/5.12.12/ | Index of /archive/qt/5.12/5.12.12
http://dts.digitser.cn/zh-CN/download/qt.html | 下载软件包 - Qt 5.15.2 for Windows 32&64 官方安装包,离线安装包,中文文档编制
https://pan.baidu.com/s/1CUA4BLYcnQAJgPK_Zl3iDw?_at_=1657154396808#list/path=/Digitser/Other/Qt/5.x/ | Qt_免费高速下载|百度网盘-分享无限制
http://c.biancheng.net/qt/ | Qt教程，Qt5编程入门教程（非常详细）
https://blog.csdn.net/chenchao_shenzhen/article/details/112059961 | (2条消息) 公司使用Qt到底要不要付费？_chenchao_shenzhen的博客-CSDN博客_qt收费吗
https://www.qt.io/download | Download Qt | Develop Desktop & Embedded Systems | Qt
https://cn.bing.com/search?q=%E8%AE%AF%E9%A3%9E&cvid=bd935560f65141d19b362b8b739f462b&aqs=edge..69i57j0l8.3231j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | 讯飞 - 搜索
https://cn.bing.com/search?q=%E5%A4%A9%E6%B1%A0%E5%A4%A7%E8%B5%9B&cvid=9b4d92845d5e46518da39cfba7e98270&aqs=edge.1.0l9.3502j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | 天池大赛 - 搜索
https://tianchi.aliyun.com/competition/gameList/algorithmList | 天池大数据竞赛-天池大赛-阿里云天池
https://tianchi.aliyun.com/competition/entrance/531901/forum | “英特尔创新大师杯”深度学习挑战赛赛道3：CCKS2021中文NLP地址相关性任务-天池大赛-阿里云天池
https://tianchi.aliyun.com/forum/postDetail?spm=5176.12586969.1002.3.596a2523Zt6c8V&postId=317162 | CCKS2021中文NLP地址相关性任务-赛道三-top1方案含源码-天池技术圈-天池技术讨论区
https://github.com/wodejiafeiyu/ccks2021-track3-top1?spm=5176.21852664.0.0.216b26b7iKOPHI

https://cn.bing.com/search?q=%e8%b0%b7%e6%ad%8c%e5%ad%a6%e6%9c%af%e9%95%9c%e5%83%8f%e7%bd%91%e7%ab%99%e5%af%bc%e8%88%aa+%e6%9c%80%e6%96%b0&qs=n&sp=-1&pq=%e8%b0%b7%e6%ad%8c%e5%ad%a6%e6%9c%af%e9%95%9c%e5%83%8f%e7%bd%91%e7%ab%99%e5%af%bc%e8%88%aa+%e6%9c%80%e6%96%b0&sc=0-13&sk=&cvid=685713A1961F4237B8DF57568876F0DE&first=11&FORM=PORE | 谷歌学术镜像网站导航最新 - 搜索
https://192link.com/12743.html | 分享最新速度稳定Google搜索镜像-192LINK网址导航
https://iknowyouask.com/3669.html | 谷歌学术搜索导航 – 国内可访问 - 楼友
https://www.myzaker.com/article/5cad47d48e9f0934f0151e69 | 最新可用 Sci-hub+Google 学术网址！-ZAKER新闻
https://jia110.github.io/index.htm | 2022谷歌搜索学术维基百科镜像-Google谷歌、Wikipedia维基百科、谷歌学术 github 镜像站、Sci-Hub、Internet Archive—-jia110百度先生
https://www.library.ac.cn/?f=a403-ustc | Mirror List
https://scholar.ustc.cf/scholar?start=30&q=%E6%94%AF%E5%8E%9F%E4%BD%93+%E6%B7%B1%E5%BA%A6%E5%AD%A6%E4%B9%A0+%E8%8D%A7%E5%85%89&hl=zh-CN&as_sdt=0,5 | 支原体深度学习荧光 - Google 学术搜索
https://www.ooozhi.com/1044.html | 推荐几个Google镜像网站，最新访问谷歌的方法，持续更新中【建议收藏】 _ 我知导航网
edge://newtab/ | 新建标签页
https://www.clinicalmicrobiologyandinfection.com/ | Home Page: Clinical Microbiology and Infection
https://cn.bing.com/search?q=dissertation&form=ANNTH1&refig=411af157586f4cb6ae5047cafdf72270 | dissertation - 搜索
https://cn.bing.com/search?q=pubmed&cvid=2d46409c1c004f73afbbaa795f72233f&aqs=edge.0.0j69i57j0l7.2753j0j4&FORM=ANAB01&PC=HCTS | pubmed - 搜索
https://pubmed.ncbi.nlm.nih.gov/ | PubMed
edge://newtab/ | 新建标签页
https://www.baidu.com/s?wd=Aspergillus&rsv_spt=1&rsv_iqid=0xbd28d6f00018dff6&issp=1&f=8&rsv_bp=1&rsv_idx=2&ie=utf-8&rqlang=cn&tn=88093251_25_hao_pg&rsv_enter=1&rsv_dl=tb&oq=%2526lt%253Bhaetomium&rsv_btype=t&inputT=2993&rsv_t=c88a4dtSVyPHNrmevslthwdqDW3bupehMA1XsCOnD9krOT%2Bi0pVy2b%2B1fw8Bwbves15Ua0GaCyLN&rsv_n=2&rsv_pq=d55e8b2000112c1f&rsv_sug3=4&rsv_sug1=4&rsv_sug7=100&rsv_sug2=0&rsv_sug4=2993 | Aspergillus_百度搜索
https://fanyi.baidu.com/?aldtype=85#en/zh/Cladosporium | 百度翻译-200种语言互译、沟通全世界！
https://cn.bing.com/search?q=python+%e8%bf%9b%e5%ba%a6%e6%9d%a1%e5%ba%93&qs=SC&pq=python+%e8%bf%9b%e5%ba%a6%e6%9d%a1+&sk=SC5&sc=6-11&cvid=CE89897579D340CBB7C0AE185567014E&FORM=QBRE&sp=6 | python 进度条库 - 搜索
https://blog.csdn.net/qianfengpython/article/details/118764070 | (2条消息) Python制作进度条，原来有这么多方法_程序员大新的博客-CSDN博客
https://blog.csdn.net/weixin_30709179/article/details/113497438 | (31条消息) python 项目进度管理_4种常用Python进度条库，学会就是时间管理大师！_颜京凉的博客-CSDN博客
https://www.html.cn/script/python/153558.html | Python Multiprocessing多进程使用tqdm显示进度条的实现 - html中文网
https://blog.csdn.net/CSDN_OWL/article/details/114335467 | (32条消息) tqdm.tqdm()_CSDN_OWL的博客-CSDN博客_tqdm是什么库
https://zhuanlan.zhihu.com/p/470320311 | 比tqdm更好用的进度条工具rich.progress，你值得拥有 - 知乎

https://github.com/datawhalechina/fun-rec/blob/master/docs/%E7%AC%AC%E4%BA%8C%E7%AB%A0%20%E6%8E%A8%E8%8D%90%E7%B3%BB%E7%BB%9F%E5%AE%9E%E6%88%98/2.2%E6%96%B0%E9%97%BB%E6%8E%A8%E8%8D%90%E7%B3%BB%E7%BB%9F%E5%AE%9E%E6%88%98/docs/2.2.1.1%20Mysql%E5%9F%BA%E7%A1%80.md | fun-rec/2.2.1.1 Mysql基础.md at master · datawhalechina/fun-rec
edge://newtab/ | 新建标签页
https://www.baidu.com/?tn=88093251_61_hao_pg | 百度一下，你就知道
https://www.baidu.com/s?wd=%E6%BB%9A%E5%8A%A8%3A%E4%BF%84%E4%B9%8C%E5%B1%80%E5%8A%BF%E6%9C%80%E6%96%B0%E8%BF%9B%E5%B1%95&sa=fyb_n_homepage&rsv_dl=fyb_n_homepage&from=super&cl=3&tn=baidutop10&fr=top1000&rsv_idx=2&hisfilter=1 | 滚动:俄乌局势最新进展_百度搜索
https://baijiahao.baidu.com/s?id=1725625149536159436&wfr=spider&for=pc | 乌克兰总统发推：世界必须迫使俄罗斯实现和平
https://cn.bing.com/search?q=mysql+%E6%95%99%E7%A8%8B+&qs=n&form=QBRE&sp=-1&pq=mysql+%E6%95%99%E7%A8%8B+gitbook&sc=1-16&sk=&cvid=3DBD09574AD8458796E4E7352E0785BB | mysql 教程 - 搜索
https://www.begtut.com/mysql/mysql-tutorial.html | MySQL 教程 | 新手教程
http://www.gitbook.net/mysql/index.html | MySQL在线教程
https://zhuanlan.zhihu.com/p/399141619 | MySQL面试题及答案（2021年MySQL面试题大全带答案） - 知乎
http://studyofnet.com/news/6022.html | mysql数据库基础练习（最全50个Mysql数据库查询练习题） - 开心学习
https://blog.csdn.net/original_recipe/article/details/91958663 | (31条消息) Mysql经典练习题50题_original_recipe的博客-CSDN博客_mysql练习题
https://cn.bing.com/search?q=swim+transformer+%E5%88%86%E7%B1%BB&qs=n&form=QBRE&sp=-1&pq=swim+transformer+fen%27l&sc=0-22&sk=&cvid=FBF614BB39794183B2533768993E0B48 | swim transformer 分类 - 搜索
https://blog.csdn.net/qq_29750461/article/details/120863541 | (31条消息) 【亲测】Swin-Transformer 自定义数据集图像分类_丑小鸭-CSDN博客_swin transformer图像分类
https://github.com/microsoft/Swin-Transformer | microsoft/Swin-Transformer: This is an official implementation for “Swin Transformer: Hierarchical Vision Transformer using Shifted Windows”.
https://blog.csdn.net/qq_39478403/article/details/120042232 | (31条消息) 【机器学习】详解 Swin Transformer (SwinT)_闻韶-CSDN博客_swin transformer
edge://newtab/ | 新建标签页
https://www.baidu.com/?tn=02003390_3_hao_pg | 百度一下，你就知道
https://cn.bing.com/search?q=openslide+github&qs=UT&pq=openslide+&sk=HS2CT2SC2&sc=8-10&cvid=8B95DACCBFA44D46950BF1BBE44B48A6&FORM=QBRE&sp=7 | openslide github - 搜索
https://github.com/openslide/openslide-python/blob/main/openslide/deepzoom.py#L201
http://www.pointborn.com/article/2020/7/21/857.html | Python Openslide 自动切割病理图像 | We all are data.
https://cn.bing.com/search?q=%E4%B8%80%E6%96%87%E7%9C%8B%E8%A7%81%E6%B7%B1%E5%BA%A6%E5%AD%A6%E4%B9%A0+%E6%B3%A8%E6%84%8F%E5%8A%9B%E6%9C%BA%E5%88%B6&qs=n&form=QBRE&sp=-1&pq=%E4%B8%80%E6%96%87%E7%9C%8B%E8%A7%81%E6%B7%B1%E5%BA%A6%E5%AD%A6%E4%B9%A0+%E6%B3%A8%E6%84%8F%E5%8A%9B%E6%9C%BA%E5%88%B6&sc=0-14&sk=&cvid=E6004A0C6DCD4682AC158DF273FA0793 | 一文看见深度学习注意力机制 - 搜索
https://cn.bing.com/search?q=%E4%B8%80%E6%96%87%E7%9C%8B%E5%B0%BD%E6%B7%B1%E5%BA%A6%E5%AD%A6%E4%B9%A0%E7%9A%8420%E7%A7%8D%E5%8D%B7%E7%A7%AF&cvid=d6fd993aad8c43cf9ae5942aba06eee6&aqs=edge.3.69i57j0l3.12126j0j4&FORM=ANAB01&PC=HCTS | 一文看尽深度学习的20种卷积 - 搜索
https://cn.bing.com/search?q=pinbox&cvid=092ae1756e064dff83707c54fe638588&aqs=edge..69i57.2289j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | pinbox - 搜索
http://localhost:53440/a1cea3b8588568ef9e54b5f90f6222ea/?utm_campaign=startup&utm_content=&utm_medium=lantern&utm_source=windows#/
https://withpinbox.com/items | Pinbox - 跨平台收藏工具

https://zentao.medbot.top/pro/my-work-task.html | 地盘-我的任务 - 禅道
https://cn.bing.com/search?q=%E5%9B%BE%E5%83%8F%E5%88%86%E7%B1%BB&form=ANNTH1&refig=b4f52da7f7264fc8bbd1dc3c86f41d52 | 图像分类 - 搜索
https://blog.csdn.net/Avery123123/article/details/106769742 | (30条消息) 计算机视觉—–图像分类综述_Avery123123的博客-CSDN博客_图像分类算法综述
https://www.baidu.com/s?ie=utf-8&f=8&rsv_bp=1&rsv_idx=2&tn=88093251_84_hao_pg&wd=efficientnetv2%20%E8%AE%AD%E7%BB%83%E5%9B%BE%E7%89%87%E5%A4%A7%E5%B0%8F&rsv_spt=1&oq=efficientnetv2%2520%25E5%259B%25BE%25E7%2589%2587%25E5%25A4%25A7%25E5%25B0%258F&rsv_pq=ac7cf53f00195728&rsv_t=2072MJUKQJk8FCAjjnhJJuj3kdiNb9v4mAYj7o2a%2BEN0m9be2x5xLZcAmsYeNQBzWXfgzCVJrb1P&rqlang=cn&rsv_enter=1&rsv_dl=tb&rsv_btype=t&inputT=6730&rsv_sug3=31&rsv_sug1=20&rsv_sug7=100&rsv_sug2=0&rsv_sug4=983599 | efficientnetv2 训练图片大小_百度搜索
https://www.cnblogs.com/deephub/p/15899555.html | EfficientNetV2 - 通过NAS、Scaling和Fused-MBConv获得更小的模型和更快的训练 - deephub - 博客园
https://blog.csdn.net/weixin_47397155/article/details/122673465 | (30条消息) EfficientNetV2简述（图像分类篇）ToolMan萤火虫的博客-CSDN博客_efficientnet图像分类
https://www.jianshu.com/p/0222efa91a6c | EfficientNetV2 - 简书
https://blog.csdn.net/amusi1994/article/details/115419177 | (30条消息) EfficientNetV2: Smaller Models and Faster Training_阿木寺的博客-CSDN博客
https://blog.csdn.net/weixin_47397155/article/details/122673465 | (30条消息) EfficientNetV2简述（图像分类篇）_ToolMan萤火虫的博客-CSDN博客_efficientnet图像分类
https://blog.csdn.net/qq_37541097/article/details/116933569 | (30条消息) EfficientNetV2网络详解_霹雳吧啦Wz-CSDN博客_efficientnetv2
http://www.360doc.com/content/21/0308/22/73546223_965868631.shtml | CVPR2021最新接收论文合集！22个方向100+篇论文汇总｜持续更新
https://blog.csdn.net/amusi1994/article/details/113131743 | (30条消息) CVPR2021最全论文合集_阿木寺的博客-CSDN博客_cvpr论文
https://github.com/amusi/CVPR2021-Papers-with-Code#Image-Classification | amusi/CVPR2021-Papers-with-Code: CVPR 2021 论文和开源项目合集
https://zhuanlan.zhihu.com/p/423816917 | CVPR-2021论文总结 - 知乎
https://openaccess.thecvf.com/CVPR2021?day=all | CVPR 2021 Open Access Repository
https://zhuanlan.zhihu.com/p/354043252 | CVPR 2021 结果出炉！最新600篇CVPR’21论文分方向汇总（更新中） - 知乎
https://github.com/extreme-assistant/CVPR2021-Paper-Code-Interpretation/blob/master/CVPR2021.md#NAS | CVPR2021-Paper-Code-Interpretation/CVPR2021.md at master · extreme-assistant/CVPR2021-Paper-Code-Interpretation
https://github.com/google/automl/blob/master/efficientnetv2/main.py | automl/main.py at master · google/automl
https://cn.bing.com/search?q=Swin+Transformers&cvid=a00cfb5f8f4f454ab8f4d3b4204f950e&aqs=edge..69i57.1079j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | Swin Transformers - 搜索
https://zhuanlan.zhihu.com/p/367111046 | 图解Swin Transformer - 知乎
https://cn.bing.com/search?q=arxiv+%E5%8A%A0%E9%80%9F&cvid=dc03f8fce0f34143a82cbbd407ee255d&aqs=edge..69i57j0l3j69i64.12904j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | arxiv 加速 - 搜索
https://zhuanlan.zhihu.com/p/110217113 | 加速国内访问 Arxiv 论文的一些方法 - 知乎
http://xxx.itp.ac.cn/pdf/2104.00298.pdf | EfficientNetV2: Smaller Models and Faster Training
https://cn.bing.com/search?q=Fused+MBConv&qs=n&form=QBRE&sp=-1&pq=fused+mbconv&sc=0-12&sk=&cvid=7F9414DD46794E269121C9C3C9DA0B05 | Fused MBConv - 搜索
https://zhuanlan.zhihu.com/p/468036014 | 论文推荐：EfficientNetV2 - 通过NAS、Scaling和Fused-MBConv获得更小的模型和更快的训练速度 - 知乎
https://blog.csdn.net/qq_42057562/article/details/118513829 | (30条消息) MBConv模块的注意力机制_申小隆的博客-CSDN博客_mbconv 卷积
https://www.ebaina.com/articles/140000012508 | 【深度学习】移动翻转瓶颈卷积的实现_专栏_易百纳技术社区
https://cn.bing.com/search?q=NAS+%E7%BD%91%E7%BB%9C&qs=n&form=QBRE&sp=-1&pq=nas+%E7%BD%91%E7%BB%9C&sc=0-6&sk=&cvid=674B21DBF388442AAF72BC5533D7BBD8 | NAS 网络 - 搜索
https://zhuanlan.zhihu.com/p/60414004 | NAS（神经结构搜索）综述 - 知乎
https://cn.bing.com/search?q=%e6%b7%b1%e5%ba%a6%e5%ad%a6%e4%b9%a0+%e6%a8%a1%e5%9e%8b+sota+%e6%80%bb%e7%bb%93&qs=n&sp=-1&pq=%e6%b7%b1%e5%ba%a6%e5%ad%a6%e4%b9%a0+%e6%a8%a1%e5%9e%8b+sota+%e6%80%bb%e7%bb%93&sc=0-15&sk=&cvid=7362230C740A40CC98216B862EB6B505&first=10&FORM=PORE | 深度学习模型 sota 总结 - 搜索
https://blog.csdn.net/pxiongw/article/details/122748518 | (30条消息) 2021 CV NLP CTR 多模态深度学习超越SOTA新模型整理_沧海明月的博客-CSDN博客
https://zhuanlan.zhihu.com/p/433086709 | 【总结】推荐系统——精排篇【2】 - 知乎
https://zhuanlan.zhihu.com/p/452019097 | 【MICCAI-2021医学图像分割领域60+篇文章速读概览（文末附总结）】 - 知乎
https://blog.csdn.net/weixin_43509698/article/details/116309952 | (30条消息) 深度学习重要思想及sota模型总结_繁华落尽，寻一世真情的博客-CSDN博客_sota模型
https://cloud.tencent.com/developer/article/1854586 | Transformer+CNN=sota！上限和下限都很高，Facebook AI一个门控就搞定 - 云+社区 - 腾讯云
https://www.zhihu.com/question/322077321 | 大佬们，我刚开始接触图像分割，对于图像分割深度学习这块一头雾水，可以给出一个从零开始学习的路线吗？ - 知乎
https://cloud.tencent.com/developer/article/1944947 | 加速100倍，性能媲美SOTA，浙大提出无数据知识蒸馏新方法FastDFKD - 云+社区 - 腾讯云
https://cloud.tencent.com/developer/article/1537240 | 【ICCV2019教程】手把手带你重现SOTA深度学习模型，附相关资料下载 - 云+社区 - 腾讯云
https://blog.csdn.net/Dream_csd/article/details/115134437 | (30条消息) 查看深度学习SOTA论文和代码网站_Dream_csd的博客-CSDN博客_sota网站
https://paperswithcode.com/task/few-shot-image-classification | Few-Shot Image Classification | Papers With Code
https://zhuanlan.zhihu.com/p/54730387 | 一个新的追踪深度学习state-of-the-art的GitHub项目 - 知乎
https://github.com/cmsflash | cmsflash (Shen Zhuoran)
https://cmsflash.github.io/blog/ | Blog | Shen Zhuoran
https://zhuanlan.zhihu.com/p/84419694 | 15大领域、127个任务，这里有最全的机器学习SOTA模型 - 知乎
https://www.jiqizhixin.com/sota/tech-task/bf907858-3bb4-415a-a439-416fa652285d | 机器之心SOTA
https://cn.bing.com/search?q=Few-Shot+Image+Classification&cvid=a5b491a61ace4a589a0e760657f0328d&aqs=edge..69i57j69i60.953j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | Few-Shot Image Classification - 搜索
https://blog.csdn.net/qq_36104364/article/details/106942299 | (32条消息) 论文阅读笔记《DeepEMD: Few-Shot Image Classification with Differentiable Earth Mover’s Distance》_深视-CSDN博客_deepemd
https://www.msra.cn/zh-cn/news/features/ijcai-2021-uncertainty-aware-few-shot-image-classification | IJCAI 2021 | 不确定性感知小样本图像分类模型，实现SOTA性能
https://cn.bing.com/search?q=%E6%B7%B1%E5%BA%A6%E5%AD%A6%E4%B9%A0yolov5+%E7%89%B9%E5%BE%81%E6%8F%90%E5%8F%96&qs=n&form=QBRE&sp=-1&pq=%E6%B7%B1%E5%BA%A6%E5%AD%A6%E4%B9%A0yolov5+%E7%89%B9%E5%BE%81%E6%8F%90%E5%8F%96&sc=0-15&sk=&cvid=4650C2E267FF4BD78829F4CCC6CE17DA | 深度学习yolov5 特征提取 - 搜索
https://blog.csdn.net/hei_hei_hei/article/details/122080242 | (18条消息) 深度学习笔记_hei_hei_hei_的博客-CSDN博客
https://cn.bing.com/search?q=CNN+%e7%bd%91%e7%bb%9c%e7%bb%93%e6%9e%84+2022&qs=n&sp=-1&pq=cnn+%e7%bd%91%e7%bb%9c%e7%bb%93%e6%9e%84+20&sc=0-11&sk=&cvid=D2D59EE8DC2B4CAE8D078547C7ADBD14&first=101&FORM=PORE | CNN 网络结构 2022 - 搜索
https://new.qq.com/omn/20220226/20220226A044JW00.html | 基于Transformer的人工神经网络，将有机结构的图像转换为分子结构_腾讯新闻
https://www.jiqizhixin.com/articles/2022-02-27-6 | 将大核卷积分三步，清华胡事民、南开程明明团队全新视觉骨干VAN，超越SOTA ViT和CNN | 机器之心
https://blog.51cto.com/u_15069443/2613608 | 经典卷积神经网络（CNN）结构可视化工具_mb5fe948249bc3d的技术博客_51CTO博客
https://javajgs.com/archives/44626 | 卷积神经网络(CNN)张量(图像)的尺寸和参数计算(深度学习)-Java架构师必看
https://blog.csdn.net/Jormungand123/article/details/123196829 | (18条消息) CNN架构_Jormungand123的博客-CSDN博客
https://www.zhihu.com/question/447419811 | 有哪些常用的神经网络模型？ - 知乎
https://www.likecs.com/show-204386802.html | CNN网络结构进化概述 - 爱码网
https://blog.csdn.net/zhuoqingjoking97298/article/details/121766308 | (18条消息) 一文精简介绍CNN基本结构_TSINGHUAJOKING-CSDN博客_cnn结构图
https://www.pinyue84.com/content/53026116.html | 「深度学习」CNN网络结构发展过程_科技通讯_品阅巴士
https://zhuanlan.zhihu.com/p/423320331 | 研一菜鸡图像分类笔记整理——lenet,alexnet,vgg,googlenet,resnet,densenet,,,, - 知乎
https://github.com/tklk610/DenseNet-with-pytorch | tklk610/DenseNet-with-pytorch: 使用pytorch实现DenseNet，完成完整的代码框架，从建立数据集、设置参数、训练网络到推理测试。
https://zhuanlan.zhihu.com/p/385606979 | Pytorch DenseNet训练自己的数据集 - 知乎
https://github.com/bamos/densenet.pytorch | bamos/densenet.pytorch: A PyTorch implementation of DenseNet.
https://github.com/liuzhuang13/DenseNet | liuzhuang13/DenseNet: Densely Connected Convolutional Networks, In CVPR 2017 (Best Paper Award).
https://cn.bing.com/search?q=paperwithcode%E7%BD%91%E7%AB%99&cvid=4b0cee3454c844448c8604397195b43a&aqs=edge.2.69i57j0l7.9663j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | paperwithcode网站 - 搜索
https://paperswithcode.com/sota/image-classification-on-imagenet-real | ImageNet ReaL Benchmark (Image Classification) | Papers With Code
https://cn.bing.com/search?q=resnet+evaluate+pytorch&qs=n&form=QBRE&sp=-1&pq=resnet+evaluate+p%27y%27torch&sc=0-25&sk=&cvid=E898CFB628734486B43D3473CD1770CA | resnet evaluate pytorch - 搜索
https://blog.csdn.net/u014453898/article/details/95337290 | 修改pytorch提供的resnet接口实现Kaggle猫狗识别_ZJE-CSDN博客_pytorch resnet 训练
https://www.cnblogs.com/wangxiaobei2019/p/13475186.html#pytorch_deeppy%E6%96%87%E4%BB%B6%E4%B8%8B%E8%BD%BD | Pytorch_加载自己的数据集并训练 - wangxiaobei2019 - 博客园
https://blog.csdn.net/hhhhhhhhhhwwwwwwwwww/article/details/117374755 | (18条消息) 【图像分类】实战——使用ResNet实现猫狗分类（pytorch）_AI浩-CSDN博客_resnet分类实战
https://blog.csdn.net/qq_38412266/article/details/112868835 | (18条消息) PyTorch Resnet实现图像分类_姚先生97的博客-CSDN博客_resnet实现图像分类
https://zhuanlan.zhihu.com/p/157134695 | 通过Pytorch实现ResNet18 - 知乎
https://zhuanlan.zhihu.com/p/472590934 | resnet的pytorch代码详解 - 知乎
https://www.cnblogs.com/wzyuan/p/9880342.html | 解读 pytorch对resnet的官方实现 - wuzeyuan - 博客园
https://blog.csdn.net/wuzhuoxi7116/article/details/106309943 | (18条消息) ResNet网络详解及Pytorch代码实现（超详细帮助你掌握ResNet原理及实现）_wuzhuoxi7116的博客-CSDN博客_resnet代码pytorch
https://cn.bing.com/search?q=efficientnetv2+pytorch&form=ANNTH1&refig=8213ce17307e4429b885a3a945e79b8e | efficientnetv2 pytorch - 搜索
https://blog.csdn.net/qq_37541097/article/details/116933569 | (18条消息) EfficientNetV2网络详解_霹雳吧啦Wz-CSDN博客_efficientnetv2
https://github.com/WZMIAOMIAO/deep-learning-for-image-processing/tree/master/pytorch_classification/Test11_efficientnetV2 | deep-learning-for-image-processing/pytorch_classification/Test11_efficientnetV2 at master · WZMIAOMIAO/deep-learning-for-image-processing
https://blog.csdn.net/hhhhhhhhhhwwwwwwwwww/article/details/117535310 | (18条消息) 【图像分类】实战——使用EfficientNetV2实现图像分类（Pytorch）_AI浩-CSDN博客_efficientnetv2
https://cn.bing.com/search?q=tensor%E8%BD%ACnumpy+list&qs=n&form=QBRE&sp=-1&pq=tensor%E8%BD%ACnumpy+list&sc=0-17&sk=&cvid=8C1596B5A4E6430DA3C8BB731E940BA2 | tensor转numpy list - 搜索
https://blog.csdn.net/qq_43790749/article/details/100031241 | list,tensor,numpy相互转化_枭志的博客-CSDN博客_tensor转list
https://blog.csdn.net/weixin_40446557/article/details/87893856 | Pytorch-模型的保存和导入_秀呀秀的博客-CSDN博客_pytorch如何保存模型
https://blog.csdn.net/weixin_40522801/article/details/106563354 | Pytorch：模型的保存与加载 torch.save()、torch.load()、torch.nn.Module.load_state_dict()_宁静致远*的博客-CSDN博客_load
https://cn.bing.com/search?q=pytorch+%e5%af%bc%e5%85%a5%e6%a8%a1%e5%9e%8b&FORM=QSRE3 | pytorch 导入模型 - 搜索
https://blog.csdn.net/qq_26884501/article/details/119415714 | pytorch模型的保存导入与推理_DAICONG’s Blog-CSDN博客
edge://newtab/ | 新建标签页
https://cn.bing.com/search?q=pytorch+transforms&qs=n&form=QBRE&sp=-1&pq=pytorch+transform&sc=8-17&sk=&cvid=D3AAAC4BB600491080632D6FE9F67057 | pytorch transforms - 搜索
https://zhuanlan.zhihu.com/p/130985895 | Pytorch数据预处理：transforms的使用方法 - 知乎
https://cn.bing.com/search?q=mobilenetv1+v2&qs=n&form=QBRE&sp=-1&pq=mobilenetv1+v2&sc=1-14&sk=&cvid=82B963D2C0004C2D92B7646E945B8D26 | mobilenetv1 v2 - 搜索
https://zhuanlan.zhihu.com/p/50045821 | MobileNet v1 和 MobileNet v2 - 知乎
https://blog.csdn.net/weixin_39958313/article/details/104578040 | (2条消息) MobileNetV1、V2、V3总结_Dreamer_Du的博客-CSDN博客_mobilenet v1 v2 v3

https://cn.bing.com/search?q=sklearn+split&cvid=53a26fae297144e595eb6eaa8a354b99&aqs=edge..69i57j0.7831j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | sklearn split - 搜索
https://scikit-learn.org/stable/modules/generated/sklearn.model_selection.train_test_split.html | sklearn.model_selection.train_test_split — scikit-learn 1.0.2 documentation
https://zhuanlan.zhihu.com/p/248634166 | sklearn函数：train_test_split（分割训练集和测试集） - 知乎
https://cn.bing.com/search?q=from+Pytorch_learning+import+pytorch_deep+as+pyd&cvid=e76887f15431456f99519cb699b383d1&aqs=edge..69i57.1064j0j4&FORM=ANAB01&PC=HCTS | from Pytorch_learning import pytorch_deep as pyd - 搜索
https://www.cnblogs.com/wangxiaobei2019/p/13330366.html | pytorch_5.5 LeNet - wangxiaobei2019 - 博客园
https://www.cnblogs.com/duanhx/articles/9655228.html | LeNet详解 - CharlesQAQ - 博客园
https://blog.csdn.net/Brikie/article/details/121587196 | 度量学习和pytorch-metric-learning的使用_Brikie的博客-CSDN博客_pytorch_metric_learning
https://blog.csdn.net/qq_34243930/article/details/106695877 | pytorch之多GPU使用——#CUDA_VISIBLE_DEVICES使用 #torch.nn.DataParallel() #报错解决_夏普通-CSDN博客
https://ptorch.com/docs/1/torch-cuda | torch.cuda - pytorch中文网
https://blog.csdn.net/u012927281/article/details/116635043 | torch.device(‘cuda‘) 与 torch.device(‘cuda:0‘) 的区别简析_Andy.Wang的博客-CSDN博客_device=’cuda:0
https://zhuanlan.zhihu.com/p/166161217#:~:text=%E4%BD%BF%E7%94%A8torch.cuda.set_device%20%28%29%E5%8F%AF%E4%BB%A5%E6%9B%B4%E6%96%B9%E4%BE%BF%E5%9C%B0%E5%B0%86%E6%A8%A1%E5%9E%8B%E5%92%8C%E6%95%B0%E6%8D%AE%E5%8A%A0%E8%BD%BD%E5%88%B0%E5%AF%B9%E5%BA%94GPU%E4%B8%8A%2C%20%E7%9B%B4%E6%8E%A5%E5%AE%9A%E4%B9%89%E6%A8%A1%E5%9E%8B%E4%B9%8B%E5%89%8D%E5%8A%A0%E5%85%A5%E4%B8%80%E8%A1%8C%E4%BB%A3%E7%A0%81%E5%8D%B3%E5%8F%AF%20torch.cuda.set_device%20%28gpu_id%29,%23%E5%8D%95%E5%8D%A1%20torch.cuda.set_device%20%28%27cuda%3A%27%2Bstr%20%28gpu_ids%29%29%20%23%E5%8F%AF%E6%8C%87%E5%AE%9A%E5%A4%9A%E5%8D%A1 | 在pytorch中指定显卡 - 知乎
https://cn.bing.com/search?q=torch.cuda+%E5%92%8C+to&qs=n&form=QBRE&sp=-1&pq=torch.cuda+%E5%92%8C+to&sc=0-15&sk=&cvid=F8CBF2EEC347439084BA4A5681FD8640 | torch.cuda 和 to - 搜索
https://www.w3cschool.cn/article/79305038.html | pytorch中to(device) 和cuda()有什么区别？如何使用？ | w3c笔记
https://pytorch-cn.readthedocs.io/zh/latest/package_references/torch-cuda/ | torch.cuda - PyTorch中文文档
https://cn.bing.com/search?q=loss+%E5%A2%9E%E5%A4%A7&cvid=2df53d4594e546db92e0a73a478a741c&aqs=edge..69i57.2769j0j4&FORM=ANAB01&PC=HCTS | loss 增大 - 搜索
https://www.baidu.com/s?wd=torch%20%E4%BF%9D%E5%AD%98%E6%9C%80%E5%A5%BD%E7%9A%84%E6%A8%A1%E5%9E%8B&rsv_spt=1&rsv_iqid=0x84131ff400006886&issp=1&f=8&rsv_bp=1&rsv_idx=2&ie=utf-8&tn=baiduhome_pg&rsv_enter=1&rsv_dl=ib&rsv_n=2&rsv_sug3=1 | torch 保存最好的模型_百度搜索
https://zhuanlan.zhihu.com/p/94337733 | pytorch学习系列（9）：只保存精度最高的模型 - 知乎
https://www.codenong.com/cs106056437/ | Pytorch上在最好的模型上继续训练模型（model 和 optimizer模型如何使用） | 码农家园
https://zhuanlan.zhihu.com/p/174577398 | 谈一谈保存训练精度最优模型的那点事 - 知乎
https://qastack.cn/programming/42703500/best-way-to-save-a-trained-model-in-pytorch | 在PyTorch中保存经过训练的模型的最佳方法？
https://www.cnblogs.com/jiangkejie/p/14606279.html | pytorch保存模型并记录最优模型 - 那抹阳光1994 - 博客园
https://blog.csdn.net/weixin_43593330/article/details/108180092#:~:text=PyTorch%20%E6%8F%90%E4%BE%9B%E4%BA%86%E4%B8%A4%E7%A7%8D%20%E4%BF%9D%E5%AD%98%20%E8%AE%AD%E7%BB%83%E5%A5%BD%E7%9A%84%20%E6%A8%A1%E5%9E%8B%20%E7%9A%84%E6%96%B9%E6%B3%95%E3%80%82%20%E7%AC%AC%E4%B8%80%E7%A7%8D%E6%98%AF%E5%8F%AA%20%E4%BF%9D%E5%AD%98%E6%A8%A1%E5%9E%8B,%28torch.load%20%28PATH%29%29%E7%AC%AC%E4%BA%8C%E7%A7%8D%E6%96%B9%E6%B3%95%20%E4%BF%9D%E5%AD%98%20%E6%95%B4%E4%B8%AA%20%E6%A8%A1%E5%9E%8B%20%EF%BC%9A%23%20%E4%BF%9D%E5%AD%98%20tor | Pytorch保留验证集上最好的模型_我是天才很好-CSDN博客_pytorch保存最佳模型
https://cn.bing.com/search?q=crossentropyloss&cvid=aea8eff776ed4a14a425f5b58a6c2f69&aqs=edge.1.69i57j0l8.13110j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | crossentropyloss - 搜索
https://cn.bing.com/search?q=%e5%9b%be%e5%83%8f%e5%ba%94%e7%94%a8%e6%9c%8d%e5%8a%a1%e5%8c%96&cvid=a8547d9b2783491bb9e92693e098c3da&aqs=edge..69i57.12576j0j4&PC=HCTS&first=16&FORM=PORE | 图像应用服务化 - 搜索

https://zhuanlan.zhihu.com/p/361181741 | 浅谈L2正则化为什么可以缓解过拟合 - 知乎
https://zhuanlan.zhihu.com/p/376000306 | 理解L1, L2正则化的正确姿势 - 知乎
https://zhuanlan.zhihu.com/p/376282535 | 【面试看这篇就够了】L1、L2正则化理解 - 知乎
https://zhuanlan.zhihu.com/p/35356992 | L1正则化与L2正则化 - 知乎
https://zhuanlan.zhihu.com/p/29360425 | 深入理解L1、L2正则化 - 知乎
https://cloud.tencent.com/developer/article/1456966 | 带答案面经分享-L1正则&L2正则 - 腾讯云开发者社区-腾讯云
https://www.cnblogs.com/XDU-Lakers/p/11773552.html | L1与L2正则化的对比及多角度阐述为什么正则化可以解决过拟合问题 - 控球强迫症 - 博客园

https://zhuanlan.zhihu.com/p/73442055 | 经典并发同步模式：生产者-消费者设计模式 - 知乎
https://blog.csdn.net/x541211190/article/details/109405170 | 设计模式之生产者消费者模式_五道口-CSDN博客_设计模式生产者消费者模式
https://blog.csdn.net/CHINACR07/article/details/78825351?utm_medium=distribute.pc_relevant.none-task-blog-2%7Edefault%7EBlogCommendFromBaidu%7Edefault-17.control&depth_1-utm_source=distribute.pc_relevant.none-task-blog-2%7Edefault%7EBlogCommendFromBaidu%7Edefault-17.control | 设计模式之生成者与消费者模式_ChinaCR07的博客-CSDN博客_生成消费模式
https://www.jb51.net/article/166648.htm | 详细介绍Python进度条tqdm的使用_python_脚本之家
http://www.obgy.cn/diagnose-fuke/5818179baa3566f411ef1e5a | 最新指南：阴道微生态评价的临床应用专家共识-中国妇产科网
https://www.cnblogs.com/davis12/p/14456227.html#:~:text=%E6%88%91%E4%BB%AC%20docker%20run%20%E5%90%AF%E5%8A%A8%E5%AE%B9%E5%99%A8%E7%9A%84%E6%97%B6%E5%80%99%EF%BC%8C%E5%B8%B8%E9%9C%80%E8%A6%81%E5%B0%86%E5%85%B6%E5%9C%A8%E5%90%8E%E5%8F%B0%E8%BF%90%E8%A1%8C%EF%BC%8C%E9%80%9A%E5%B8%B8%E6%88%91%E4%BB%AC%E8%AE%BE%E7%BD%AE%E5%8F%82%E6%95%B0%20-d%20%E5%8D%B3%E5%8F%AF%E3%80%82%20%E4%BD%86%E5%90%8E%E5%8F%B0%E8%BF%90%E8%A1%8C%EF%BC%8C%E5%85%B6%E5%AE%9E%E6%98%AF%E6%9C%89%E5%89%8D%E6%8F%90%E7%9A%84%EF%BC%8C%E5%A6%82%E6%9E%9C%E6%B2%A1%E6%9C%89%E5%89%8D%E5%8F%B0%E8%BF%9B%E7%A8%8B%EF%BC%8C%E9%82%A3%E4%B9%88%E5%AE%9E%E9%99%85%E8%BF%90%E8%A1%8C%E5%AE%8C,docker%20run%20%E5%91%BD%E4%BB%A4%E5%90%8E%EF%BC%8C%E4%BC%9A%E5%A4%84%E4%BA%8E%E9%80%80%E5%87%BA%E7%8A%B6%E6%80%81%EF%BC%8C%E5%8D%B3%20exited%20%E3%80%82%20Docker%E5%AE%B9%E5%99%A8%E5%90%8E%E5%8F%B0%E8%BF%90%E8%A1%8C%EF%BC%8C%E5%BF%85%E9%A1%BB%E6%9C%89%E4%B8%80%E4%B8%AA%E5%89%8D%E5%8F%B0%E8%BF%9B%E7%A8%8B%E3%80%82%20%E5%AE%B9%E5%99%A8%E8%BF%90%E8%A1%8C%E7%9A%84%E5%91%BD%E4%BB%A4%E5%A6%82%E6%9E%9C%E4%B8%8D%E6%98%AF%E9%82%A3%E4%BA%9B%E4%B8%80%E7%9B%B4%E6%8C%82%E8%B5%B7%E7%9A%84%E5%91%BD%E4%BB%A4%EF%BC%88%E6%AF%94%E5%A6%82%E8%BF%90%E8%A1%8Cping%EF%BC%8Csleep%EF%BC%89%EF%BC%8C%E5%B0%B1%E6%98%AF%E4%BC%9A%E8%87%AA%E5%8A%A8%E9%80%80%E5%87%BA%E7%9A%84%E3%80%82 | docker容器之后台运行 - Marathon-Davis - 博客园
https://www.baidu.com/s?wd=%E8%AF%AD%E9%9B%80&rsv_spt=1&rsv_iqid=0xf4014b9e00004c0c&issp=1&f=3&rsv_bp=1&rsv_idx=2&ie=utf-8&tn=44004473_27_oem_dg&rsv_enter=1&rsv_dl=ts_0&rsv_sug3=6&rsv_sug1=6&rsv_sug7=101&rsv_sug2=0&rsv_btype=i&prefixsug=yuque&rsp=0&inputT=2443&rsv_sug4=3409 | 语雀_百度搜索
https://zhuanlan.zhihu.com/p/374853843 | 404 - 知乎
https://cn.bing.com/search?q=sold2&cvid=7266c34eacbe401aa10706b9ebbc92d3&aqs=edge..69i57.2767j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | sold2 - 搜索
edge://newtab/ | 新建标签页

https://github.com/KevinMusgrave/pytorch-metric-learning/blob/master/src/pytorch_metric_learning/trainers/base_trainer.py | pytorch-metric-learning/base_trainer.py at master · KevinMusgrave/pytorch-metric-learning
edge://newtab/ | 新建标签页
https://www.baidu.com/s?wd=pytorch%20trainer%20%E5%AE%9E%E7%8E%B0&rsv_spt=1&rsv_iqid=0xa5f492f400062fc3&issp=1&f=8&rsv_bp=1&rsv_idx=2&ie=utf-8&rqlang=cn&tn=88093251_95_hao_pg&rsv_enter=1&rsv_dl=tb&rsv_t=393fAtsKLkWP2DmzyrpOdf8CIwYoeE4Y7162hT%2BQhQEu5kVxOAE7pV5Om9BSEzTRx6b7Zvak%2BBM7&rsv_sug1=5&rsv_sug7=100&oq=pytorch%2520trainer&rsv_pq=994e80b2001212b3&rsv_sug3=14&rsv_sug2=0&rsv_btype=t&inputT=63038&rsv_sug4=63124 | pytorch trainer 实现_百度搜索
https://zhuanlan.zhihu.com/p/414843341 | Pytorch封装一个优雅的Trainer类，并实现可扩展性 - 知乎
https://zhuanlan.zhihu.com/p/449181811 | 300行代码实现一个优雅的PyTorch Trainer - 知乎
https://cn.bing.com/search?q=pytorch+trainer&qs=n&form=QBRE&sp=-1&pq=pytorch+trainer&sc=1-15&sk=&cvid=A21A177FA46647FABA10173650FF0949 | pytorch trainer - 搜索
https://www.jianshu.com/p/c88df856dbc8 | Pytorch trainer 建议写法 - 简书
https://zhuanlan.zhihu.com/p/198982185 | keras train_on_batch详解 - 知乎
https://blog.csdn.net/huaxiaoxin1/article/details/107510389 | Pytorch节省显存、加速训练的小技巧_huaxiaoxin1的博客-CSDN博客_pytorch节省显存
https://jishuin.proginn.com/p/763bfbd31a84 | 加速 PyTorch 模型训练的 9 个技巧-技术圈
https://zhuanlan.zhihu.com/p/345279772 | 加快PyTorch训练速度！掌握这17种方法，让你省时省力！ - 知乎
https://github.com/Megvii-Nanjing/ML-GCN/blob/master/engine.py | ML-GCN/engine.py at master · Megvii-Nanjing/ML-GCN
https://cn.bing.com/search?q=predicting+fluo+unlabeled+images&form=ANNTH1&refig=15f3de1cb9ce451aa774df4b29a81f94 | predicting fluo unlabeled images - 搜索
edge://newtab/ | 新建标签页
https://www.baidu.com/?tn=44004473_21_oem_dg | 百度一下，你就知道

https://github.com/ultralytics/yolov5/blob/cce313d67d5a8169c3508e08ec52dae3602e439d/data/coco.yaml | yolov5/coco.yaml at cce313d67d5a8169c3508e08ec52dae3602e439d · ultralytics/yolov5
https://cn.bing.com/search?q=yolov5%E8%AE%AD%E7%BB%83%E8%87%AA%E5%B7%B1%E7%9A%84%E6%95%B0%E6%8D%AE%E9%9B%86&cvid=218eb3006fcc477a91c97dcf1bf44c83&aqs=edge.4.69i59l2j0j69i59j0j69i60l3j69i65.5176j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | yolov5训练自己的数据集 - 搜索
https://zhuanlan.zhihu.com/p/448525505 | YOLOv5超详细的入门级教程（训练篇）——训练自己的数据集 - 知乎
https://www.baidu.com/s?wd=docker%20load%20import%20save%20export&rsv_spt=1&rsv_iqid=0xc58ee61f0017e317&issp=1&f=8&rsv_bp=1&rsv_idx=2&ie=utf-8&rqlang=cn&tn=baiduhome_pg&rsv_enter=1&rsv_dl=tb&rsv_t=82a27ILnJHXYgWcgpxygVyfWHNlX%2Fr52z1242ZUi26cBC9cHwaoldPIe4%2BaOTEhPl1VP&oq=docker%2520load%2520import&rsv_btype=t&rsv_pq=d5c2eedd0003e654&inputT=7276&rsv_sug3=69&rsv_sug1=52&rsv_sug7=100&rsv_sug2=0&rsv_sug4=7494 | docker load import save export_百度搜索
https://www.cnblogs.com/qq931399960/p/11379252.html | docker save/load以及export/import使用测试 - 大坑水滴 - 博客园
https://www.cnblogs.com/sunsky303/p/11024672.html | docker save load export import的区别 - sunsky303 - 博客园
https://www.cnblogs.com/linjiqin/p/8604756.html | Docker 本地导入镜像/保存镜像/载入镜像/删除镜像 - Ruthless - 博客园
https://www.kancloud.cn/woshigrey/docker/945800 | 导入镜像的 docker load 命令 · Docker命令手册 · 看云
https://www.cnblogs.com/fansys/articles/10831307.html | WSL运行docker容器 - 雪域熊猫 - 博客园
https://blog.csdn.net/awodwde/article/details/122047721 | (29条消息) Docker的常用命令（使用wsl）_awodwde的博客-CSDN博客
https://blog.csdn.net/zhengun/article/details/84916780 | (29条消息) docker启动服务命令_zhengun的专栏-CSDN博客
https://www.jianshu.com/p/4912f7ae8331 | Docker service 启动关闭 - 简书

https://cn.bing.com/search?q=tnesorflow+%E9%9D%A2%E8%AF%95%E9%A2%98&cvid=ab9c762f88914cbc8e9c182a29905d60&aqs=edge..69i57.8431j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | tnesorflow 面试题 - 搜索
http://www.srcmini.com/33181.html | 最新推荐！TensorFlow面试题及其答案合集 - 面试题 - srcmini
https://cn.bing.com/search?q=%E5%A6%87%E7%A7%91%E5%BE%AE%E7%94%9F%E6%80%81++%E6%BB%B4%E8%99%AB+%E8%8D%A7%E5%85%89&qs=n&form=QBRE&sp=-1&pq=%E5%A6%87%E7%A7%91%E5%BE%AE%E7%94%9F%E6%80%81+%E6%BB%B4%E8%99%AB+%E8%8D%A7%E5%85%89&sc=0-11&sk=&cvid=26B7618483464119BADB9881DD5291EF | 妇科微生态滴虫荧光 - 搜索
https://www.qdsk.cn/proDetail-2604.html | 三凯医学妇科双重荧光染色液，妇科荧光染色液，双重荧光染色液
https://www.sohu.com/a/359656949_100102218 | 妇科检查报告单看不懂？看这里就够了！_阴道
http://www.obgy.cn/diagnose-fuke/5818179baa3566f411ef1e5a | 最新指南：阴道微生态评价的临床应用专家共识-中国妇产科网
https://zhuanlan.zhihu.com/p/409947864 | 知识篇——阴道微生态报告解读 - 知乎
https://news.medlive.cn/obgyn/info-progress/show-138577_206.html | 大数据在妇产科微生态领域中的应用现状及展望_大数据在妇产科微生态领域中的应用现状及展望_医脉通
http://www.tmicrobe.com/exchange/id/11 | Nature Medicine|不同种族妊娠期女性阴道微生物组的动态变化（IF:32.621） - 学术交流 - 苏州图灵微生物科技有限公司
https://zhuanlan.zhihu.com/p/115863445 | 阴道微生态报告怎么看？解密阴道微生态，让你远离妇科炎症 - 知乎
https://www.baidu.com/s?wd=%E5%85%8D%E7%96%AB%E8%8D%A7%E5%85%89%E6%9F%93%E8%89%B2%E6%8A%80%E6%9C%AF&rsv_spt=1&rsv_iqid=0xd236d4a500052e03&issp=1&f=8&rsv_bp=1&rsv_idx=2&ie=utf-8&tn=88093251_85_hao_pg&rsv_enter=1&rsv_dl=ib&rsv_sug3=3&rsv_sug1=1&rsv_sug7=001&rsv_n=2 | 免疫荧光染色技术_百度搜索
https://cn.bing.com/search?q=%E6%BB%B4%E8%99%AB+%E5%85%8D%E7%96%AB%E8%8D%A7%E5%85%89&qs=n&form=QBRE&sp=-1&pq=%E6%BB%B4%E8%99%AB+mian%27yi%E8%8D%A7%E5%85%89&sc=0-12&sk=&cvid=DD0459F6775F44CAB05808AB42818324 | 滴虫免疫荧光 - 搜索
https://max.book118.com/html/2017/0320/96187324.shtm#:~:text=%E5%85%8D%E7%96%AB%E8%8D%A7%E5%85%89%E6%B3%95%E6%A3%80%E6%B5%8B%E9%98%B4%E9%81%93%E6%AF%9B%E6%BB%B4%E8%99%AB%E4%B9%9F%E5%B7%B2%E8%A2%AB%E5%B9%BF%E6%B3%9B%E7%A0%94%E7%A9%B6%EF%BC%8C%E8%AF%A5%E6%B3%95%E6%95%8F%E6%84%9F%E6%80%A7%E7%BA%A6%E4%B8%BA83%25%EF%BD%9E86%25,%EF%BC%8C%E5%8D%95%E5%85%8B%E9%9A%86%E6%8A%97%E4%BD%93%E7%89%B9%E5%BC%82%E6%80%A7%E9%AB%98%EF%BC%8C%E6%9C%89%E6%9C%9B%E6%9B%BF%E4%BB%A3%E5%9F%B9%E5%85%BB%E6%B3%95%E6%88%90%E4%B8%BA%E4%B8%B4%E5%BA%8A%E5%BF%AB%E9%80%9F%E8%AF%8A%E6%96%AD%E7%9A%84%E7%BB%8F%E6%B5%8E%E3%80%81%E5%AE%9E%E7%94%A8%E6%96%B9%E6%B3%95%E3%80%82 | 免疫荧光技术在生殖感染检验领域的应用价值.doc-原创力文档
https://cn.bing.com/search?q=%E6%BB%B4%E8%99%AB&qs=n&form=QBRE&sp=-1&pq=%E6%BB%B4%E8%99%AB&sc=8-2&sk=&cvid=73390B2732724740995C56B50602067C | 滴虫 - 搜索
https://baike.baidu.com/item/%E6%BB%B4%E8%99%AB/6304230 | 滴虫_百度百科
https://xuewen.cnki.net/R2006120340000008.html | 毛滴虫与滴虫病_知网百科
http://www.cogonline.com/info/132585179439468287 | 细菌性阴道病诊治指南（2021修订版）
https://www.sohu.com/a/448157420_120054526 | 细菌性阴道病诊治指南（2021修订版）_诊断
https://cn.bing.com/search?q=VFocal&cvid=8a0d2cb5c7d0454fb8845699d04c487c&aqs=edge..69i57.3998j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | VFocal - 搜索
https://cn.bing.com/search?q=sci-hub&cvid=f23da524cf0c4213ac81c7d485addeed&aqs=edge.1.69i57j0l8.2444j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | sci-hub - 搜索
https://www.scihub.net.cn/ | Sci-Hub:SciHub科研学术网址导航
https://www.scihub.net.cn/sci-hub/ | 文献检索-SciHub
https://sci-hub.se/10.1038/s41591-019-0465-8 | Sci-Hub | Racioethnic diversity in the dynamics of the vaginal microbiome during pregnancy. Nature Medicine | 10.1038/s41591-019-0465-8

https://github.com/opconty/pytorch_ctpn/blob/89ea784d7776a08962c6d5bc5591730e15156d07/ctpn_utils.py#L289 | pytorch_ctpn/ctpn_utils.py at 89ea784d7776a08962c6d5bc5591730e15156d07 · opconty/pytorch_ctpn
https://cn.bing.com/search?q=ctpn+github&cvid=d4d12a077f774cc1847405ba4a0355b4&aqs=edge..69i57j69i60l3.5341j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | ctpn github - 国内版 Bing
https://github.com/tianzhi0549/CTPN/blob/master/src/text_proposal_graph_builder.py | CTPN/text_proposal_graph_builder.py at master · tianzhi0549/CTPN
https://cn.bing.com/search?q=get_successions&cvid=81092aa7e45f43d398151b33b4414ead&aqs=edge..69i57.824j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | get_successions - 国内版 Bing
https://www.cnblogs.com/skyfsm/p/10054386.html | 【OCR技术系列之六】文本检测CTPN的代码实现 - Madcola - 博客园
https://cn.bing.com/search?q=CRNN+GitHub&cvid=cc375d66a3ab4386bbef37623806936a&aqs=edge..69i57j69i60l3.4877j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | CRNN GitHub - 国内版 Bing
https://github.com/xmy0916/pytorch_crnn | xmy0916/pytorch_crnn: 基于pytorch写的CRNN文字识别简化写法帮助入门
https://blog.csdn.net/qq_37668436/article/details/113642808 | (26条消息) 从零写CRNN文字识别 —— （1）准备工作_只会git clone的程序员的博客-CSDN博客
https://cn.bing.com/search?q=CTCLoss+%E5%8E%9F%E7%90%86&qs=n&form=QBRE&sp=-1&pq=ctcloss+%E5%8E%9F%E7%90%86&sc=0-10&sk=&cvid=1D9FA7FEB1C241F18822E9386A96EA8A | CTCLoss 原理 - 国内版 Bing
https://blog.csdn.net/yifen4234/article/details/80334516 | (28条消息) CTC loss 理解_体重51.00-CSDN博客_ctc loss
https://blog.csdn.net/jackytintin/article/details/79425866 | (28条消息) 【Learning Notes】CTC 原理及实现_丁丁的博客-CSDN博客_ctc 实现
https://distill.pub/2017/ctc/ | Sequence Modeling with CTC
https://blog.csdn.net/justsolow/article/details/105251789 | (28条消息) CTCloss从理论到训练_justsolow的博客-CSDN博客_ctc loss
https://runebook.dev/zh-CN/docs/pytorch/generated/torch.nn.ctcloss | PyTorch - CTCLoss - 连接主义时空分类的损失。计算连续（非分段）时间序列与目标序列之间的损失。CTCLoss计算输入与目标可能对齐的概率之和，从而产生相对于每个输入节点可微分的损耗 - 中文
https://zhuanlan.zhihu.com/p/23293860 | CTC实现——compute ctc loss（1） - 知乎
https://blog.csdn.net/fanhenghui/article/details/102687201 | (28条消息) Pytorch使用CRNN CTCLoss实现OCR系统_fanhenghui的专栏-CSDN博客_pytorch实现ocr
https://zhuanlan.zhihu.com/p/33464788#::text=%E5%9C%A8pytorch,%E8%BF%98%E4%BF%9D%E5%AD%98%E4%BA%86%E5%85%B6%E6%A2%AF%E5%BA%A6%E3%80%82 | 基于CTC的语音识别基础与实现 - 知乎
https://github.com/baidu-research/warp-ctc | baidu-research/warp-ctc: Fast parallel CTC.
https://zhuanlan.zhihu.com/p/67415439 | 如何优雅的使用pytorch内置torch.nn.CTCLoss的方法 - 知乎
http://ailab.aiwin.org.cn/competitions/65#phases | AIWIN - 比赛
https://blog.csdn.net/qq_24819773/article/details/104605994 | (29条消息) 【项目实践】中文文字检测与识别项目（CTPN+CRNN+CTC Loss原理讲解）_ChaucerG的博客-CSDN博客_crnn ctc

https://cn.bing.com/search?q=notion+page%E7%9B%B8%E4%BA%92&qs=n&form=QBRE&sp=-1&pq=notion+page%E7%9B%B8%E4%BA%92&sc=0-13&sk=&cvid=10BF0DA92BCB4957AE402F14877FAB22 | notion page相互 - 国内版 Bing
https://cn.bing.com/search?q=Oral+Health+%E7%BF%BB%E8%AF%91&qs=n&form=QBRE&sp=-1&pq=oral+health+%E7%BF%BB%E8%AF%91&sc=0-14&sk=&cvid=A10F0D45144D4727994C9E4ACD3D03CA | Oral Health 翻译 - 国内版 Bing
https://funletu.com/newsflashes/12486.html | Notion网页版中文版 – Funletu
https://www.notion.so/Copy-of-ff27c383193d4023b9099e2ae9c70d88?showMoveTo=true | Copy of 康奈尔笔记
https://www.zhihu.com/question/377868821 | notion什么时候可中文啊…? - 知乎
https://linmi.cc/notiongroup | 加入 Notion 中文社区社群 - Linmi
https://post.smzdm.com/p/amxzrwdk/ | Notion中文客户端汉化版教程_软件应用_什么值得买
https://www.notion.so/Copy-of-ff27c383193d4023b9099e2ae9c70d88 | Copy of 康奈尔笔记
https://www.notionwizard.com/download-notion/#google_vignette | Download Notion: Mac and Windows
https://zhuanlan.zhihu.com/p/440773804 | Notion中文客户端汉化版教程 - 知乎
https://pan.baidu.com/s/185CK_tlr0_IEwxe35xgNvw | Notion客户端汉化文件.zip_免费高速下载|百度网盘-分享无限制
https://zhuanlan.zhihu.com/p/397634631 | 【notion中文汉化方法】真实有效，成功率100% - 知乎
https://greasyfork.org/zh-CN/scripts/396279-notion-translate | Notion-Translate
https://cn.bing.com/search?q=Attention_ocr&cvid=6bcfa7f45a274518acebc2663714b037&aqs=edge..69i57j0l8.1350j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | Attention_ocr - 国内版 Bing
https://www.ptorch.com/news/228.html | Pytorch AttentionOCR 中文端到端的文字识别(Attention_ocr.pytorch) - pytorch中文网
https://github.com/chenjun2hao/Attention_ocr.pytorch | chenjun2hao/Attention_ocr.pytorch: This repository implements the the encoder and decoder model with attention model for OCR
https://zhuanlan.zhihu.com/p/138589087 | 自然场景文本检测识别 - Attention OCR - 知乎
https://github.com/search?q=Attention_ocr.pytorch | Search · Attention_ocr.pytorch

https://github.com/xmy0916/pytorch_crnn | xmy0916/pytorch_crnn: 基于pytorch写的CRNN文字识别~简化写法帮助入门
https://blog.csdn.net/jackytintin/article/details/79425866 | (28条消息) 【Learning Notes】CTC 原理及实现_丁丁的博客-CSDN博客_ctc 实现
https://blog.csdn.net/justsolow/article/details/105251789 | (28条消息) CTCloss从理论到训练_justsolow的博客-CSDN博客_ctc loss
https://zhuanlan.zhihu.com/p/23293860 | CTC实现——compute ctc loss（1） - 知乎
https://zhuanlan.zhihu.com/p/67415439 | 如何优雅的使用pytorch内置torch.nn.CTCLoss的方法 - 知乎
http://ailab.aiwin.org.cn/competitions/65#phases | AIWIN - 比赛
edge://newtab/ | 新建标签页
https://cn.bing.com/search?q=ocr+%E7%BB%BC%E8%BF%B0&cvid=af3fafac95594f9c9ddc353a804b5d12&aqs=edge.3.0j69i57j0j69i59j0l2j69i61j69i60l2.4237j0j4&FORM=ANAB01&PC=HCTS | ocr 综述 - 搜索
https://zhuanlan.zhihu.com/p/428777199 | OCR综述 (持续更新) - 知乎
https://zhuanlan.zhihu.com/p/81876053 | OCR综述论文笔记 - 知乎

https://cn.bing.com/search?q=%E8%BF%88%E7%A7%91%E9%BE%99%E9%98%B4%E9%81%93%E6%BB%B4%E8%99%AB%E5%85%8D%E7%96%AB&cvid=198f436614a54153abc03e460a8f30a7&aqs=edge..69i57.18382j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | 迈科龙阴道滴虫免疫 - 国内版 Bing
https://www.docin.com/p-514920378.html | 免疫荧光试剂盒临床应用.ppt - 豆丁网Docin
https://cn.bing.com/search?q=%E6%95%8F%E6%84%9F%E6%80%A7%E5%92%8C%E7%89%B9%E5%BC%82%E6%80%A7&form=ANNTH1&refig=cba65f2970c94facab944d6cd328850e | 敏感性和特异性 - 国内版 Bing
https://www.jianshu.com/p/22d947ffb71e | 敏感性和特异性（sensitivity and specificity） - 简书
https://cn.bing.com/search?q=%E7%A7%81%E7%AB%8B%E5%9B%BD%E9%99%85%E5%AD%A6%E6%A0%A1+%E8%AF%BE%E7%A8%8B&form=ANNTH1&refig=ce56e2c5679b4a0193c5b2c1d1614a24 | 私立国际学校课程 - 国内版 Bing
https://cn.bing.com/search?q=%E4%B8%80%E8%87%B4%E6%80%A7%E7%B3%BB%E6%95%B0kappa&cvid=a2912c83e3ee4b519eb7e3e4f23b4b6a&aqs=edge..69i57.7061j0j4&FORM=ANAB01&PC=HCTS | 一致性系数kappa - 国内版 Bing
https://zhuanlan.zhihu.com/p/142954589 | 【一致性检验指标】Kappa(cappa)系数 - 知乎
https://zindi.africa/competitions/hulkshare-recommendation-algorithm-challenge | Hulkshare Recommendation Algorithm Challenge - Zindi
https://zhuanlan.zhihu.com/p/48300815 | [LaTeX 绘图指南 - 001] TikZ 的简介、资源以及学习方法 - 知乎

https://paddlepedia.readthedocs.io/en/latest/tutorials/computer_vision/OCR/OCR.html | OCR综述 — PaddleEdu documentation
https://blog.csdn.net/weixin_44119362/article/details/114289607 | (28条消息) 一点就分享系列（实践篇3-上篇）— 修改YOLOV5 之”魔刀小试“+ Trick心得分享+V5精髓部分源码解读_啥都会一点的小程的博客-CSDN博客_yolov5魔改
https://blog.csdn.net/Yong_Qi2015/article/details/122206955 | (28条消息) YOLO-Z | 记录修改YOLOv5以适应小目标检测的实验过程_3D视觉工坊-CSDN博客
https://cn.bing.com/search?q=yolov5+%E4%BF%AE%E6%94%B9neck&cvid=28b4685fe4bd4188a824ac70bc25527c&aqs=edge..69i57.17694j0j1&pglt=41&FORM=ANSPA1&PC=HCTS | yolov5 修改neck - 搜索
https://www.bilibili.com/video/av498627710 | 【yolov5】【第三期】小白手把手教你修改yolov5网络结构，看完就会动手改_哔哩哔哩_bilibili
https://blog.csdn.net/m0_61899108/article/details/122800871 | (28条消息) Yolov5总结文档(理论、代码、实验结果)_m0_61899108的博客-CSDN博客
https://zhuanlan.zhihu.com/p/451080127 | 又改YOLO | 项目如何改进YOLOv5？这篇告诉你如何修改让检测更快、更稳！！！ - 知乎
https://blog.csdn.net/wa1tzy/article/details/114492726 | (28条消息) YOLOv5系列（3）——YOLOv5修改网络结构_wa1tzy的博客-CSDN博客_yolov5修改网络结构
https://cn.bing.com/search?q=bi-fpn+github&qs=n&form=QBRE&sp=-1&pq=bi-fpn+github&sc=0-13&sk=&cvid=4A1526C075984798A5B374C2DB09C78B | bi-fpn github - 搜索
https://github.com/kentaroy47/efficientdet.pytorch/blob/master/BiFPN.py | efficientdet.pytorch/BiFPN.py at master · kentaroy47/efficientdet.pytorch
https://zhuanlan.zhihu.com/p/438719378 | Bi FPN - 知乎
https://zhuanlan.zhihu.com/p/430181461 | 目标检测之EfficientDet（BiFPN） - 知乎
https://cn.bing.com/search?q=efficientnet&qs=n&form=QBRE&sp=-1&pq=efficientnet&sc=9-12&sk=&cvid=52790B29BD5F40829B2B84EDF24B68D6 | efficientnet - 搜索
https://zhuanlan.zhihu.com/p/67508423 | 谷歌大脑：EfficientNet–阅读笔记 - 知乎
https://zhuanlan.zhihu.com/p/366738106 | 【论文解读】一文看懂EfficientnetB0~B7模型所有细节 - 知乎
https://blog.csdn.net/qq_34107425/article/details/104151460 | (28条消息) 【一看就懂】EfficientNet详解。凭什么EfficientNet号称当今最强？_忽逢桃林的博客-CSDN博客_efficientnet
https://cloud.tencent.com/developer/article/1610524 | EfficientNet详解：当前最强网络 - 云+社区 - 腾讯云
https://zhuanlan.zhihu.com/p/111115509 | EfficientNet-B0解读 - 知乎
https://blog.csdn.net/qq_37541097/article/details/114434046 | (28条消息) EfficientNet网络详解_霹雳吧啦Wz-CSDN博客_efficient网络
https://cn.bing.com/search?q=%e7%bb%86%e7%b2%92%e5%ba%a6%e5%9b%be%e5%83%8f%e5%88%86%e7%b1%bb&qs=LT&pq=%e7%bb%86%e7%b2%92%e5%ba%a6&sk=LT2&sc=10-3&cvid=EFC5B5A478014B39855E68284545EC7C&FORM=QBRE&sp=3 | 细粒度图像分类 - 搜索
https://cn.bing.com/search?q=%E7%BB%86%E7%B2%92%E5%BA%A6%E7%9B%AE%E6%A0%87%E8%AF%86%E5%88%AB&qs=n&form=QBRE&sp=-1&pq=%E7%BB%86%E7%B2%92%E5%BA%A6mu%27b%E8%AF%86%E5%88%AB&sc=1-9&sk=&cvid=8C6976560147483096BA52B455F0BAF6 | 细粒度目标识别 - 搜索
https://cn.bing.com/search?q=%E7%94%9F%E6%88%90%E5%AF%B9%E6%8A%97%E7%BD%91%E7%BB%9C+%E7%9F%A5%E4%B9%8E%E4%B8%93%E6%A0%8F&qs=n&form=QBRE&sp=-1&pq=%E7%94%9F%E6%88%90%E5%AF%B9%E6%8A%97%E7%BD%91%E7%BB%9C+%E7%9F%A5%E4%B9%8E%E4%B8%93%E6%A0%8F&sc=0-11&sk=&cvid=2BC1486C60904A4A85F099B0B0592349 | 生成对抗网络知乎专栏 - 搜索
https://zhuanlan.zhihu.com/p/261871560 | 生成对抗网络（GAN）的数学原理全解 - 知乎
https://zhuanlan.zhihu.com/p/272746911 | 白话生成对抗网络GAN及代码实现 - 知乎
https://zhuanlan.zhihu.com/p/34287744 | 生成对抗网络（GAN） - 知乎
https://cn.bing.com/search?q=python+%e5%a4%9a%e8%bf%9b%e7%a8%8b+%e5%88%9b%e5%bb%ba%e6%96%87%e4%bb%b6%e5%a4%b9%e6%8a%a5%e9%94%99file+exists&qs=n&sp=-1&pq=python+%e5%a4%9a%e8%bf%9b%e7%a8%8b+%e5%88%9b%e5%bb%ba%e6%96%87%e4%bb%b6%e5%a4%b9%e6%8a%a5%e9%94%99file+exists&sc=0-29&sk=&cvid=92FF27EEE8C9406A978A49CF559392FF&first=11&FORM=PORE | python 多进程创建文件夹报错file exists - 搜索
https://blog.csdn.net/weixin_42560937/article/details/106077014 | (28条消息) python os.path.isdir/exists函数在多线程中误判，导致os.mkdir函数执行报错_出来吧大毛球的博客-CSDN博客
edge://newtab/ | 新建标签页
https://www.baidu.com/?tn=88093251_56_hao_pg | 百度一下，你就知道

https://cn.bing.com/search?q=%E8%BF%88%E7%A7%91%E9%BE%99%E9%98%B4%E9%81%93%E6%BB%B4%E8%99%AB%E5%85%8D%E7%96%AB&cvid=198f436614a54153abc03e460a8f30a7&aqs=edge..69i57.18382j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | 迈科龙阴道滴虫免疫 - 国内版 Bing
https://www.docin.com/p-514920378.html | 免疫荧光试剂盒临床应用.ppt - 豆丁网Docin
https://cn.bing.com/search?q=%E6%95%8F%E6%84%9F%E6%80%A7%E5%92%8C%E7%89%B9%E5%BC%82%E6%80%A7&form=ANNTH1&refig=cba65f2970c94facab944d6cd328850e | 敏感性和特异性 - 国内版 Bing
https://www.jianshu.com/p/22d947ffb71e | 敏感性和特异性（sensitivity and specificity） - 简书
https://cn.bing.com/search?q=%E7%A7%81%E7%AB%8B%E5%9B%BD%E9%99%85%E5%AD%A6%E6%A0%A1+%E8%AF%BE%E7%A8%8B&form=ANNTH1&refig=ce56e2c5679b4a0193c5b2c1d1614a24 | 私立国际学校课程 - 国内版 Bing
https://cn.bing.com/search?q=%E4%B8%80%E8%87%B4%E6%80%A7%E7%B3%BB%E6%95%B0kappa&cvid=a2912c83e3ee4b519eb7e3e4f23b4b6a&aqs=edge..69i57.7061j0j4&FORM=ANAB01&PC=HCTS | 一致性系数kappa - 国内版 Bing
https://zhuanlan.zhihu.com/p/142954589 | 【一致性检验指标】Kappa(cappa)系数 - 知乎
edge://newtab/ | 新建标签页
https://pc.yiyouliao.com/msn/article.html?recId=c7f61c7315ab411e876c4834d959af94_s&infoId=II015XEKRWX1USJ | 易有料 - 智能信息流一站式服务平台
https://www.baidu.com/?tn=44004473_35_oem_dg | 百度一下，你就知道

https://cn.bing.com/search?q=tnesorflow+%E9%9D%A2%E8%AF%95%E9%A2%98&cvid=ab9c762f88914cbc8e9c182a29905d60&aqs=edge..69i57.8431j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | tnesorflow 面试题 - 国内版 Bing
http://www.srcmini.com/33181.html | 最新推荐！TensorFlow面试题及其答案合集 - 面试题 - srcmini
https://cn.bing.com/search?q=%E5%A6%87%E7%A7%91%E5%BE%AE%E7%94%9F%E6%80%81++%E6%BB%B4%E8%99%AB+%E8%8D%A7%E5%85%89&qs=n&form=QBRE&sp=-1&pq=%E5%A6%87%E7%A7%91%E5%BE%AE%E7%94%9F%E6%80%81+%E6%BB%B4%E8%99%AB+%E8%8D%A7%E5%85%89&sc=0-11&sk=&cvid=26B7618483464119BADB9881DD5291EF | 妇科微生态滴虫荧光 - 国内版 Bing
https://www.qdsk.cn/proDetail-2604.html | 三凯医学妇科双重荧光染色液，妇科荧光染色液，双重荧光染色液
https://www.sohu.com/a/359656949_100102218 | 妇科检查报告单看不懂？看这里就够了！_阴道
http://www.obgy.cn/diagnose-fuke/5818179baa3566f411ef1e5a | 最新指南：阴道微生态评价的临床应用专家共识-中国妇产科网
https://zhuanlan.zhihu.com/p/409947864 | 知识篇——阴道微生态报告解读 - 知乎
https://news.medlive.cn/obgyn/info-progress/show-138577_206.html | 大数据在妇产科微生态领域中的应用现状及展望_大数据在妇产科微生态领域中的应用现状及展望_医脉通
http://www.tmicrobe.com/exchange/id/11 | Nature Medicine|不同种族妊娠期女性阴道微生物组的动态变化（IF:32.621） - 学术交流 - 苏州图灵微生物科技有限公司
https://zhuanlan.zhihu.com/p/115863445 | 阴道微生态报告怎么看？解密阴道微生态，让你远离妇科炎症 - 知乎
https://www.baidu.com/s?wd=%E5%85%8D%E7%96%AB%E8%8D%A7%E5%85%89%E6%9F%93%E8%89%B2%E6%8A%80%E6%9C%AF&rsv_spt=1&rsv_iqid=0xd236d4a500052e03&issp=1&f=8&rsv_bp=1&rsv_idx=2&ie=utf-8&tn=88093251_85_hao_pg&rsv_enter=1&rsv_dl=ib&rsv_sug3=3&rsv_sug1=1&rsv_sug7=001&rsv_n=2 | 免疫荧光染色技术_百度搜索
https://cn.bing.com/search?q=%E6%BB%B4%E8%99%AB+%E5%85%8D%E7%96%AB%E8%8D%A7%E5%85%89&qs=n&form=QBRE&sp=-1&pq=%E6%BB%B4%E8%99%AB+mian%27yi%E8%8D%A7%E5%85%89&sc=0-12&sk=&cvid=DD0459F6775F44CAB05808AB42818324 | 滴虫免疫荧光 - 国内版 Bing
https://max.book118.com/html/2017/0320/96187324.shtm#:~:text=%E5%85%8D%E7%96%AB%E8%8D%A7%E5%85%89%E6%B3%95%E6%A3%80%E6%B5%8B%E9%98%B4%E9%81%93%E6%AF%9B%E6%BB%B4%E8%99%AB%E4%B9%9F%E5%B7%B2%E8%A2%AB%E5%B9%BF%E6%B3%9B%E7%A0%94%E7%A9%B6%EF%BC%8C%E8%AF%A5%E6%B3%95%E6%95%8F%E6%84%9F%E6%80%A7%E7%BA%A6%E4%B8%BA83%25%EF%BD%9E86%25,%EF%BC%8C%E5%8D%95%E5%85%8B%E9%9A%86%E6%8A%97%E4%BD%93%E7%89%B9%E5%BC%82%E6%80%A7%E9%AB%98%EF%BC%8C%E6%9C%89%E6%9C%9B%E6%9B%BF%E4%BB%A3%E5%9F%B9%E5%85%BB%E6%B3%95%E6%88%90%E4%B8%BA%E4%B8%B4%E5%BA%8A%E5%BF%AB%E9%80%9F%E8%AF%8A%E6%96%AD%E7%9A%84%E7%BB%8F%E6%B5%8E%E3%80%81%E5%AE%9E%E7%94%A8%E6%96%B9%E6%B3%95%E3%80%82 | 免疫荧光技术在生殖感染检验领域的应用价值.doc-原创力文档
https://cn.bing.com/search?q=%E6%BB%B4%E8%99%AB&qs=n&form=QBRE&sp=-1&pq=%E6%BB%B4%E8%99%AB&sc=8-2&sk=&cvid=73390B2732724740995C56B50602067C | 滴虫 - 国内版 Bing
https://baike.baidu.com/item/%E6%BB%B4%E8%99%AB/6304230 | 滴虫_百度百科
https://xuewen.cnki.net/R2006120340000008.html | 毛滴虫与滴虫病_知网百科
http://www.cogonline.com/info/132585179439468287 | 细菌性阴道病诊治指南（2021修订版）
https://www.sohu.com/a/448157420_120054526 | 细菌性阴道病诊治指南（2021修订版）_诊断
https://cn.bing.com/search?q=VFocal&cvid=8a0d2cb5c7d0454fb8845699d04c487c&aqs=edge..69i57.3998j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | VFocal - 国内版 Bing
https://cn.bing.com/search?q=sci-hub&cvid=f23da524cf0c4213ac81c7d485addeed&aqs=edge.1.69i57j0l8.2444j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | sci-hub - 国内版 Bing
https://www.scihub.net.cn/ | Sci-Hub:SciHub科研学术网址导航
https://www.scihub.net.cn/sci-hub/ | 文献检索-SciHub
https://sci-hub.se/10.1038/s41591-019-0465-8 | Sci-Hub | Racioethnic diversity in the dynamics of the vaginal microbiome during pregnancy. Nature Medicine | 10.1038/s41591-019-0465-8

https://cn.bing.com/search?q=notion+page%E7%9B%B8%E4%BA%92&qs=n&form=QBRE&sp=-1&pq=notion+page%E7%9B%B8%E4%BA%92&sc=0-13&sk=&cvid=10BF0DA92BCB4957AE402F14877FAB22 | notion page相互 - 国内版 Bing
https://cn.bing.com/search?q=Oral+Health+%E7%BF%BB%E8%AF%91&qs=n&form=QBRE&sp=-1&pq=oral+health+%E7%BF%BB%E8%AF%91&sc=0-14&sk=&cvid=A10F0D45144D4727994C9E4ACD3D03CA | Oral Health 翻译 - 国内版 Bing
https://funletu.com/newsflashes/12486.html | Notion网页版中文版 – Funletu
https://www.notion.so/Copy-of-ff27c383193d4023b9099e2ae9c70d88?showMoveTo=true | Copy of 康奈尔笔记
https://www.zhihu.com/question/377868821 | notion什么时候可中文啊…? - 知乎
https://linmi.cc/notiongroup | 加入 Notion 中文社区社群 - Linmi
https://post.smzdm.com/p/amxzrwdk/ | Notion中文客户端汉化版教程_软件应用_什么值得买
https://www.notion.so/Copy-of-ff27c383193d4023b9099e2ae9c70d88 | Copy of 康奈尔笔记
https://www.notionwizard.com/download-notion/#google_vignette | Download Notion: Mac and Windows
https://zhuanlan.zhihu.com/p/440773804 | Notion中文客户端汉化版教程 - 知乎
https://pan.baidu.com/s/185CK_tlr0_IEwxe35xgNvw | Notion客户端汉化文件.zip_免费高速下载|百度网盘-分享无限制
https://zhuanlan.zhihu.com/p/397634631 | 【notion中文汉化方法】真实有效，成功率100% - 知乎
https://greasyfork.org/zh-CN/scripts/396279-notion-translate | Notion-Translate
https://cn.bing.com/search?q=Attention_ocr&cvid=6bcfa7f45a274518acebc2663714b037&aqs=edge..69i57j0l8.1350j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | Attention_ocr - 国内版 Bing
https://www.ptorch.com/news/228.html | Pytorch AttentionOCR 中文端到端的文字识别(Attention_ocr.pytorch) - pytorch中文网
https://github.com/chenjun2hao/Attention_ocr.pytorch | chenjun2hao/Attention_ocr.pytorch: This repository implements the the encoder and decoder model with attention model for OCR
https://zhuanlan.zhihu.com/p/138589087 | 自然场景文本检测识别 - Attention OCR - 知乎
https://github.com/search?q=Attention_ocr.pytorch | Search · Attention_ocr.pytorch

https://github.com/opconty/pytorch_ctpn/blob/89ea784d7776a08962c6d5bc5591730e15156d07/ctpn_utils.py#L289 | pytorch_ctpn/ctpn_utils.py at 89ea784d7776a08962c6d5bc5591730e15156d07 · opconty/pytorch_ctpn
https://cn.bing.com/search?q=ctpn+github&cvid=d4d12a077f774cc1847405ba4a0355b4&aqs=edge..69i57j69i60l3.5341j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | ctpn github - 国内版 Bing
https://github.com/tianzhi0549/CTPN/blob/master/src/text_proposal_graph_builder.py | CTPN/text_proposal_graph_builder.py at master · tianzhi0549/CTPN
https://cn.bing.com/search?q=get_successions&cvid=81092aa7e45f43d398151b33b4414ead&aqs=edge..69i57.824j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | get_successions - 国内版 Bing
https://www.cnblogs.com/skyfsm/p/10054386.html | 【OCR技术系列之六】文本检测CTPN的代码实现 - Madcola - 博客园
https://cn.bing.com/search?q=CRNN+GitHub&cvid=cc375d66a3ab4386bbef37623806936a&aqs=edge..69i57j69i60l3.4877j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | CRNN GitHub - 国内版 Bing
https://github.com/xmy0916/pytorch_crnn | xmy0916/pytorch_crnn: 基于pytorch写的CRNN文字识别简化写法帮助入门
https://blog.csdn.net/qq_37668436/article/details/113642808 | (26条消息) 从零写CRNN文字识别 —— （1）准备工作_只会git clone的程序员的博客-CSDN博客
https://cn.bing.com/search?q=CTCLoss+%E5%8E%9F%E7%90%86&qs=n&form=QBRE&sp=-1&pq=ctcloss+%E5%8E%9F%E7%90%86&sc=0-10&sk=&cvid=1D9FA7FEB1C241F18822E9386A96EA8A | CTCLoss 原理 - 国内版 Bing
https://blog.csdn.net/yifen4234/article/details/80334516 | (26条消息) CTC loss 理解_体重51.00-CSDN博客_ctc loss
https://blog.csdn.net/jackytintin/article/details/79425866 | (26条消息) 【Learning Notes】CTC 原理及实现_丁丁的博客-CSDN博客_ctc 实现
https://distill.pub/2017/ctc/ | Sequence Modeling with CTC
https://blog.csdn.net/justsolow/article/details/105251789 | (26条消息) CTCloss从理论到训练_justsolow的博客-CSDN博客_ctc loss
https://runebook.dev/zh-CN/docs/pytorch/generated/torch.nn.ctcloss | PyTorch - CTCLoss - 连接主义时空分类的损失。计算连续（非分段）时间序列与目标序列之间的损失。CTCLoss计算输入与目标可能对齐的概率之和，从而产生相对于每个输入节点可微分的损耗 - 中文
https://zhuanlan.zhihu.com/p/23293860 | CTC实现——compute ctc loss（1） - 知乎
https://blog.csdn.net/fanhenghui/article/details/102687201 | (26条消息) Pytorch使用CRNN CTCLoss实现OCR系统_fanhenghui的专栏-CSDN博客_pytorch实现ocr
https://zhuanlan.zhihu.com/p/33464788#::text=%E5%9C%A8pytorch,%E8%BF%98%E4%BF%9D%E5%AD%98%E4%BA%86%E5%85%B6%E6%A2%AF%E5%BA%A6%E3%80%82 | 基于CTC的语音识别基础与实现 - 知乎
https://zhuanlan.zhihu.com/p/67415439 | 如何优雅的使用pytorch内置torch.nn.CTCLoss的方法 - 知乎
http://ailab.aiwin.org.cn/competitions/65#phases | AIWIN - 比赛
https://cn.bing.com/search?q=%E6%9F%A5%E7%9C%8B%E7%94%B5%E8%B4%B9%E8%AF%A6%E7%BB%86&form=ANNTH1&refig=ecbf9a923bf247cbad95e4d6fb06457e | 查看电费详细 - 国内版 Bing
https://cn.bing.com/search?q=%E4%B8%80%E6%88%B7+%E7%87%83%E6%B0%94+%E6%AD%A3%E5%B8%B8%E7%94%A8%E9%87%8F&qs=n&form=QBRE&sp=-1&pq=%E4%B8%80%E6%88%B7+%E7%87%83%E6%B0%94+%E6%AD%A3%E5%B8%B8%E7%94%A8%E9%87%8F&sc=0-10&sk=&cvid=B9DFDD969C914C82BF312640ADC45485 | 一户燃气正常用量 - 国内版 Bing
https://www.baidu.com/?tn=88093251_22_hao_pg | 百度一下，你就知道

https://zhuanlan.zhihu.com/p/81876053 | OCR综述论文笔记 - 知乎
http://xxx.itp.ac.cn/pdf/1811.04256.pdf | 1811.04256.pdf
https://cn.bing.com/search?q=ocr%e8%ae%ba%e6%96%87+%e7%9f%a5%e4%b9%8e&qs=n&sp=-1&pq=ocrlun%27wen+%e7%9f%a5%e4%b9%8e&sc=0-13&sk=&cvid=A7741319301441DA8D92A2526555158A&first=59&FORM=PORE | ocr论文知乎 - 国内版 Bing
https://www.cnblogs.com/thinc/p/14092201.html | PP-OCR论文翻译 - 7hinc - 博客园
https://www.cnblogs.com/hithongming/ | HongmingYou - 博客园
https://zhuanlan.zhihu.com/p/428777199 | OCR综述 (持续更新) - 知乎
https://zhuanlan.zhihu.com/p/149417999 | OCR核心技术-CRNN-CTC论文解读（3） - 知乎
https://www.zhihu.com/people/F-Zero/posts | Gavin - 知乎
https://zhuanlan.zhihu.com/p/115991987 | CVPR2020论文解读：OCR场景文本识别 - 知乎
https://zhuanlan.zhihu.com/p/265282666 | ECCV 2020 论文大盘点 - OCR 篇 - 知乎
https://www.zhihu.com/topic/19574441/hot | OCR（光学字符识别） - 知乎
https://cn.bing.com/search?q=ocr+%E7%BB%BC%E8%BF%B0+2022&qs=n&form=QBRE&sp=-1&pq=ocr+%E7%BB%BC%E8%BF%B0+2022&sc=0-11&sk=&cvid=70F86BA5089E479D923350D5004CE23B | ocr 综述 2022 - 国内版 Bing
https://blog.csdn.net/amusi1994/article/details/118098111 | (25条消息) 最新！CVPR 2021 OCR领域论文大盘点（22篇）_阿木寺的博客-CSDN博客
https://paddlepedia.readthedocs.io/en/latest/tutorials/computer_vision/OCR/OCR.html#:~:text=OCR%20%28Optical%20Character,Recognition%2C%E5%85%89%E5%AD%A6%E5%AD%97%E7%AC%A6%E8%AF%86%E5%88%AB%29%E6%98%AF%E6%8C%87%E5%AF%B9%E5%9B%BE%E5%83%8F%E8%BF%9B%E8%A1%8C%E5%88%86%E6%9E%90%E8%AF%86%E5%88%AB%E5%A4%84%E7%90%86%EF%BC%8C%E8%8E%B7%E5%8F%96%E6%96%87%E5%AD%97%E5%92%8C%E7%89%88%E9%9D%A2%E4%BF%A1%E6%81%AF%E7%9A%84%E8%BF%87%E7%A8%8B%EF%BC%8C%E6%98%AF%E5%85%B8%E5%9E%8B%E7%9A%84%E8%AE%A1%E7%AE%97%E6%9C%BA%E8%A7%86%E8%A7%89%E4%BB%BB%E5%8A%A1%EF%BC%8C%E9%80%9A%E5%B8%B8%E7%94%B1%E6%96%87%E6%9C%AC%E6%A3%80%E6%B5%8B%E5%92%8C%E6%96%87%E6%9C%AC%E8%AF%86%E5%88%AB%E4%B8%A4%E4%B8%AA%E5%AD%90%E4%BB%BB%E5%8A%A1%E6%9E%84%E6%88%90%E3%80%82%20%E6%96%87%E5%AD%97%E6%A3%80%E6%B5%8B%EF%BC%9A%E5%B0%86%E5%9B%BE%E7%89%87%E4%B8%AD%E7%9A%84%E6%96%87%E5%AD%97%E5%8C%BA%E5%9F%9F%E4%BD%8D%E7%BD%AE%E6%A3%80%E6%B5%8B%E5%87%BA%E6%9D%A5%EF%BC%88%E5%A6%82%E5%9B%BE1%20%28b%29%E6%89%80%E7%A4%BA%EF%BC%89%EF%BC%9B | OCR综述 — PaddleEdu documentation
https://github.com/search?q=PaddleOCR&type=repositories | Search · PaddleOCR
https://github.com/PaddlePaddle/PaddleOCR/blob/release/1.1/doc/doc_en/models_list_en.md | PaddleOCR/models_list_en.md at release/1.1 · PaddlePaddle/PaddleOCR
https://codalab.lisn.upsaclay.fr/competitions/695#learn_the_details | CodaLab - Competition
http://xxx.itp.ac.cn//pdf/2105.02373.pdf | 2105.02373.pdf
https://github.com/wanghaisheng/awesome-ocr | wanghaisheng/awesome-ocr: A curated list of promising OCR resources
https://codalab.lisn.upsaclay.fr/my/ | CodaLab - Competitions (My CodaLab)
https://codalab.lisn.upsaclay.fr/competitions/707#participate | CodaLab - Competition
https://codalab.lisn.upsaclay.fr/competitions/1400 | CodaLab - Competition
https://codalab.lisn.upsaclay.fr/competitions/1235 | CodaLab - Competition
https://github.com/datawhalechina/competition-baseline/tree/master/competition/WSDM2022 | competition-baseline/competition/WSDM2022 at master · datawhalechina/competition-baseline
https://xmrec.github.io/wsdmcup/ | WSDM Cup on Cross-Market Recommendation Competition | Cross-Market Recommendation (XMRec)
https://competitions.codalab.org/my/ | CodaLab - Competitions (My CodaLab)
https://cn.bing.com/search?q=BVCC+dataset+Mosnet&qs=n&form=QBRE&sp=-1&pq=bvcc+dataset+mosnet&sc=0-19&sk=&cvid=2B21E7225E4A47AD8A7EE15195823CB9 | BVCC dataset Mosnet - 国内版 Bing
https://jishuin.proginn.com/p/763bfbd6786e | NLP 民工的乐园-技术圈

https://zhuanlan.zhihu.com/p/64410344 | 一点一点理解faster-rcnn - 知乎
https://zhuanlan.zhihu.com/p/183261974 | 你一定从未看过如此通俗易懂的YOLO系列(从v1到v5)模型解读 (上) - 知乎
https://blog.csdn.net/u013066730/article/details/100577501 | (26条消息) 【FasterRCNN】边框回归(Bounding Box Regression)详解_mjiansun的专栏-CSDN博客_faster rcnn回归
https://cn.bing.com/search?q=%e4%bc%98%e5%8c%96%e5%99%a8++&qs=n&sp=-1&pq=%e4%bc%98%e5%8c%96%e5%99%a8+&sc=8-4&sk=&cvid=30381981A9E14308ADDC102CED2950B2&first=10&FORM=PORE | 优化器 - 国内版 Bing
https://zhuanlan.zhihu.com/p/58236906 | 深度学习优化器总结 - 知乎
https://www.cnblogs.com/froml77/p/14956375.html | (◍´꒳`◍) Hi, L77 - 深度学习的优化器（各类 optimizer 的原理、优缺点及数学推导）
https://zhuanlan.zhihu.com/p/261695487 | 优化器(Optimizer) - 知乎
https://zhuanlan.zhihu.com/p/208178763 | 机器学习2 – 优化器（SGD、SGDM、Adagrad、RMSProp、Adam） - 知乎
https://blog.csdn.net/qq_42109740/article/details/105401197 | (26条消息) 深度学习各类优化器详解（动量、NAG、adam、Adagrad、adadelta、RMSprop、adaMax、Nadam、AMSGrad）_恩泽君的博客-CSDN博客_优化器动量
https://www.jianshu.com/p/d99b83f4c1a6/ | 如何选择优化器 optimizer - 简书
https://cn.bing.com/search?q=ML-GCN&qs=n&form=QBRE&sp=-1&pq=ml-gcn&sc=0-6&sk=&cvid=EAA8E8EB99BA4FB3A71B192046AED163 | ML-GCN - 国内版 Bing
https://zhuanlan.zhihu.com/p/62212910 | 旷视研究院提出ML-GCN：基于图卷积网络的多标签图像识别模型 - 知乎
https://cloud.tencent.com/developer/article/1841815 | 终于有人总结了图神经网络！ - 云+社区 - 腾讯云
https://blog.csdn.net/xys430381_1/article/details/89640699 | (26条消息) 细粒度图像分类（FGVC）—综述_xys430381_1的专栏-CSDN博客
https://zhuanlan.zhihu.com/p/75199948 | 【图像分类】细粒度图像分类 - 知乎
https://cn.bing.com/search?q=Zetane+Engine&cvid=134198b0b33c4af6bc75ea8f3ced08e0&aqs=edge..69i57.815j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | Zetane Engine - 国内版 Bing
https://www.thepaper.cn/newsDetail_forward_16390115 | 上传ML模型，一键可视化，终于能看懂神经网络到底在干啥了_湃客_澎湃新闻-The Paper
https://blog.csdn.net/weixin_47196664/article/details/122646580 | (26条消息) 神器来啦！一键可视化，终于能看懂神经网络到底在干啥了_深度学习技术前沿-CSDN博客
https://cn.bing.com/search?q=yolov3+bbox+sigmoid&qs=n&form=QBRE&sp=-1&pq=yolov3+bbox+sigmoid&sc=0-19&sk=&cvid=1EE20C1441484F98A47C93FEE1CFABF0 | yolov3 bbox sigmoid - 国内版 Bing
https://blog.csdn.net/qq_38262266/article/details/104231371 | (26条消息) YOLO-V3 Bbox预测解读_沐雨金鳞-CSDN博客_bbox yolo
https://blog.csdn.net/qq_34199326/article/details/84109828 | (26条消息) 史上最详细的Yolov3边框预测分析_逍遥王的博客-CSDN博客_yolov3锚框
https://zhuanlan.zhihu.com/p/50595699 | 揭秘YOLOv3鲜为人知的关键细节 - 知乎
https://zhuanlan.zhihu.com/p/49995236 | 超详细的Yolo检测框预测分析 - 知乎
https://blog.csdn.net/z0n1l2/article/details/84076775 | (26条消息) yolov3 中box坐标的处理_大泽之国-CSDN博客_yolov3后处理
https://cloud.tencent.com/developer/article/1820939 | 基于深度学习的目标检测算法面试必备（RCNN~YOLOv5） - 云+社区 - 腾讯云

https://cn.bing.com/search?q=pixelshuffle%e5%8e%9f%e7%90%86&FORM=R5FD | pixelshuffle原理 - 国内版 Bing
https://www.jianshu.com/p/71d6a9374899 | 上采样和PixelShuffle - 简书
https://www.jianshu.com/search?q=pixel&page=1&type=note | 搜索 - 简书
https://www.jianshu.com/search?q=pixel%20shuffle&page=1&type=note | 搜索 - 简书
https://www.jianshu.com/p/7cc6821a2714 | 显著性目标检测之Learning to Promote Saliency Detectors - 简书
https://www.baidu.com/s?wd=pixelshuffle%20%E8%AF%A6%E8%A7%A3&rsv_spt=1&rsv_iqid=0xb5a02f8e0009ef23&issp=1&f=8&rsv_bp=1&rsv_idx=2&ie=utf-8&rqlang=cn&tn=67074732_8_dg&rsv_enter=1&rsv_dl=tb&oq=pixelshuffle%2520pytorch&rsv_btype=t&rsv_t=ccd3J7YfmS8s28oojsde%2BVzNyrDwN3cHXac%2FqTq%2BYcEv4x7A1vkCsZh6kLw0bEVqXxhMOw&rsv_pq=8d77ee6c00121b37&inputT=5542&rsv_sug3=72&rsv_sug1=41&rsv_sug7=100&rsv_sug2=0&rsv_sug4=6359 | pixelshuffle 详解_百度搜索
https://blog.csdn.net/g11d111/article/details/82855946 | (24条消息) PyTorch学习笔记(10)——上采样和PixelShuffle_g11d111的博客-CSDN博客_nn.pixelshuffle
https://www.cnblogs.com/LXP-Never/p/13181168.html | 神经网络中的降维和升维方法 (tensorflow & pytorch) - 凌逆战 - 博客园
https://www.cnblogs.com/jiangkejie/p/12919155.html | pytorch中的上采样（上采样，转置卷积，上池化，PixelShuffle） - 那抹阳光1994 - 博客园
https://blog.csdn.net/bby1987/article/details/113446309 | (24条消息) python numpy实现PixelShuffle及其逆变换（depth_to_space，space_to_depth）_吹吹自然风-CSDN博客
https://blog.csdn.net/qq249356520/article/details/95318645 | (24条消息) 超分辨率网络ESPCN中的pixel shuffle–几种代码思路（基于TF, Pytorch）_qq249356520的博客-CSDN博客_pixel shuffle
https://blog.csdn.net/g11d111/article/details/82855946 | (24条消息) PyTorch学习笔记(10)——上采样和PixelShuffle_g11d111的博客-CSDN博客_nn.pixelshuffle
https://blog.csdn.net/u014636245/article/details/98071626 | (24条消息) PixelShuffle特征图的上采样的方法_TomRen-CSDN博客_pixel shuffle
https://blog.csdn.net/qq_14845119/article/details/107557449 | (24条消息) 上采样方法大PK（Upsample，Interpolate，resize，Transposed convolution，deconv，Unpool，Pixelshuffle）_年轻即出发，-CSDN博客_upsample和interpolate
https://blog.csdn.net/weixin_43206928/article/details/88964128 | (24条消息) 上采样方法_ldd_unique的博客-CSDN博客_上采样
https://www.zhihu.com/question/342143752 | Upsample、ConvTranspose2d、conv后PixelShuffle用法上有什么区别？ - 知乎
https://cn.bing.com/search?q=%E8%BD%AC%E7%BD%AE%E5%8D%B7%E7%A7%AF&cvid=b98b4c8067ea4cafa10855fb61691670&aqs=edge.2.69i57j0l8.4575j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | 转置卷积 - 国内版 Bing
https://zhuanlan.zhihu.com/p/113525131 | 转置卷积（反卷积） - 知乎
https://jishuin.proginn.com/p/763bfbd2aecd | 一文搞懂转置卷积（反卷积）-技术圈

https://cn.bing.com/search?q=opencv+%E7%AA%81%E5%87%BA&qs=n&form=QBRE&sp=-1&pq=opencv+%E7%AA%81%E5%87%BA&sc=1-9&sk=&cvid=D4181EB7561244C2BF6A5B87052D3545 | opencv 突出 - 国内版 Bing
https://blog.csdn.net/augusdi/article/details/9021467 | (20条消息) OpenCV轮廓检测_Augusdi的专栏-CSDN博客_opencv轮廓检测
https://blog.csdn.net/hackerjoy/article/details/8296097 | (20条消息) opencv for python 之突出点检测_Flying cat-CSDN博客
https://blog.csdn.net/yizhang_ml/article/details/86994193 | (20条消息) 角点（corner point）、关键点(key point)、特征点(feature point)概念辨析_yizhang_ml的博客-CSDN博客_角点
https://blog.csdn.net/u010712012/article/details/84947831 | (20条消息) 局部特征——角点检测（Harris，Shi-Tomasi）IT荻的博客-CSDN博客_角点
https://cn.bing.com/search?q=opencv+%e5%bd%a2%e7%8a%b6%e6%a3%80%e6%b5%8b&qs=SC&pq=opencv+xing%27z&sk=SC2&sc=8-13&cvid=659533CA9DD1434BB0418DC35264234B&FORM=QBRE&sp=3 | opencv 形状检测 - 国内版 Bing
https://blog.csdn.net/WZZ18191171661/article/details/91045219 | (20条消息) Python+Opencv实现多种形状的检测_WZZ18191171661的博客-CSDN博客_python 检测
https://blog.csdn.net/Yong_Qi2015/article/details/108675561 | (20条消息) 使用OpenCV进行模糊检测（拉普拉斯算子）_3D视觉工坊-CSDN博客
https://cn.bing.com/search?q=imutils&form=ANNTH1&refig=2a6981e0b327465590092e0fc8c1b117 | imutils - 国内版 Bing
https://cn.bing.com/search?q=python+cv2+%E6%A3%80%E6%B5%8B%E5%AE%9E%E5%BF%83%E6%A4%AD%E5%9C%86&qs=n&form=QBRE&sp=-1&pq=python+cv2+%E6%A3%80%E6%B5%8B%E5%AE%9E%E5%BF%83%E6%A4%AD%E5%9C%86&sc=0-17&sk=&cvid=CC37422B61954A42AFE0BF7D487AB37A | python cv2 检测实心椭圆 - 国内版 Bing
https://cn.bing.com/search?q=%E5%9C%9F%E9%BB%84%E8%89%B2+hsv&cvid=011cadf8958347079d5c864483a3f3f3&aqs=edge..69i57j0.17067j0j4&FORM=ANAB01&PC=HCTS | 土黄色 hsv - 国内版 Bing
https://www.cnblogs.com/long5683/p/9757135.html | OpenCV——RGB和HSV颜色空间 - 一抹烟霞 - 博客园
https://blog.csdn.net/xddwz/article/details/110232977 | (22条消息) python-opencv之圆检测_xddwz的博客-CSDN博客
https://zhuanlan.zhihu.com/p/133531570 | python opencv :使用Hough变换检测圆形，数钱。 - 知乎
https://blog.csdn.net/qq_41662180/article/details/108147118 | (22条消息) Python+opencv — cv2.HoughCircles（）圆形检测，并顺时针给圆编序_不知名的小白66的博客-CSDN博客_cv2 圆形检测
https://blog.csdn.net/qq_42544886/article/details/106701552 | (20条消息) Python3+Opencv 学习笔记（三）图像轮廓处理+描绘矩形框+ 矩形框加文字_辞的博客-CSDN博客
https://blog.csdn.net/qq_33897832/article/details/88931748 | (20条消息) python+opencv图像处理(提取矩形方框提取颜色滤波轮廓检测直线检测)若风的博客-CSDN博客_python矩形检测
https://blog.csdn.net/lionkawoo/article/details/83928116 | (20条消息) 【Python+OpenCV】基于Harris角点的边缘提取以及矩形四角点检测_lionkawoo的博客-CSDN博客_opencv检测矩形4个点坐标
https://blog.csdn.net/wsp_1138886114/article/details/82945328 | (20条消息) OpenCV—Python 轮廓检测绘出矩形框（findContours\ boundingRect\rectangle）_wsp_1138886114的博客-CSDN博客_python 矩形检测
https://blog.csdn.net/HuangZhang_123/article/details/80511270 | (20条消息) Python 使用Opencv实现边缘检测以及轮廓检测_Xy-Huang的博客-CSDN博客_python 轮廓检测
https://blog.csdn.net/m0_51402531/article/details/121066693 | (20条消息) OpenCV——Canny边缘检测（cv2.Canny()）_m0_51402531的博客-CSDN博客_cv2.canny
https://blog.csdn.net/Easen_Yu/article/details/89365497 | (20条消息) cv2.findContours() 轮廓检测_Easen_Yu的博客-CSDN博客
https://blog.csdn.net/gaoranfighting/article/details/34877549 | (20条消息) opencv.cv2.findContours()函数_gaoranfighting的专栏-CSDN博客_cv2.findcontours
https://cn.bing.com/search?q=%E6%BE%B3%E5%A4%A7%E5%88%A9%E4%BA%9A%E6%B3%95%E5%AE%9A%E5%81%87%E6%97%A5&cvid=775bc42c6d794f799a2393030564f52b&aqs=edge.1.69i57j0l7.10679j0j4&FORM=ANAB01&PC=HCTS | 澳大利亚法定假日 - 国内版 Bing
https://cn.bing.com/search?q=%E5%BD%93%E4%BD%BF%E7%94%A8sigmoid%E4%BD%9C%E4%B8%BA%E6%BF%80%E6%B4%BB%E5%87%BD%E6%95%B0%E7%9A%84%E6%97%B6%E5%80%99%EF%BC%8C%E5%B8%B8%E7%94%A8%E4%BA%A4%E5%8F%89%E7%86%B5%E6%8D%9F%E5%A4%B1%E5%87%BD%E6%95%B0%E8%80%8C%E4%B8%8D%E7%94%A8%E5%9D%87%E6%96%B9%E8%AF%AF%E5%B7%AE%E6%8D%9F%E5%A4%B1%E5%87%BD%E6%95%B0&cvid=8dc4d032fafb49d1834698c0c2015e84&aqs=edge..69i57.815j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | 当使用sigmoid作为激活函数的时候，常用交叉熵损失函数而不用均方误差损失函数 - 国内版 Bing
https://blog.csdn.net/Robin___/article/details/105225581 | (22条消息) 当使用Sigmoid或者Softmax作为激活函数时，损失函数推荐使用交叉熵损失函数而不是均方差损失函数_Robin_____的博客-CSDN博客_softmax 均方差
https://blog.csdn.net/lcalqf/article/details/71170171 | (20条消息) opencv cv2.THRESH_OTSU 二值化_lcalqf的专栏-CSDN博客_thresh_otsu
https://blog.csdn.net/qq_39507748/article/details/104506306 | (20条消息) opencv学习笔记七：使用cv2.threshold实现全局和OTSU阈值二值化_耐心的小黑的博客-CSDN博客_cv2 otsu
https://blog.csdn.net/qq_36584673/article/details/121398502 | (23条消息) 【Python+OpenCV】二值化函数cv2.threshold()的用法详解_十八岁小白想成大牛-CSDN博客_cv.threshold python

edge://newtab/ | 新建标签页
https://www.baidu.com/?tn=44004473_35_oem_dg | 百度一下，你就知道
https://cn.bing.com/search?q=kaggle&qs=n&form=QBRE&sp=-1&pq=kaggle&sc=8-6&sk=&cvid=C223CAFF8DB14071AB0512ED14FCF486 | kaggle - 国内版 Bing
https://www.kaggle.com/search?q=super+ctr | Search | Kaggle
https://blog.csdn.net/springtostring/article/details/108157070 | (20条消息) DeepFM Pytorch实现(Criteo数据集验证)_springtostring的博客-CSDN博客_deepfm pytorch实现
https://github.com/datawhalechina/fun-rec/blob/master/codes/base_models/DeepFM.py | fun-rec/DeepFM.py at master · datawhalechina/fun-rec
https://blog.csdn.net/springtostring/article/details/108157070 | (20条消息) DeepFM Pytorch实现(Criteo数据集验证)_springtostring的博客-CSDN博客_deepfm pytorch实现
https://cn.bing.com/search?q=%E5%A6%82%E4%BD%95%E9%98%85%E8%AF%BB%E8%AE%BA%E6%96%87&cvid=925fd3bbc458481daa49c11f0ea3f0ed&aqs=edge.0.0l4.9879j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | 如何阅读论文 - 国内版 Bing
https://www.1024sky.cn/blog/article/64332 | 怎样读论文？
https://www.jianshu.com/p/4aff99acfce8 | 如何高效读论文？ - 简书
https://zhuanlan.zhihu.com/p/393554663 | 如何阅读一篇论文 - 知乎
https://zhuanlan.zhihu.com/p/354260804 | 【论文阅读】如何高效阅读学术论文 - 知乎

https://cn.bing.com/search?q=%E7%BD%91%E7%BB%9C%E8%B0%83%E5%8F%82+%E8%AE%BA%E6%96%87%E5%88%86%E4%BA%AB&qs=n&form=QBRE&sp=-1&pq=%E7%BD%91%E7%BB%9C%E8%B0%83%E5%8F%82+%E8%AE%BA%E6%96%87%E5%88%86%E4%BA%AB&sc=0-9&sk=&cvid=C27A18FC134344C68CBEA28AA279EF30 | 网络调参论文分享 - 国内版 Bing
https://blog.csdn.net/fengdu78/article/details/119847498 | (20条消息) 【深度学习】知乎高赞：深度学习如何调参？_fengdu78的博客-CSDN博客
https://zhuanlan.zhihu.com/p/164615837 | 【Deep Learning】深度学习调参技巧的总结 - 知乎
https://blog.csdn.net/weixin_39653948/article/details/105962427 | (20条消息) 调参：深度学习模型24种优化策略_Constant dripping wears the stone-CSDN博客
https://cn.bing.com/search?q=python++defaultdict+setdefault&qs=n&form=QBRE&sp=-1&pq=python+defaultdict+setdefault&sc=0-29&sk=&cvid=A9FCA1C0FF88430AB9E555D9DA47C096 | python defaultdict setdefault - 国内版 Bing
https://foofish.net/setdefault-defaultdict.html | 提升逼格的两个函数：setdefault 与 defaultdict - FooFish
https://blog.csdn.net/weixin_41636030/article/details/103273243 | (20条消息) defaultdict(function_factory), setdefault和dict.items()用法_tomwang0322的博客-CSDN博客
https://blog.csdn.net/WSH_ONLY/article/details/83243112 | (20条消息) python技巧（setdefault 与 defaultdict）_小花猪的博客-CSDN博客
https://www.runoob.com/python/att-dictionary-setdefault.html | Python 字典(Dictionary) setdefault()方法 | 菜鸟教程
https://www.html.cn/script/python/154246.html | Python之字典添加元素的几种方法 - html中文网
https://www.runoob.com/python/python-dictionary.html | Python 字典(Dictionary) | 菜鸟教程
edge://newtab/ | 新建标签页
https://www.baidu.com/?tn=44004473_27_oem_dg | 百度一下，你就知道
https://www.baidu.com/s?cl=3&tn=baidutop10&fr=top1000&rsv_idx=2&rsv_dl=fyb_n_homepage&sa=fyb_n_homepage&hisfilter=1&wd=%E6%B1%9F%E6%AD%8C%E5%A6%88%E5%A6%88%E4%B8%BA%E8%B5%B7%E8%AF%89%E5%88%98%E9%91%AB%E8%8A%B1%E8%B4%B9120%E4%B8%87 | 江歌妈妈为起诉刘鑫花费120万_百度搜索
https://www.baidu.com/s?tn=baidutop10&rsv_idx=2&wd=%E6%AF%94%E5%88%A9%E6%97%B6%E5%BB%BA%E8%AE%AE%E8%B5%B4%E5%8D%8E%E5%8B%BF%E5%B8%A6%E6%89%8B%E6%9C%BA%3F%E4%B8%AD%E6%96%B9%E5%9B%9E%E5%BA%94&ie=utf-8&rsv_pq=c950366a000ae911&oq=%E6%B1%9F%E6%AD%8C%E5%A6%88%E5%A6%88%E4%B8%BA%E8%B5%B7%E8%AF%89%E5%88%98%E9%91%AB%E8%8A%B1%E8%B4%B9120%E4%B8%87&rsv_t=a91dDKxjnDQBXyF8%2FEYdDgDaHdSrF2zXktbjeCjNWwDJMlY7lQSVZnH6a50uSJDFMQ&rqid=c950366a000ae911&rsf=e8aaf5a5087a5d9f97114b54040228f9_1_15_10&rsv_dl=0_right_fyb_pchot_20811&sa=0_right_fyb_pchot_20811 | 比利时建议赴华勿带手机?中方回应_百度搜索
https://www.baidu.com/s?ie=utf-8&f=8&rsv_bp=1&rsv_idx=2&tn=baidutop10&wd=hsv%20%E7%B2%89%E7%BA%A2%E8%89%B2&oq=%25E5%25A5%25B3%25E5%25AD%2590%25E5%25BC%2580%25E8%25BD%25A6%2526lt%253B0%25E5%2585%25AC%25E9%2587%258C%25E7%25BB%2599%25E8%2580%2581%25E5%2585%25AC%25E5%2589%258D%25E5%25A6%25BB%25E5%25A5%25B3%25E5%2584%25BF%25E9%2580%2581%25E9%25A5%25AD&rsv_pq=f6d9b3240004609b&rsv_t=ea0e5J%2FZ6Gr5OGQGuofDiC9Vd5dAT0fZQHPl9nJG7S17sAMcZCY%2BSCMCv8Zkd6NBiQ&rqlang=cn&rsv_enter=1&rsv_dl=tb&rsv_sug3=13&rsv_sug1=17&rsv_sug7=100&rsv_sug2=0&rsv_btype=t&inputT=5282&rsv_sug4=6609 | hsv 粉红色_百度搜索
https://www.cnblogs.com/wangyblzu/p/5710715.html | OpenCV中HSV颜色模型及颜色分量范围 - wangyblzu - 博客园
https://blog.csdn.net/WarGames_dc/article/details/89305438 | (20条消息) HSV颜色空间中颜色（红、黄、绿、青、蓝、紫、粉红、砖红、品红）对应的灰度范围_WarGames_dc的博客-CSDN博客_粉色hsv
https://cn.bing.com/search?q=python+csv&cvid=885032f71a554d3fbb26777f251c5a84&aqs=edge.0.0l9.5126j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | python csv - 国内版 Bing
http://c.biancheng.net/python_spider/csv-module.html | Python csv模块（读写文件）
http://localhost:8888/lab/tree/%E7%AC%AC1%E7%AB%A0%20Pandas%E5%9F%BA%E7%A1%80.ipynb | 第1章 Pandas基础.ip (2) - JupyterLab
edge://newtab/ | 新建标签页
https://gitee.com/Liam1104/projects | 仓库 - hu1104 (Liam1104) - Gitee.com
https://gitee.com/Liam1104/record_myself/blob/master/pool_plot.py | pool_plot.py · hu1104/record_myself - 码云 - 开源中国
https://zentao.medbot.top/pro/task-view-3090.html | TASK#3090 模型优化 / 抗酸杆菌分析AIV1 - 禅道
https://cn.bing.com/search?q=%E5%A2%9E%E9%87%8F%E5%AD%A6%E4%B9%A0&cvid=7aa01611a0cd4063bfa9fcb5145be23a&aqs=edge.0.0l9.3063j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | 增量学习 - 国内版 Bing
https://zhuanlan.zhihu.com/p/353273834 | 增量学习(Incremental Learning)小综述 - 知乎
https://cn.bing.com/search?q=%E7%BF%BB%E8%AF%91&cvid=d5c29f7b0c8b4ad1901eb7ada3467417&aqs=edge.0.0l9.1381j0j4&FORM=ANAB01&PC=HCTS | 翻译 - 国内版 Bing
https://cn.bing.com/search?q=Real-time+automatic+helmet+detection+of+motorcyclists+in+urban+traffic+using+improved+YOLOv5+detector&cvid=cc4b825d60794bf38ae8aadb47a10efa&aqs=edge..69i57.1006j0j4&FORM=ANAB01&PC=HCTS | Real-time automatic helmet detection of motorcyclists in urban traffic using improved YOLOv5 detector - 国内版 Bing
https://ietresearch.onlinelibrary.wiley.com/doi/epdf/10.1049/ipr2.12295 | Real‐time automatic helmet detection of motorcyclists in urban traffic using improved YOLOv5 detector
https://cn.bing.com/search?q=Triplet+attention&qs=n&form=QBRE&sp=-1&pq=triplet+attention&sc=0-17&sk=&cvid=9CBA2C74709E4716947A6417BCBD5132 | Triplet attention - 国内版 Bing
https://arxiv.org/pdf/2010.03045.pdf | 2010.03045.pdf
https://cn.bing.com/search?q=Z-Pool&cvid=fb3f2bad6512410b810d98d5c75b7ea1&aqs=edge..69i57.717j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | Z-Pool - 国内版 Bing
https://blog.csdn.net/yx868yx/article/details/107284707 | (21条消息) CV中的attention机制SE、CBAM、Triplet、DCANet_yx868yx的博客-CSDN博客
https://cn.bing.com/search?q=%E7%BF%BB%E8%AF%91&cvid=d5c29f7b0c8b4ad1901eb7ada3467417&aqs=edge.0.0l9.1381j0j4&FORM=ANAB01&PC=HCTS | 翻译 - 国内版 Bing
https://translate.google.cn/?sl=auto&tl=zh-CN&text=%20setting%20an%20extra%20small%20feature%20map%20inclusively&op=translate | Google 翻译
https://zhuanlan.zhihu.com/p/57208761 | [机器学习]偏差和方差的理解 - 知乎
https://blog.csdn.net/weixin_42096202/article/details/108970057 | (23条消息) Triplet Attention -接近无参注意力| Rotate to Attend: Convolutional Triplet Attention Module_chenzy_hust的博客-CSDN博客_triplet注意力

https://cn.bing.com/search?q=opencv+%E7%AA%81%E5%87%BA&qs=n&form=QBRE&sp=-1&pq=opencv+%E7%AA%81%E5%87%BA&sc=1-9&sk=&cvid=D4181EB7561244C2BF6A5B87052D3545 | opencv 突出 - 国内版 Bing
https://blog.csdn.net/augusdi/article/details/9021467 | (20条消息) OpenCV轮廓检测_Augusdi的专栏-CSDN博客_opencv轮廓检测
https://blog.csdn.net/hackerjoy/article/details/8296097 | (20条消息) opencv for python 之突出点检测_Flying cat-CSDN博客
https://blog.csdn.net/yizhang_ml/article/details/86994193 | (20条消息) 角点（corner point）、关键点(key point)、特征点(feature point)概念辨析_yizhang_ml的博客-CSDN博客_角点
https://blog.csdn.net/u010712012/article/details/84947831 | (20条消息) 局部特征——角点检测（Harris，Shi-Tomasi）IT荻的博客-CSDN博客_角点
https://cn.bing.com/search?q=opencv+%e5%bd%a2%e7%8a%b6%e6%a3%80%e6%b5%8b&qs=SC&pq=opencv+xing%27z&sk=SC2&sc=8-13&cvid=659533CA9DD1434BB0418DC35264234B&FORM=QBRE&sp=3 | opencv 形状检测 - 国内版 Bing
https://blog.csdn.net/WZZ18191171661/article/details/91045219 | (20条消息) Python+Opencv实现多种形状的检测_WZZ18191171661的博客-CSDN博客_python 检测
https://blog.csdn.net/Yong_Qi2015/article/details/108675561 | (20条消息) 使用OpenCV进行模糊检测（拉普拉斯算子）_3D视觉工坊-CSDN博客
https://cn.bing.com/search?q=imutils&form=ANNTH1&refig=2a6981e0b327465590092e0fc8c1b117 | imutils - 国内版 Bing
https://cn.bing.com/search?q=python+cv2+%E6%A3%80%E6%B5%8B%E5%AE%9E%E5%BF%83%E6%A4%AD%E5%9C%86&qs=n&form=QBRE&sp=-1&pq=python+cv2+%E6%A3%80%E6%B5%8B%E5%AE%9E%E5%BF%83%E6%A4%AD%E5%9C%86&sc=0-17&sk=&cvid=CC37422B61954A42AFE0BF7D487AB37A | python cv2 检测实心椭圆 - 国内版 Bing
https://cn.bing.com/search?q=%E5%9C%9F%E9%BB%84%E8%89%B2+hsv&cvid=011cadf8958347079d5c864483a3f3f3&aqs=edge..69i57j0.17067j0j4&FORM=ANAB01&PC=HCTS | 土黄色 hsv - 国内版 Bing
https://www.cnblogs.com/long5683/p/9757135.html | OpenCV——RGB和HSV颜色空间 - 一抹烟霞 - 博客园
https://blog.csdn.net/xddwz/article/details/110232977 | (22条消息) python-opencv之圆检测_xddwz的博客-CSDN博客
https://zhuanlan.zhihu.com/p/133531570 | python opencv :使用Hough变换检测圆形，数钱。 - 知乎
https://blog.csdn.net/qq_41662180/article/details/108147118 | (22条消息) Python+opencv — cv2.HoughCircles（）圆形检测，并顺时针给圆编序_不知名的小白66的博客-CSDN博客_cv2 圆形检测
https://blog.csdn.net/qq_42544886/article/details/106701552 | (20条消息) Python3+Opencv 学习笔记（三）图像轮廓处理+描绘矩形框+ 矩形框加文字_辞的博客-CSDN博客
https://blog.csdn.net/qq_33897832/article/details/88931748 | (20条消息) python+opencv图像处理(提取矩形方框提取颜色滤波轮廓检测直线检测)若风的博客-CSDN博客_python矩形检测
https://blog.csdn.net/lionkawoo/article/details/83928116 | (20条消息) 【Python+OpenCV】基于Harris角点的边缘提取以及矩形四角点检测_lionkawoo的博客-CSDN博客_opencv检测矩形4个点坐标
https://blog.csdn.net/wsp_1138886114/article/details/82945328 | (20条消息) OpenCV—Python 轮廓检测绘出矩形框（findContours\ boundingRect\rectangle）_wsp_1138886114的博客-CSDN博客_python 矩形检测
https://blog.csdn.net/HuangZhang_123/article/details/80511270 | (20条消息) Python 使用Opencv实现边缘检测以及轮廓检测_Xy-Huang的博客-CSDN博客_python 轮廓检测
https://blog.csdn.net/m0_51402531/article/details/121066693 | (20条消息) OpenCV——Canny边缘检测（cv2.Canny()）_m0_51402531的博客-CSDN博客_cv2.canny
https://blog.csdn.net/Easen_Yu/article/details/89365497 | (20条消息) cv2.findContours() 轮廓检测_Easen_Yu的博客-CSDN博客
https://blog.csdn.net/gaoranfighting/article/details/34877549 | (20条消息) opencv.cv2.findContours()函数_gaoranfighting的专栏-CSDN博客_cv2.findcontours
https://cn.bing.com/search?q=%E6%BE%B3%E5%A4%A7%E5%88%A9%E4%BA%9A%E6%B3%95%E5%AE%9A%E5%81%87%E6%97%A5&cvid=775bc42c6d794f799a2393030564f52b&aqs=edge.1.69i57j0l7.10679j0j4&FORM=ANAB01&PC=HCTS | 澳大利亚法定假日 - 国内版 Bing
https://cn.bing.com/search?q=%E5%BD%93%E4%BD%BF%E7%94%A8sigmoid%E4%BD%9C%E4%B8%BA%E6%BF%80%E6%B4%BB%E5%87%BD%E6%95%B0%E7%9A%84%E6%97%B6%E5%80%99%EF%BC%8C%E5%B8%B8%E7%94%A8%E4%BA%A4%E5%8F%89%E7%86%B5%E6%8D%9F%E5%A4%B1%E5%87%BD%E6%95%B0%E8%80%8C%E4%B8%8D%E7%94%A8%E5%9D%87%E6%96%B9%E8%AF%AF%E5%B7%AE%E6%8D%9F%E5%A4%B1%E5%87%BD%E6%95%B0&cvid=8dc4d032fafb49d1834698c0c2015e84&aqs=edge..69i57.815j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | 当使用sigmoid作为激活函数的时候，常用交叉熵损失函数而不用均方误差损失函数 - 国内版 Bing
https://blog.csdn.net/Robin___/article/details/105225581 | (22条消息) 当使用Sigmoid或者Softmax作为激活函数时，损失函数推荐使用交叉熵损失函数而不是均方差损失函数_Robin_____的博客-CSDN博客_softmax 均方差
https://blog.csdn.net/lcalqf/article/details/71170171 | (20条消息) opencv cv2.THRESH_OTSU 二值化_lcalqf的专栏-CSDN博客_thresh_otsu
https://blog.csdn.net/qq_39507748/article/details/104506306 | (20条消息) opencv学习笔记七：使用cv2.threshold实现全局和OTSU阈值二值化_耐心的小黑的博客-CSDN博客_cv2 otsu
https://blog.csdn.net/qq_36584673/article/details/121398502 | (20条消息) 【Python+OpenCV】二值化函数cv2.threshold()的用法详解_十八岁小白想成大牛-CSDN博客_cv.threshold python

https://cloud.tencent.com/developer/article/1689064 | 一文读懂 RoIPooling、RoIAlign 和 RoIWarp - 云+社区 - 腾讯云
https://www.jianshu.com/p/2a5ffca8b861 | 从ROI Pooling到ROI Align - 简书
https://blog.csdn.net/weixin_43745234/article/details/119608202 | (23条消息) faster-rcnn详解（rpn，损失函数）_weixin_43745234的博客-CSDN博客_rpn损失函数
https://zhuanlan.zhihu.com/p/72579976 | Faster RCNN的损失函数(Loss Function) - 知乎
https://blog.csdn.net/qq_41828351/article/details/97614901 | (23条消息) Faster R-CNN由深入简_壹砂的博客-CSDN博客
https://www.jianshu.com/p/5ea1b0e8ea9e | Faster RCNN 算法_面试版 - 简书
https://www.jianshu.com/p/aa659406d8c4 | 超详细解读Faster R-CNN-FPN - 简书
https://blog.csdn.net/Mr_health/article/details/84970776 | (23条消息) 【Faster RCNN】损失函数理解_Mr_health的博客-CSDN博客_faster rcnn损失函数
https://www.cnblogs.com/zyly/p/9247863.html | 第三十一节，目标检测算法之 Faster R-CNN算法详解 - 大奥特曼打小怪兽 - 博客园
https://zhuanlan.zhihu.com/p/161540817 | RoI Pooling 与 RoI Align 的区别 - 知乎
https://zhuanlan.zhihu.com/p/75434800 | 目标检测(Object Detection)-Faster RCNN中的ROI Pooling层 - 知乎
https://zhuanlan.zhihu.com/p/59186710 | Faster RCNN之RPN理解 - 知乎
https://zhuanlan.zhihu.com/p/73138740 | ROI Pooling和ROI Align - 知乎
https://blog.csdn.net/gbyy42299/article/details/80352418 | (23条消息) faster rcnn中Roi pooling层的个人理解_gbyy42299的博客-CSDN博客_faster rcnn roi

https://cn.bing.com/search?q=yolo%E5%8F%98%E4%BD%93+2022&qs=n&form=QBRE&sp=-1&pq=yolo%E5%8F%98%E4%BD%93+2022&sc=0-11&sk=&cvid=B0EE29608B5F40259F914AA5B291A966 | yolo变体 2022 - 国内版 Bing
https://zhuanlan.zhihu.com/p/439940127 | arXiv每日更新-2021.12.1（今日关键词：segmentation, recognition, 3d） - 知乎
https://www.163.com/dy/article/G624276P05380EIV.html | CV岗位面试题：简单说下YOLOv1,v2,v3,v4各自的特点与发展史|算法_网易订阅
https://zhuanlan.zhihu.com/p/453052622 | 目标检测 | 刚刚过去的2021年，谁是YOLO系列的最强王者？ - 知乎
https://cloud.tencent.com/developer/article/1897556 | 再改YOLO | YOLO-ReT让边缘端也可以实时检测 - 云+社区 - 腾讯云
https://jishuin.proginn.com/p/763bfbd6ca5f | 2021 年最受欢迎的 YOLO 变体-技术圈
https://blog.csdn.net/amusi1994/article/details/87460874 | (20条消息) 腾讯优图：开源 YOLO 系列代码（含 YOLOv3 以及各种 Backbone）_阿木寺的博客-CSDN博客_yolo开源
https://blog.csdn.net/HiWangWenBing/article/details/122369147 | (20条消息) [YOLO专题-21]：YOLO V5 - ultralytics代码解析-网络配置文件与总体结构_文火冰糖（王文兵）的博客-CSDN博客
https://jishuin.proginn.com/p/763bfbd5917b | YOLO 系目标检测算法家族全景图！-技术圈
https://zhuanlan.zhihu.com/p/403231242 | YoloALL：解决不同版本YOLO的选择进行目标检测推理（源代码已开源） - 知乎
https://cn.bing.com/search?q=arxiv&cvid=4c2c3591d8d44a1daacf8269afa6936c&aqs=edge.2.69i57j0l8.3119j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | arxiv - 国内版 Bing
https://arxiv.org/search/?query=yolo&searchtype=all&source=header | Search | arXiv e-print repository
https://cn.bing.com/search?q=mmoe&form=ANNTH1&refig=c1119d5731b64294b627ca1fe447770b | mmoe - 国内版 Bing

https://www.baidu.com/s?wd=%E8%AE%BE%E8%AE%A1%E6%A8%A1%E5%BC%8F&rsv_spt=1&rsv_iqid=0x85fe7d8800001c2f&issp=1&f=8&rsv_bp=1&rsv_idx=2&ie=utf-8&tn=44004473_35_oem_dg&rsv_enter=1&rsv_dl=ib&rsv_sug3=13&rsv_sug1=18&rsv_sug7=100 | 设计模式_百度搜索
https://cn.bing.com/search?q=%E7%94%9F%E4%BA%A7%E8%80%85%E6%B6%88%E8%B4%B9%E8%80%85%E6%A8%A1%E5%BC%8F&cvid=ce7fbd3c17a847fdac10a149c36e0baa&aqs=edge.3.69i57j0l8.4247j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | 生产者消费者模式 - 国内版 Bing
https://zhuanlan.zhihu.com/p/73442055 | 经典并发同步模式：生产者-消费者设计模式 - 知乎
https://cn.bing.com/search?q=%E5%A4%9A%E7%BA%BF%E7%A8%8B%E8%AE%BE%E8%AE%A1%E6%A8%A1%E5%BC%8F&cvid=93faddbc9d124c26b84c8e5d11e463b2&aqs=edge..69i57j0l2.736j0j4&FORM=ANAB01&PC=HCTS | 多线程设计模式 - 国内版 Bing
edge://newtab/ | 新建标签页

https://cn.bing.com/search?q=bing&cvid=5c623989044f49159d867c3d52401a95&aqs=edge..69i57j0l6j69i60j69i65.1214j0j4&FORM=ANAB01&PC=HCTS | bing - 国内版 Bing
https://cn.bing.com/search?q=ocr+%e7%bb%bc%e8%bf%b0&sp=-1&pq=ocr+%e7%bb%bc%e8%bf%b0&sc=3-6&qs=n&sk=&cvid=C1126F0984C247088D1C89DB55B8C051&first=30&FORM=PORE | ocr 综述 - 国内版 Bing
https://jishuin.proginn.com/p/763bfbd6c8a4 | 最新的一篇视觉Transformer综述！-技术圈
https://kakack.github.io/2020/10/%E6%96%87%E5%AD%97%E8%AF%86%E5%88%ABOCR%E6%96%B9%E6%B3%95%E7%BB%BC%E8%BF%B0%E4%B8%8E%E5%BA%94%E7%94%A8/ | 文字识别ocr方法综述与应用 | Kaka Chen
https://www.bilibili.com/read/cv5724782/ | 25篇最新CV领域综述性论文速递！涵盖15个方向：目标检测/图像处理/姿态估计/医学影像 - 哔哩哔哩
https://www.zhihu.com/question/375768882 | 深度学习ocr识别准确率高的算法是哪些? - 知乎
https://paddlepedia.readthedocs.io/en/latest/tutorials/computer_vision/OCR/OCR.html | OCR综述 — PaddleEdu documentation
https://zhuanlan.zhihu.com/p/75925807 | 写给自己的OCR入门指导（上） - 知乎
https://zhuanlan.zhihu.com/p/65707543 | 文字识别方法整理（20152019） - 知乎
https://blog.csdn.net/amusi1994/article/details/118098111 | (20条消息) 最新！CVPR 2021 OCR领域论文大盘点（22篇）_阿木寺的博客-CSDN博客
https://mp.weixin.qq.com/s?__biz=MzUxNjcxMjQxNg%3D%3D&chksm=f9a1102aced6993c73c32f866c15998ed383c61c2bb832802a15def80a1b83f4dbfd38766fa6&idx=2&mid=2247525029&scene=21&sn=ce3d40ed6ebe9134805f0d04c7983880#wechat_redirect | 最新！CVPR 2021 视觉Transformer论文大盘点（43篇）
https://www.leiphone.com/category/aijuejinzhi/CwL9pKfavCd1v8uM.html | 复杂场景下的 OCR 如何实现——深度学习算法综述 | 雷峰网
https://www.cnblogs.com/skyfsm/p/7923015.html | 【OCR技术系列之一】字符识别技术总览 - Madcola - 博客园
https://zhuanlan.zhihu.com/p/71784363 | 一个关于OCR/STR深度学习算法的综述 - 知乎
https://blog.csdn.net/poppyty/article/details/118189855 | (20条消息) OCR论文综述（含文字识别、文本检测、端到端和数据集合）_poppyty的博客-CSDN博客_ocr文字识别论文
https://blog.csdn.net/wsp_1138886114/article/details/100040857 | (20条消息) OCR综述—持续更新_wsp_1138886114的博客-CSDN博客_ocr论文
https://zhuanlan.zhihu.com/p/81876053 | OCR综述论文笔记 - 知乎
https://blog.csdn.net/weixin_44575517/article/details/109287623 | (20条消息) OCR综述_21世纪打工人的博客-CSDN博客_ocr综述
https://www.sohu.com/a/488190512_121124011 | OCR方向目前最火的repo，绝绝子！_PaddleOCR
https://www.cnblogs.com/xiaoxia722/p/14627482.html | PaddleOCR详解 - xiaoxia0722 - 博客园
https://github.com/pengcao/chinese_ocr | pengcao/chinese_ocr: 中文ocr识别
https://www.csdn.net/tags/NtjaEgysMjgwMDctYmxvZwO0O0OO0O0O.html | (20条消息) ocr识别综述 - CSDN
https://blog.csdn.net/poppyty/article/details/118189855?utm_medium=distribute.pc_aggpage_search_result.none-task-blog-2aggregatepage~~first_rank_ecpm_v1~~rank_v31_ecpm-8-118189855.pc_agg_new_rank&utm_term=ocr%E8%AF%86%E5%88%AB%E7%BB%BC%E8%BF%B0&spm=1000.2123.3001.4430 | (20条消息) OCR论文综述（含文字识别、文本检测、端到端和数据集合）_poppyty的博客-CSDN博客_ocr文字识别论文
https://zhuanlan.zhihu.com/p/311696610?ivk_sa=1024320u | 【语义分割】对象上下文表示：OCR论文笔记 - 知乎
https://blog.csdn.net/chenghaoy/article/details/117294392 | (20条消息) 【ACM 2020 - Text Recognition in the Wild：A Survey】OCR识别综述_chenghaoy的博客-CSDN博客
https://blog.csdn.net/qq_36268755/article/details/105998417 | (20条消息) 【综述】【场景文字识别】Text Recognition in the Wild: A Survey_序章-CSDN博客_场景文字识别算法

https://zhuanlan.zhihu.com/p/56792817 | 目标检测任务的优化策略tricks - 知乎
https://zhuanlan.zhihu.com/p/102817180 | 目标检测比赛中的tricks（已更新更多代码解析） - 知乎
https://zhuanlan.zhihu.com/p/345657476 | 目标检测比赛思路、tricks集锦、资料汇总 - 知乎
https://www.kaggle.com/nvnnghia/yolov5-pseudo-labeling | YoloV5 Pseudo Labeling | Kaggle
https://github.com/Ying-Kang/TianCHI_guangdong/blob/main/eda.ipynb?spm=5176.12282029.0.0.26b21aaa83A2vL&file=eda.ipynb | TianCHI_guangdong/eda.ipynb at main · Ying-Kang/TianCHI_guangdong
https://blog.csdn.net/weixin_44836143/article/details/105952819 | (29条消息) 目标检测（降低误检测率及小目标检测系列笔记）_Z.w.j的博客-CSDN博客_目标检测误检率高怎么办
https://blog.csdn.net/mdjxy63/article/details/88604526 | (29条消息) 目标检测当中的模型融合_mdjxy63的博客-CSDN博客_目标检测模型融合
https://blog.csdn.net/dulingwen/article/details/89669928 | (29条消息) 使用python批量修改图像尺寸、文件名称、xml文件内容、生成无标记xml文件！_dulingwen的博客-CSDN博客
https://cn.bing.com/search?q=%E4%BC%AA%E6%A0%87%E7%AD%BE%E7%AD%96%E7%95%A5&cvid=d1b608781c864be9bc2638a6e8c8ecfb&aqs=edge..69i57.918j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | 伪标签策略 - 国内版 Bing
https://zhuanlan.zhihu.com/p/116065447 | 伪标签学习 ——一种简单有效的半监督学习方法 - 知乎
edge://newtab/ | 新建标签页
edge://history/all?q=kaggle%20target | 历史记录
https://www.kaggle.com/ryanholbrook/target-encoding#Example---MovieLens1M | Target Encoding | Kaggle
https://cn.bing.com/search?q=python+4%E4%B8%AA%E5%80%BC+%E5%88%86%E7%BB%99%E4%B8%A4%E4%B8%AA%E5%8F%98%E9%87%8F&cvid=70207ed0838a464a905409a02bee4613&aqs=edge..69i57.17295j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | python 4个值分给两个变量 - 国内版 Bing
https://blog.csdn.net/qq_27283619/article/details/88832064 | (29条消息) Python3中实现一行输入多个数字并进行赋值_Jock2018的博客-CSDN博客_python 输入3个数字
https://cn.bing.com/search?q=python+tutor&cvid=2c894a4457024faebd3da6eb2ecfc455&aqs=edge.1.69i57j69i59j69i60l3.7780j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | python tutor - 国内版 Bing
https://pythontutor.com/visualize.html#mode=display | Python Tutor - Visualize Python, Java, JavaScript, C, C++, Ruby code execution
edge://newtab/ | 新建标签页
https://cn.bing.com/search?q=pathlib+Path+%E8%BD%AC%E5%AD%97%E7%AC%A6%E4%B8%B2&cvid=7fd7f533b5e347939829712eb968adbb&aqs=edge..69i57.12432j0j4&FORM=ANAB01&PC=HCTS | pathlib Path 转字符串 - 国内版 Bing
edge://newtab/ | 新建标签页
https://cn.bing.com/search?q=%E6%9E%81%E8%A7%86%E8%A7%92&cvid=a2926503cad74d79b37beb1e84b84a5d&aqs=edge..69i57.10383j0j4&FORM=ANAB01&PC=HCTS | 极视角 - 国内版 Bing
https://cvmart.net/topList | 极市开发者平台
https://github.com/4paradigm/autox | 4paradigm/AutoX: AutoX is an efficient automl tool, which is mainly aimed at data mining competitions with tabular data.

https://blog.csdn.net/junruitian/article/details/105689571 | (18条消息) 常见损失函数和评价指标总结（附公式&代码）_junruit的博客-CSDN博客_最小化指数损失函数
https://zhuanlan.zhihu.com/p/58883095 | 常见的损失函数(loss function)总结 - 知乎
https://zhuanlan.zhihu.com/p/91511706 | 常用损失函数和评价指标总结 - 知乎
https://jishuin.proginn.com/p/763bfbd5c12f | 一文看尽深度学习中的15种损失函数-技术圈
https://blog.csdn.net/GrayOnDream/article/details/102955297 | (20条消息) 23种激活函数_梦的灰色边沿…-CSDN博客_silu激活函数
https://cn.bing.com/search?q=%E7%94%A8%E4%BA%A4%E5%8F%89%E7%86%B5%E6%8D%9F%E5%A4%B1%E5%87%BD%E6%95%B0%E8%80%8C%E4%B8%8D%E7%94%A8%E5%9D%87%E6%96%B9%E8%AF%AF%E5%B7%AE%E6%8D%9F%E5%A4%B1%E5%87%BD%E6%95%B0&qs=n&form=QBRE&sp=-1&pq=%E7%94%A8%E4%BA%A4%E5%8F%89%E7%86%B5%E6%8D%9F%E5%A4%B1%E5%87%BD%E6%95%B0%E8%80%8C%E4%B8%8D%E7%94%A8%E5%9D%87%E6%96%B9%E8%AF%AF%E5%B7%AE%E6%8D%9F%E5%A4%B1%E5%87%BD%E6%95%B0&sc=0-19&sk=&cvid=FBCD39BEE30F4A33AC9B2DAD0C925663 | 用交叉熵损失函数而不用均方误差损失函数 - 国内版 Bing
https://www.cnblogs.com/shine-lee/p/12032066.html | 直观理解为什么分类问题用交叉熵损失而不用均方误差损失? - shine-lee - 博客园
https://zhuanlan.zhihu.com/p/104130889 | 为什么分类问题的损失函数采用交叉熵而不是均方误差MSE？ - 知乎
https://www.nowcoder.com/discuss/807115?channel=-1&source_id=discuss_terminal_discuss_sim_nctrack&ncTraceId=82bb8eed4d174c488b354d5acd5d4972.251.16414528403236408 | (1条未读通知) 算法工程师-机器学习面试高频知识点_资源分享_牛客网
https://zhuanlan.zhihu.com/p/70804197 | 为什么用交叉熵做损失函数 - 知乎
https://blog.csdn.net/qq_29679623/article/details/99441913 | (20条消息) 为什么sigmoid激活函数，使用交叉熵损失函数更好。详细推导过程：_奔跑的汉堡包的博客-CSDN博客_sigmoid 交叉熵

edge://newtab/ | 新建标签页
https://github.com/datawhalechina/fun-rec/blob/master/docs/%E7%AC%AC%E4%BA%8C%E7%AB%A0%20%E6%8E%A8%E8%8D%90%E7%B3%BB%E7%BB%9F%E5%AE%9E%E6%88%98/2.1%E7%AB%9E%E8%B5%9B%E5%AE%9E%E8%B7%B5/jupyter/2.1%20%E8%B5%9B%E9%A2%98%E7%90%86%E8%A7%A3%2BBaseline.ipynb
https://blog.csdn.net/Swocky/article/details/105809498 | (28条消息) 【深度学习】深度学习模型训练的tricks总结_无极阁-CSDN博客_深度学习的tricks
https://aistudio.baidu.com/aistudio/index | 飞桨AI Studio - 人工智能学习实训社区
https://cn.bing.com/search?q=shutil+copy+%E6%B2%A1%E6%9C%89%E6%9D%83%E9%99%90+operation+not+Permitted&qs=n&form=QBRE&sp=-1&pq=shutil+copy+%E6%B2%A1%E6%9C%89%E6%9D%83%E9%99%90+operation+not+permitte&sc=0-39&sk=&cvid=BF02E253A1774320B3307B883BAB8659 | shutil copy 没有权限 operation not Permitted - 国内版 Bing
https://stackoverflow.com/questions/11835833/why-would-shutil-copy-raise-a-permission-exception-when-cp-doesnt | python - Why would shutil.copy() raise a permission exception when cp doesn’t? - Stack Overflow
https://blog.csdn.net/Bit_Coders/article/details/113867208 | (29条消息) 解决shutil.copyfile常见错误：PermissionError: [Errno 13] Permission denied_人工智能算法与工程实践-CSDN博客_shutil.error
https://www.cnblogs.com/keeptg/p/10944109.html | 用Python复制文件的9个方法（转） - keeptg - 博客园
https://www.cnblogs.com/Micro0623/p/12778798.html | part12：Python 文件I/O（pathlib模块：PurePath、Path，os.path，fnmatch，open，with，linecache，os模块操作文件和目录，tempfile） - 远方那一抹云 - 博客园
https://cuiqingcai.com/6598.html | Python3 中使用 Pathlib 模块进行文件操作 | 静觅
https://blog.csdn.net/dorlolo/article/details/120507924 | (29条消息) python os模块和pathlib模块复制文件方法对比_dorlolo的博客-CSDN博客
https://learnku.com/python/t/61521?order_by=created_at& | 为什么pathlib.Path.glob() 和rglob()一样支持递归？ | Python | Python 技术论坛
https://aistudio.baidu.com/aistudio/projectdetail/3142643 | XMRec WSDM CUP 亚马逊跨境电商推荐topline - 飞桨AI Studio - 人工智能学习实训社区
https://competitions.codalab.org/competitions/36050#participate-get-data | CodaLab - Competition
https://competitions.codalab.org/competitions/36050#participate-get_data | CodaLab - Competition
https://github.com/LogicJake/competition_baselines/blob/master/competitions/wsdm_iqiyi/baseline.ipynb | competition_baselines/baseline.ipynb at master · LogicJake/competition_baselines
https://competitions.codalab.org/competitions/36050#learn_the_details | CodaLab - Competition
http://challenge.ai.iqiyi.com/detail?raceId=61600f6cef1b65639cd5eaa6 | 用户留存预测挑战赛
https://cn.bing.com/search?q=python+command%E5%91%BD%E4%BB%A4%E8%A1%8C&cvid=c8239641b46448dea05f962cc59fc626&aqs=edge..69i57.9843j0j4&FORM=ANAB01&PC=HCTS | python command命令行 - 国内版 Bing
https://cloud.tencent.com/developer/article/1540943 | 在python中运行命令行命令的四种方案 - 云+社区 - 腾讯云
https://mp.weixin.qq.com/s/1IADiDb60NQVzUBhLnbkhA | 全网首发！Kaggle操作完整指南！
https://mp.weixin.qq.com/s/S4Ud59yWOjD0FM4EH0iJog | WSDM：亚马逊跨境电商推荐赛Topline分享
edge://newtab/ | 新建标签页
https://www.baidu.com/?tn=44004473_35_oem_dg | 百度一下，你就知道
http://localhost:8888/lab/tree/%E7%AC%AC3%E7%AB%A0%20%E5%88%86%E7%BB%84.ipynb | 第3章分组.ipynb - JupyterLab
https://github.com/datawhalechina/fun-rec/tree/master/docs | fun-rec/docs at master · datawhalechina/fun-rec
https://cn.bing.com/search?q=item+CF&cvid=b969102c0b7b415e92b26fb6abd44c00&aqs=edge..69i57j69i60.2219j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | item CF - 国内版 Bing
https://blog.csdn.net/weixin_37999899/article/details/89456091 | (32条消息) 协同过滤（Item cf）_人生苦短，我用Python-CSDN博客_item协同过滤
https://cn.bing.com/search?q=ocr+pytorch&form=ANNTH1&refig=b5fe1fdd6f7c47ac994332f3f3486cd0 | ocr pytorch - 国内版 Bing
https://zhuanlan.zhihu.com/p/150704536 | PytorchOCR：一个基于pytorch的OCR算法库 - 知乎
https://cn.bing.com/search?q=OCR++2021+%E7%9F%A5%E4%B9%8E%E4%B8%93%E6%A0%8F&qs=n&form=QBRE&sp=-1&pq=ocr+2021+%E7%9F%A5%E4%B9%8E%E4%B8%93%E6%A0%8F&sc=0-13&sk=&cvid=ECD306642A1040A0A7DF0787F5A15502 | OCR 2021 知乎专栏 - 国内版 Bing
https://zhuanlan.zhihu.com/p/385011404 | CVPR 2021 论文大盘点-文本检测篇 - 知乎
http://xxx.itp.ac.cn/pdf/2105.05486.pdf | 2105.05486.pdf
https://zhuanlan.zhihu.com/p/451588262 | TencentOCR 斩获 ICDAR 2021 三项冠军 - 知乎
https://zhuanlan.zhihu.com/p/134115239 | 更更好用的中文OCR—— cnocr-V2.0 - 知乎
https://zhuanlan.zhihu.com/p/65707543 | 文字识别方法整理（2015~2019） - 知乎
https://zhuanlan.zhihu.com/p/259733173 | PP-OCR：实用超轻量级文本识别系统设计详解 - 知乎
https://www.zhihu.com/column/c_1270045584586620928 | OCR专栏 - 知乎
https://zhuanlan.zhihu.com/p/45376274 | OCR技术简介 - 知乎
https://www.zhihu.com/question/36500536 | 图像识别中，目标分割、目标识别、目标检测和目标跟踪这几个方面区别是什么？ - 知乎
https://www.cnblogs.com/skyfsm/p/9776611.html | 【OCR技术系列之五】自然场景文本检测技术综述（CTPN, SegLink, EAST） - Madcola - 博客园
https://cloud.tencent.com/developer/news/437208 | 金连文：“文字检测与识别：现状及展望” - 云+社区 - 腾讯云
https://jishuin.proginn.com/p/763bfbd5caaa | 复旦邱锡鹏组最新综述：A Survey of Transformers！-技术圈
https://blog.csdn.net/qq_18315295/article/details/104355658 | (18条消息) 文本检测与识别（OCR）资源汇总（持续更新）_chenbo的博客-CSDN博客_文本检测与识别
https://zhuanlan.zhihu.com/p/127012959 | 25篇最新CV领域综述论文速递！涵盖15个方向：目标检测/图像处理/姿态估计/医学影像/人脸识别等 - 知乎
https://zhuanlan.zhihu.com/p/38655369 | 自然场景文本检测识别技术综述 - 知乎

edge://newtab/ | 新建标签页
https://competitions.codalab.org/competitions/36050#participate-get-data | CodaLab - Competition
https://competitions.codalab.org/competitions/36050# | CodaLab - Competition
https://github.com/hamedrab/wsdm22_cup_xmrec | hamedrab/wsdm22_cup_xmrec: WSDM 2022 CUP - Cross-Market Recommendation - Starter Kit
https://cn.bing.com/search?q=%e7%9b%ae%e6%a0%87%e6%a3%80%e6%b5%8b%e6%af%94%e8%b5%9btrick&FORM=QSRE1 | 目标检测比赛trick - 国内版 Bing
https://zhuanlan.zhihu.com/p/345657476 | 目标检测比赛思路、tricks集锦、资料汇总 - 知乎
https://github.com/summeridea/-Kaggle-Click-Through-Rate-Prediction- | summeridea/-Kaggle-Click-Through-Rate-Prediction-: CTR-Prediction
https://github.com/songgc/display-advertising-challenge | songgc/display-advertising-challenge: Criteo/Kaggle Competition of CTR prediction
https://blog.csdn.net/weixin_44836143/article/details/105952819 | (24条消息) 目标检测（降低误检测率及小目标检测系列笔记）_Z.w.j的博客-CSDN博客
https://blog.csdn.net/moxibingdao/article/details/112598106 | (24条消息) 目标检测多模型集成方法总结_我爱计算机视觉-CSDN博客
https://blog.csdn.net/mdjxy63/article/details/88604526 | (24条消息) 目标检测当中的模型融合_mdjxy63的博客-CSDN博客_目标检测模型融合
https://blog.csdn.net/gaotihong/article/details/104527801 | (24条消息) 目标检测之四：预测框微调/投票法/模型融合_gaotihong的博客-CSDN博客_目标检测模型融合
https://zhuanlan.zhihu.com/p/141685352 | 盘点目标检测中的特征融合技巧（根据YOLO v4总结） - 知乎
https://blog.csdn.net/Yemiekai/article/details/117839686 | (24条消息) 加权框融合 WBF（Weighted Boxes Fusion: combining boxes for object detection models）_Yemiekai的博客-CSDN博客
https://blog.csdn.net/moxibingdao/article/details/106667623 | (24条消息) 目标检测比赛中的trick_我爱计算机视觉-CSDN博客
https://blog.csdn.net/justsolow/article/details/105587929 | (24条消息) 目标检测比赛的trick_justsolow的博客-CSDN博客
https://blog.csdn.net/e01528/article/details/82354477 | (24条消息) 目标检测比赛的奇技淫巧（tricks）_Snoopy_Dream-CSDN博客_目标检测比赛经验
https://zhuanlan.zhihu.com/p/102817180 | 目标检测比赛中的tricks（已更新更多代码解析） - 知乎
https://zhuanlan.zhihu.com/p/138855612 | 目标检测中的Tricks - 知乎
https://zhuanlan.zhihu.com/p/56792817 | 目标检测任务的优化策略tricks - 知乎
https://zhuanlan.zhihu.com/p/53849733 | CNN训练分类任务的优化策略(tricks) - 知乎
https://blog.csdn.net/tomy2426214836/article/details/106414703 | (24条消息) 迁移学习的几种迁移方式_无崖子的博客-CSDN博客_迁移学习方法
https://zhuanlan.zhihu.com/p/137968529 | 深度学习模型训练过程 - 知乎
https://blog.csdn.net/xys430381_1/article/details/85141702 | (24条消息) 深度学习优化策略—炼丹trick之正确使用BN（训练和测试/预测时怎么用）_xys430381_1的专栏-CSDN博客
https://blog.csdn.net/qq_41375318/article/details/112603611 | (24条消息) 深度检测模型比赛训练技巧（Tricks）_Major_S的博客-CSDN博客
https://zhuanlan.zhihu.com/p/100252944 | 半监督深度学习训练和实现小Tricks - 知乎
https://blog.csdn.net/syysyf99/article/details/96861770 | (24条消息) 深度学习训练技巧总结_夏日山间过-CSDN博客_深度学习训练
https://zhuanlan.zhihu.com/p/144737551 | 深度学习中网络训练技巧 - 知乎
https://zhuanlan.zhihu.com/p/57208761 | [机器学习]偏差和方差的理解 - 知乎
https://www.kaggle.com/c/kkbox-music-recommendation-challenge/discussion/46169 | WSDM - KKBox’s Music Recommendation Challenge | Kaggle
https://cn.bing.com/search?q=WSDM&cvid=0b60cc2b113743f484158531c72d1535&aqs=edge..69i57j0l6j69i60l2.1166j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | WSDM - 国内版 Bing
https://cn.bing.com/search?q=GBDT+%2B+LR++CTR+ID%E7%B1%BB+%E5%92%8C+%E9%9D%9EID&qs=n&form=QBRE&sp=-1&pq=gbdt+%2B+lr+ctr+id%E7%B1%BB+%E5%92%8C+%E9%9D%9Eid&sc=0-23&sk=&cvid=73C6F0F901A141A0BCF56C2BB16463BC | GBDT + LR CTR ID类和非ID - 国内版 Bing
https://www.cnblogs.com/zhazhaacmer/p/13904575.html | 好好学习，天天向上
https://www.jianshu.com/p/73adce3fd975 | GBDT+LR CTR预估-Kaggle实例[有数据集] - 简书
https://zhuanlan.zhihu.com/p/147674729 | CTR预估之GDBT+LR - 知乎
https://cn.bing.com/search?q=CTR+%E9%9D%9EID%E7%B1%BB%E6%A0%91+%E7%A4%BA%E4%BE%8B&qs=n&form=QBRE&sp=-1&pq=ctr+%E9%9D%9Eid%E7%B1%BB%E6%A0%91+%E7%A4%BA%E4%BE%8B&sc=0-12&sk=&cvid=3800281A16F741B98CEDB282DE375CD9 | CTR 非ID类树示例 - 国内版 Bing
https://zhuanlan.zhihu.com/p/152525097 | 推荐系统（5）基于LR+XGBoost预估电信客户流失 - 知乎
https://cn.bing.com/search?q=%e8%a7%a3%e5%86%b3github%e5%9b%bd%e5%86%85%e6%89%93%e4%b8%8d%e5%bc%80%e7%9a%84%e8%a7%a3%e5%86%b3%e6%96%b9%e6%b3%95&FORM=R5FD1 | 解决github国内打不开的解决方法 - 国内版 Bing
https://get.qiaobuqiao.com/post-5022 | 最新亲测有效GitHub打不开、进不去、被墙无法访问的解决方法，打开慢不稳定也可以试一下 | Get巧不巧
https://cn.bing.com/search?q=python+%E4%BB%A5%E7%AE%A1%E7%90%86%E5%91%98%E8%BA%AB%E4%BB%BD&cvid=0c12f71b1533444db3c989b7b389ab2d&aqs=edge..69i57.9648j0j4&FORM=ANAB01&PC=HCTS | python 以管理员身份 - 国内版 Bing
https://zhuanlan.zhihu.com/p/430699625 | github网站打不开怎么办（无需改host） - 知乎
https://zhuanlan.zhihu.com/p/139783331 | python路径操作新标准：pathlib 模块 - 知乎
https://github.com/jackhuntcn?tab=repositories | jackhuntcn (Heng Zheng) / Repositories
https://www.jianshu.com/p/16aa7aa9fa6a | windows命令之copy——复制文件或隐藏文件 - 简书
https://www.jb51.net/article/18986.htm | cmd move命令移动文件（夹）_DOS/BAT_脚本之家
https://www.cnblogs.com/jepson6669/p/9737645.html | 修改Hosts文件提示没有权限怎么办 - Jepson6669 - 博客园
https://github.com/datawhalechina/fun-rec/tree/407fbd8459e4941b36d1b52ded9d779ee59d588b/docs | fun-rec/docs at 407fbd8459e4941b36d1b52ded9d779ee59d588b · datawhalechina/fun-rec
https://github.com/datawhalechina/fun-rec/blob/407fbd8459e4941b36d1b52ded9d779ee59d588b/codes/base_models/DIN.py | fun-rec/DIN.py at 407fbd8459e4941b36d1b52ded9d779ee59d588b · datawhalechina/fun-rec
https://github.com/SYC1123?tab=repositories | SYC1123 / Repositories
https://github.com/datawhalechina/fun-rec/blob/master/docs/%E7%AC%AC%E4%BA%8C%E7%AB%A0%20%E6%8E%A8%E8%8D%90%E7%B3%BB%E7%BB%9F%E5%AE%9E%E6%88%98/2.1%E7%AB%9E%E8%B5%9B%E5%AE%9E%E8%B7%B5/jupyter/2.1%20%E8%B5%9B%E9%A2%98%E7%90%86%E8%A7%A3%2BBaseline.ipynb | fun-rec/2.1 赛题理解+Baseline.ipynb at master · datawhalechina/fun-rec
https://cn.bing.com/search?q=%E8%85%BE%E8%AE%AF+%E6%8A%80%E6%9C%AF%E5%9B%A2%E9%98%9F&qs=n&form=QBRE&sp=-1&pq=%E8%85%BE%E8%AE%AF+%E6%8A%80%E6%9C%AF%E5%9B%A2%E9%98%9F&sc=2-7&sk=&cvid=FCFBA83E495F47ED8C3E89B68E9CA6BE | 腾讯技术团队 - 国内版 Bing
https://www.infoq.cn/article/20-outstanding-enterprise-technology-blog/?luicode=10000359 | 最值得关注的20个优秀企业技术博客-InfoQ
https://zhuanlan.zhihu.com/p/76375442 | 推荐系统阅读笔记：PGPR - 知乎
https://www.zhihu.com/people/bi-shu-xian | Gaby - 知乎
https://www.zhihu.com/org/ji-shi-jiao-14/posts | 极市平台 - 知乎
https://github.com/chendayin/FangTianXiaProject/tree/master/FangTianXiaDistributed | FangTianXiaProject/FangTianXiaDistributed at master · chendayin/FangTianXiaProject
https://github.com/LMFrank/CrawlerProject | LMFrank/CrawlerProject: 爬虫项目：链家网（普通/scrapy）、虎扑、维基百科、百度地图api、房天下（分布式爬虫）、微信公众号（代理池爬取）
https://blog.csdn.net/weixin_42029733/article/details/101056059 | (24条消息) Python: 房天下小区数据爬取_Cyber的博客-CSDN博客
https://blog.csdn.net/sereasuesue/article/details/115551661 | (25条消息) 爬虫实战-爬取房天下网站全国所有城市的新房和二手房信息（最新）_sereasuesue的博客-CSDN博客_爬取房天下二手房信息
https://github.com/datawhalechina/statistical-learning-method-solutions-manual | datawhalechina/statistical-learning-method-solutions-manual: 《统计学习方法》（第二版）习题解答，在线阅读地址：https://datawhalechina.github.io/statistical-learning-method-solutions-manual
https://zhuanlan.zhihu.com/p/33177517 | 推荐系统中使用ctr排序的f(x)的设计-dnn篇之PNN模型 - 知乎
https://zhuanlan.zhihu.com/p/56651241 | 详解 PNN: 在NN中显式引入Product - 知乎
https://blog.csdn.net/qfikh/article/details/105227754 | (25条消息) DeepFM对FM、FNN、PNN、wide&deep模型_qfikh的博客-CSDN博客
https://blog.csdn.net/qfikh/article/month/2019/12 | (27条消息) 2019年12月_qfikh的博客_人鱼线_CSDN博客
https://blog.csdn.net/qfikh/article/month/2019/12 | (28条消息) 2019年12月_qfikh的博客_人鱼线_CSDN博客
https://zhuanlan.zhihu.com/p/267219459 | PNN模型 - 知乎
https://blog.csdn.net/qq_18293213/article/details/90262378 | (25条消息) 推荐算法-PNN(Product Network)_Silence的博客-CSDN博客_pnn
https://cn.bing.com/search?q=wide%26deep+pytorch&cvid=d8151e3fe57047c7b9e24136a9e007bc&aqs=edge.2.69i57j69i59l2j69i61.3549j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | wide&deep pytorch - 国内版 Bing
https://zhuanlan.zhihu.com/p/132708525 | Wide&Deep模型原理与实现 - 知乎
https://github.com/kaitolucifer/wide-and-deep-learning-keras | kaitolucifer/wide-and-deep-learning-keras: Wide and Deep Learning implemented with Keras
https://blog.csdn.net/wuzhongqiang/article/details/109254498 | (26条消息) AI上推荐之 Wide&Deep与Deep&Cross模型(记忆与泛化并存的华丽转身）_Miracle8070-CSDN博客
https://cn.bing.com/search?q=FTRL%E7%AE%97%E6%B3%95&cvid=b883d3098c6348cca0089cf3f4efb0ab&aqs=edge..69i57.1094j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | FTRL算法 - 国内版 Bing
https://www.jianshu.com/p/e07ddc3ace8f | FTRL论文笔记 - 简书
https://cn.bing.com/search?q=deepFM+pytorch+%E5%AE%9E%E7%8E%B0&cvid=9ec7c28079ea4a18a8b54bf0d01fc765&aqs=edge..69i57.11077j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | deepFM pytorch 实现 - 国内版 Bing
https://github.com/guoyang9/NFM-pyorch/blob/master/model.py | NFM-pyorch/model.py at master · guoyang9/NFM-pyorch
https://zhuanlan.zhihu.com/p/33587540 | 推荐系统中使用ctr排序的f(x)的设计-dnn篇之NFM模型 - 知乎
https://wy-ei.github.io/notebook/2019/fm-family-code/ | FM, FFM, DFM, NFM, Wide & Deep 实现 - WangYu’s Notebook
https://zhuanlan.zhihu.com/p/332786045 | 清晰易懂，基于pytorch的DeepFM的完整实验代码 - 知乎
https://cn.bing.com/search?q=AFM+pytorch&cvid=c95cfb08ae3d4b83a7bec68b8332df3e&aqs=edge..69i57.3360j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | AFM pytorch - 国内版 Bing
https://www.cnblogs.com/sunupo/p/12862852.html | AFM模型 pytorch示例代码 - 懒惰的星期六 - 博客园
https://blog.csdn.net/oppo62258801/article/details/104146456 | (27条消息) 深度CTR之AFM：基于Attention网络的FM模型_数据之美的博客-CSDN博客_afm网络
https://blog.csdn.net/wuzhongqiang/article/details/109532346 | (27条消息) AI上推荐之 AFM与DIN模型（当推荐系统遇上了注意力机制）_Miracle8070-CSDN博客
https://cn.bing.com/search?q=DIN+pytorch&cvid=3459b9280c2f42cbafdd94f59dea540e&aqs=edge..69i57.5416j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | DIN pytorch - 国内版 Bing
https://blog.csdn.net/weixin_35154281/article/details/112275652 | (28条消息) 推荐系统之DIN代码详解_张先生－您好的博客-CSDN博客_din 序列推荐
https://github.com/waxxyybb/DIN-pytorch/blob/master/DataLoader.py#L74 | DIN-pytorch/DataLoader.py at master · waxxyybb/DIN-pytorch
edge://newtab/ | 新建标签页
https://labuladong.gitee.io/algo/ | labuladong 的算法小抄 :: labuladong的算法小抄
https://mp.weixin.qq.com/s/wIxflO1dvXzDlibhEcENcQ | 刷题插件 V2.4 功能手册
http://hanpy.plus/2021/03/26/din/ | DIN – Rb’s Hanpy Blog

https://zhuanlan.zhihu.com/p/268787224 | NLP实战篇之tf2自定义Layer与Model - 知乎

edge://newtab/ | 新建标签页

https://cn.bing.com/search?q=tensorflow+%e5%a4%a7%e6%95%b0%e6%8d%ae%e8%af%bb%e5%8f%96&qs=n&sp=-1&pq=tensorflow+%e5%a4%a7%e6%95%b0%e6%8d%ae%e8%af%bb%e5%8f%96&sc=0-16&sk=&cvid=14F78F355163499FB5B79CBE8615DBEE&first=20&FORM=PORE | tensorflow 大数据读取 - 搜索
https://www.cnblogs.com/chaofn/p/9309560.html | Tensorflow 大规模数据集训练方法 - SuperVan - 博客园
https://blog.csdn.net/u014061630/article/details/80712635 | (22条消息) TensorFlow 数据读取方法总结_黑暗星球-CSDN博客_tensorflow读取数据
https://blog.csdn.net/qq_41427568/article/details/85801579 | (22条消息) Tensorflow读取大数据集的方法，tf.train.string_input_producer()和tf.train.slice_input_producer()_Seeklhy的博客-CSDN博客_tf.train.string_input_producer
https://www.jianshu.com/p/c29965e6c40c | TensorFlow中的多线程 - 简书
https://cloud.tencent.com/developer/article/1454133#:~:text=TensorFl,%E7%BA%BF%E7%A8%8B%EF%BC%8C%E5%B9%B6%E8%A1%8C%E6%89%A7%E8%A1%8C%E3%80%82 | tensorflow中协调器 tf.train.Coordinator - 云+社区 - 腾讯云
https://www.helloworld.net/p/5968377563 | Tensorflow多线程输入数据处理框架（一）——队列与多线程 - HelloWorld开发者社区
https://juejin.cn/post/6955442027849285640 | 6.1 Tensorflow笔记(基础篇):队列与线程 - 掘金
https://cloud.tencent.com/developer/article/1010066 | TensorFlow 队列与多线程的应用 - 云+社区 - 腾讯云
https://www.jianshu.com/p/5a885954173d | TensorFlow 教程1：线程和队列 - 简书

https://blog.csdn.net/qq_35812205/article/details/121031616 | (22条消息) 【王喆-推荐系统】线上服务篇-(task5)部署离线模型_提出问题和解决问题同等重要-CSDN博客

https://blog.csdn.net/u014636245/article/details/84147581 | (22条消息) 【TensorFlow】多GPU训练：示例代码解析_TomRen-CSDN博客_多卡训练
https://blog.csdn.net/BmwGaara/article/details/94021563?utm_medium=distribute.wap_aggpage_search_result.none-task-blog-2~aggregatepage~first_rank_ecpm_v1~rank_aggregation-3-94021563.wap_agg_rank_aggregation&utm_term=tensorflow%E5%A4%9A%E6%9C%BA%E5%A4%9A%E5%8D%A1%E8%AE%AD%E7%BB%83 | (22条消息) 基于TensorFlow的多机多卡分布式训练搭建方法_BmwGaara的博客-CSDN博客
https://blog.csdn.net/qq_36178899/article/details/84369074 | (22条消息) 快速上手多机多卡的分布式tensorflow_Orbita_wangtao-CSDN博客
https://www.cnblogs.com/yunke-ws/p/9998674.html | 快速上手多机多卡的分布式tensorflow - ywwhxxtw - 博客园
https://blog.csdn.net/yjk13703623757/article/details/80956268/ | (22条消息) 分布式TensorFlow多主机多GPU原理与实现_yjk13703623757的博客-CSDN博客
https://blog.csdn.net/qq_38293297/article/details/105593100?utm_medium=distribute.wap_aggpage_search_result.none-task-blog-2~aggregatepage~first_rank_ecpm_v1~rank_aggregation-2-105593100.wap_agg_rank_aggregation&utm_term=tensorflow%E5%A4%9A%E6%9C%BA%E5%A4%9A%E5%8D%A1%E8%AE%AD%E7%BB%83 | (22条消息) Tensorflow之单机多卡程序[1.0和2.0版本]_晴晴_Amanda的博客-CSDN博客
https://blog.csdn.net/weixin_39559333/article/details/111962722 | (22条消息) tensorflow 多卡测试_tensorflow使用多个gpu训练_weixin_39559333的博客-CSDN博客
https://cn.bing.com/search?q=argumentparser+%E6%A8%A1%E6%9D%BF&qs=n&form=QBRE&sp=-1&pq=argumentparser+%E6%A8%A1%E6%9D%BF&sc=0-17&sk=&cvid=F343344095B04F49B4A824710CD27A59 | argumentparser 模板 - 搜索
https://blog.csdn.net/weixin_45884316/article/details/108297981 | (22条消息) Python argparse模块详解_Running Snail-CSDN博客
https://blog.csdn.net/weixin_31445091/article/details/113637972?utm_medium=distribute.pc_relevant.none-task-blog-2~default~baidujs_title~default-0.no_search_link&spm=1001.2101.3001.4242.1 | (22条消息) python argparse模块详解多级_Python argparse模块详解_章鱼老师的博客-CSDN博客
https://zhuanlan.zhihu.com/p/56922793 | argparse模块用法实例详解 - 知乎
https://blog.csdn.net/The_Time_Runner/article/details/97941409 | (22条消息) argparse.ArgumentParser()用法解析_漫步量化-CSDN博客
edge://newtab/ | 新建标签页
https://blog.csdn.net/Forrest97/article/details/105882837/ | (22条消息) Tensorflow2.0 自定义图像数据集读取加速 tf.data.Dataset.cache (二)_Forrest97的博客-CSDN博客
https://cloud.tencent.com/developer/article/1788266 | TensorFlow 2.0 - Keras Pipeline、自定义Layer、Loss、Metric - 云+社区 - 腾讯云
https://blog.csdn.net/qq_27825451/article/details/90739257 | (22条消息) tensorflow的Data Pipeline系列教程（一）——Dataset类的属性即常用方法_MIss-Y的博客-CSDN博客
https://zhuanlan.zhihu.com/p/27238630 | 十图详解tensorflow数据读取机制（附代码） - 知乎
https://blog.csdn.net/Triple_WDF/article/details/109188076 | (22条消息) TensorFlow2.0 大数据预处理Pipeline_猫ER-CSDN博客
https://pfeiwang.blog.csdn.net/article/details/106927881 | (22条消息) TensorFlow2.0 分批读取数据集、训练_猫ER-CSDN博客
https://pfeiwang.blog.csdn.net/article/details/106929356 | (22条消息) TensorFlow2.0 利用TFRecord存取数据集，分批次读取训练_猫ER-CSDN博客
https://blog.csdn.net/qq_36758914/article/details/104736266 | (22条消息) Tensorflow2.0加载和预处理数据的方法汇总_cocofisher的博客-CSDN博客
https://www.cnblogs.com/zyly/p/8982335.html | 第十二节，TensorFlow读取数据的几种方法以及队列的使用 - 大奥特曼打小怪兽 - 博客园
https://blog.csdn.net/XST1520203418/article/details/121322531 | (22条消息) tensorflow与深度学习之队列与线程_XST1520203418的博客-CSDN博客
https://www.cnblogs.com/chaofn/p/9309560.html | Tensorflow 大规模数据集训练方法 - SuperVan - 博客园
https://blog.csdn.net/SweetWind1996/article/details/103361915 | (22条消息) TensorFlow2.0 TFrecord数据集的写入、读取和训练示例详解_SweetWind1996的博客-CSDN博客
https://zhuanlan.zhihu.com/p/100896019 | TensorFlow 2.0 制作TFRecord格式的数据集 - 知乎
https://blog.csdn.net/lzs781/article/details/104741946 | (22条消息) TensorFlow2数据加载与数据集_刘之帅-CSDN博客_tensorflow2数据集
https://blog.csdn.net/nanke_4869/article/details/114211215 | (22条消息) 【Tensorflow教程笔记】TensorFlow Datasets 数据集载入_nanke_4869的博客-CSDN博客
https://blog.csdn.net/nanke_4869/article/details/114107052 | (22条消息) 【Tensorflow教程笔记】Keras Pipeline_nanke_4869的博客-CSDN博客
https://blog.csdn.net/nanke_4869/article/details/114209831 | (22条消息) 【Tensorflow教程笔记】常用模块 tf.data ：数据集的构建与预处理_nanke_4869的博客-CSDN博客
https://zhuanlan.zhihu.com/p/140343833 | [腾讯机智] tensorflow profiling工具简介——tensorflow原生工具 - 知乎
https://zhuanlan.zhihu.com/p/147283610 | 命令行参数–tf.app.flags和python argparse - 知乎
https://blog.csdn.net/lyc_yongcai/article/details/73456960 | (22条消息) TensorFlow 中 tf.app.flags.FLAGS 的用法介绍_lyc_yongcai的博客-CSDN博客
https://cn.bing.com/search?q=tensorflow+%E8%AF%BB%E5%8F%96%E5%A4%9A%E4%B8%AA%E6%96%87%E4%BB%B6&qs=n&form=QBRE&sp=-1&pq=tensorflow+%E8%AF%BB%E5%8F%96%E5%A4%9A%E4%B8%AA%E6%96%87%E4%BB%B6&sc=0-17&sk=&cvid=E6A76BA885B44DC4B42546D574D7DCF7 | tensorflow 读取多个文件 - 搜索
https://zhuanlan.zhihu.com/p/53917588 | Tensorflow数据读取-TFRecords - 知乎
https://blog.csdn.net/Smile_mingm/article/details/104552340 | (22条消息) Tensorflow2.0学习（13）：在Tf1.0中使用Dataset_一枚小白的博客-CSDN博客
https://blog.csdn.net/houyanhua1/article/details/88173403 | (22条消息) Python TensorFlow，读取文件流程，读取CSV文件案例，TextLineReader()_houyanhua1的专栏-CSDN博客
https://www.jianshu.com/p/7a634e2226aa | Tensorflow——读取csv文件作为数据集 - 简书
https://blog.csdn.net/u014311125/article/details/121290743 | (22条消息) Tensorflow的数据处理中的Dataset和Iterator_zhaoguanhua的博客-CSDN博客
https://www.cnblogs.com/huangyc/p/10339433.html#_label2 | 2. Tensorflow的数据处理中的Dataset和Iterator - hyc339408769 - 博客园
https://www.cnblogs.com/alex0521/p/11581786.html | Cifar-10数据集免费下载及Tensorflow代码实现 - 站在巴黎铁塔看东京樱花 - 博客园
https://cn.bing.com/search?q=tf.TextLineReader&cvid=0e01519e4cf2481482d0ddbe09b38e2a&aqs=edge..69i57.1097j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | tf.TextLineReader - 搜索
https://www.w3cschool.cn/tensorflow_python/tensorflow_python-2s342nbz.html | TensorFlow函数：tf.TextLineReader_w3cschool
https://cn.bing.com/search?q=tensorflow2+++%e8%87%aa%e5%ae%9a%e4%b9%89+dataset&qs=n&sp=-1&pq=tensorflow2+%e8%87%aa%e5%ae%9a%e4%b9%89+dataset&sc=0-23&sk=&cvid=2F01DD89E2B04FF09365C2ED27AC6A04&first=10&FORM=PORE | tensorflow2 自定义 dataset - 搜索
https://www.cnblogs.com/panda-blog/p/12310110.html | tensorflow学习笔记–dataset使用，创建自己的数据集 - panday - 博客园
https://cloud.tencent.com/developer/article/1559308 | TensorFlow2.0（10）：加载自定义图片数据集到Dataset - 云+社区 - 腾讯云
https://www.cnblogs.com/chenhuabin/p/11669452.html | TensorFlow2.0（6）：数据预处理中的Dataset - 奥辰 - 博客园
https://blog.csdn.net/weixin_45875105/article/details/114596386 | (22条消息) Tensorflow2.0 keras 子类式多输入多输出_廷益_飞鸟的博客-CSDN博客
https://zhuanlan.zhihu.com/p/133508396 | 【示例】tensorflow2.0加载自己的图片数据集到Dataset - 知乎
https://www.jarvis73.com/2020/11/25/Tensorflow-Datasets/ | Tensorflow 自定义数据集 (Tensorflow-Datasets Customization, TFDS)
https://blog.csdn.net/Two_Bits/article/details/109529292 | (22条消息) tensorFlow：加载自定义本地图片作为数据集训练GAN_记录本身，即已是反抗-CSDN博客
https://blog.csdn.net/qq_34914551/article/details/86286184 | (22条消息) TensorFlow中读取图像数据的三种方式_我的博客有点东西-CSDN博客
https://www.cnblogs.com/hp-lake/p/13174181.html | (oﾟvﾟ)ノ Hi - Tensorflow2 自定义数据集图片完成图片分类任务
https://www.cnblogs.com/chenhuabin/p/11863889.html | TensorFlow2.0（10）：加载自定义图片数据集到Dataset - 奥辰 - 博客园
https://blog.csdn.net/qq_41973343/article/details/103213366 | (22条消息) Tensorflow2.0学习（八） — tf.dataset自定义图像数据集_Yohann的博客-CSDN博客
https://blog.csdn.net/qq_41973343?type=blog | (22条消息) Yohann的博客_Yohann、_CSDN博客-人工智能,深度学习,Tensorflow2.0领域博主
https://blog.csdn.net/qq_41973343/article/details/103223377?spm=1001.2014.3001.5502 | (22条消息) Tensorflow2.0学习（九） — 通过keras自定义图像数据集_Yohann的博客-CSDN博客
https://zhuanlan.zhihu.com/p/94344717 | 【Tensorflow 2.0 正式版教程】tf.data.Dataset的基本使用方法 - 知乎
https://cxymm.net/article/mao_xiao_feng/75287760 | 【Tensorflow】超大规模数据集解决方案：通过线程来预取（下）_xf__mao的博客-程序员秘密_tensorflow读取大规模图像集 - 程序员秘密
https://cn.bing.com/search?q=catboost&cvid=fcc63dc5498445d1998a98236a2706de&aqs=edge..69i57j0l8.1683j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | catboost - 搜索
https://zhuanlan.zhihu.com/p/102570430 | catboost完全指南 - 知乎
http://challenge.ai.iqiyi.com/detail?raceId=61600f6cef1b65639cd5eaa6 | 用户留存预测挑战赛
https://www.datafountain.cn/competitions/542 | 图书推荐系统竞赛 - DataFountain
https://www.datafountain.cn/competitions/542/datasets | 图书推荐系统竞赛 - DataFountain

https://zhuanlan.zhihu.com/p/68717029 | Pytorch多机多卡分布式训练 - 知乎
https://blog.csdn.net/qq_31768873/article/details/115294744 | (33条消息) PyTorch训练加速技巧_好名字可以让朋友..更容易记住你的博客-CSDN博客
https://zhuanlan.zhihu.com/p/158375055 | PyTorch 21.单机多卡操作(分布式DataParallel，混合精度，Horovod) - 知乎
https://heary.cn/posts/DistributedDataParallel-DDP-PyTorch%E5%A4%9A%E8%BF%9B%E7%A8%8B%E5%B9%B6%E8%A1%8C%E8%AE%A1%E7%AE%97/ | DistributedDataParallel(DDP) - PyTorch多进程并行计算 | Heary’s Blog
https://www.cnblogs.com/douzujun/p/13426548.html | pytorch利用多个GPU并行计算 - douzujun - 博客园
https://github.com/apachecn/home | apachecn/home: ApacheCN 开源组织：公告、介绍、成员、活动、交流方式
https://blog.csdn.net/weixin_41990278/article/details/116480859 | (33条消息) 【Pytorch】单机多卡数据并行DataParallel_小苏打的学习博客-CSDN博客
https://bbs.cvmart.net/articles/1424 | 了解 Pytorch 分布式训练，这一篇足够了! - 极市社区
https://zhuanlan.zhihu.com/p/250471767 | [原创][深度][PyTorch] DDP系列第三篇：实战与技巧 - 知乎
https://github.com/pytorch/pytorch/blob/release/1.5/torch/nn/modules/_functions.py#L5 | pytorch/_functions.py at release/1.5 · pytorch/pytorch
https://zhuanlan.zhihu.com/p/419833524 | Pytorch多卡分布式训练|DistributedDataParaell（DDP）简要分析 - 知乎
https://zhuanlan.zhihu.com/p/98535650 | 当代研究生应当掌握的并行训练方法（单机多卡） - 知乎
https://blog.csdn.net/PanYHHH/article/details/111750248 | (33条消息) Pytorch torch.distributed 实现单机多卡分布式训练_PanYHHH的博客-CSDN博客
https://zhuanlan.zhihu.com/p/136902153 | PyTorch–lr_scheduler.step()和optimizer.step()的先后顺序 - 知乎
https://cloud.tencent.com/developer/article/1770678 | PyTorch 多机多卡训练：DDP 实战与技巧 - 云+社区 - 腾讯云
https://murphypei.github.io/blog/2020/09/pytorch-distributed | PyTorch 多进程分布式训练实战 | 拾荒志
https://blog.csdn.net/searobbers_duck/article/details/115209880 | (33条消息) pytorch分布式系列1——搞清torch.distributed.launch相关的环境变量_searobbers_duck的博客-CSDN博客
https://blog.csdn.net/weixin_41041772/article/details/109820870 | (33条消息) 通俗理解torch.distributed.barrier()工作原理_视觉弘毅的博客-CSDN博客_torch.distributed.barrier()
https://blog.csdn.net/junqing_wu/article/details/112732338#:~:text=dist.broadcast%20%28tensor%2C%20src%2C%20group%29%3A%20Copies%20tensor%20from%20src,but%20the%20result%20is%20stored%20in%20all%20processes. | (33条消息) torch 多进程训练(详细例程)_junqing_wu的博客-CSDN博客_dist.broadcast
https://cn.bing.com/search?q=os.environ%5B%27SLURM_NTASKS%27%5D+&qs=n&form=QBRE&sp=-1&pq=os.environ%5B%27slurm_ntasks%27%5D+&sc=0-27&sk=&cvid=9DC11BCFC63E44E9811EBDA2B25406D3 | os.environ[‘SLURM_NTASKS’] - 搜索
https://www.zhihu.com/question/451269794 | 请教pytorch之DDP训练slurm配置问题+代码？ - 知乎
https://www.jianshu.com/p/5f6cd6b50140 | 分布式通信包 - torch.distributed - 简书
https://blog.csdn.net/comway_Li/article/details/107531165 | (33条消息) pytorch 分布式多卡训练DistributedDataParallel 踩坑记_comway_Li的博客-CSDN博客
https://blog.csdn.net/weixin_38739735/article/details/110944658 | (33条消息) 使用 PyTorch 进行分布式训练_Warmer_Sweeter-CSDN博客
https://blog.csdn.net/u010734277/article/details/108324900 | (33条消息) pytorch 分布式多卡_char256-CSDN博客
https://zhuanlan.zhihu.com/p/76638962 | Pytorch 分布式训练 - 知乎
https://www.w3cschool.cn/article/48798893.html | PyTorch 如何设置随机数种子使结果可复现 | w3c笔记
https://blog.csdn.net/weixin_42587961/article/details/109363698 | (33条消息) torch.backends.cudnn.benchmark和cudnn.deterministic_windistance的博客-CSDN博客
https://zhuanlan.zhihu.com/p/141063432 | torch.backends.cudnn.deterministic - 知乎
https://www.jianshu.com/p/be9f8b90a1b8 | 使用PyTorch编写分布式应用程序 - 简书
https://www.jianshu.com/p/5f6cd6b50140 | 分布式通信包 - torch.distributed - 简书
https://blog.csdn.net/love1005lin/article/details/116404049 | (33条消息) 分布式训练 - 单机多卡（DP和DDP）_love1005lin的博客-CSDN博客
https://blog.csdn.net/lzkzls/article/details/107014702 | (19条消息) torch.nn.parallel.DistributedDataParallel 使用笔记_lzkzls的博客-CSDN博客
https://blog.csdn.net/baidu_35120637/article/details/110816619?utm_medium=distribute.pc_relevant.none-task-blog-2~default~baidujs_title~default-1.no_search_link&spm=1001.2101.3001.4242.2 | (22条消息) Pytorch之torch.nn.parallel.DistributedDataParallel_潜心-CSDN博客
https://blog.csdn.net/cdknight_happy/article/details/109366599 | (22条消息) SyncBN及其pyTorch实现_cdknight_happy的专栏-CSDN博客

https://cn.bing.com/search?q=lgboost&cvid=99dd9d56ac504a75aefc1e85287c5418&aqs=edge..69i57.11816j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | lgboost - 搜索
https://cloud.tencent.com/developer/article/1534903 | 终于有人把XGBoost 和 LightGBM 讲明白了，项目中最主流的集成算法！ - 云+社区 - 腾讯云
https://github.com/stanfordmlgroup/ngboost | stanfordmlgroup/ngboost: Natural Gradient Boosting for Probabilistic Prediction
https://zhuanlan.zhihu.com/p/91395504 | 决策树-GBDT, Xgboost，LGBM - 知乎
https://cn.bing.com/search?q=xlearn&form=ANNTH1&refig=b8d17bccdb3440eda217e96962cd4e68 | xlearn - 搜索
https://blog.csdn.net/linxid/article/details/80382569 | (33条消息) xlearn安装与实战_林夕-CSDN博客
https://cn.bing.com/search?q=notion+%e5%88%b6%e4%bd%9c%e6%a8%a1%e6%9d%bf&qs=n&sp=-1&pq=notion+%e5%88%b6%e4%bd%9c%e6%a8%a1%e6%9d%bf&sc=0-11&sk=&cvid=B2FA56F626954E5E9391A0D9A782D691&first=11&FORM=PORE | notion 制作模板 - 搜索
https://www.vantopics.com/notion-%e6%a8%a1%e6%9d%bf-%e5%a6%82%e4%bd%95%e5%89%b5%e5%bb%ba%e5%92%8c%e4%bf%ae%e6%94%b9-template-%ef%bc%9f/ | Notion 模板 | 如何創建和修改 template ？ - VanTopics 大溫話題
https://zhuanlan.zhihu.com/p/103498242 | Notion使用指南 | Database篇 - 知乎
https://blog.csdn.net/u014162133/article/details/109724200 | (33条消息) Notion使用详解一基础教程_相见不如怀念-CSDN博客_notion使用教程
https://zhuanlan.zhihu.com/p/148510374 | Notion入门&技巧 - 知乎
https://linmi.cc/1713.html | 如何快速上手 Notion | Notion 入门教程 - Linmi
https://zhuanlan.zhihu.com/p/156792848 | Notion简明教程 - 知乎
https://zhuanlan.zhihu.com/p/395123222 | Notion官方中文入门大全+可复制个人系统模板 - 知乎
https://www.notion.so/Copy-of-ff27c383193d4023b9099e2ae9c70d88 | Copy of 康奈尔笔记
https://www.ifanr.com/app/1168845 | 高效工作必备，手把手教你玩转 Notion 模板（附下载） | 爱范儿
https://www.douban.com/group/topic/132217048/ | 分享一个很赞的 notion 模板网站
https://www.notion.so/Cornell-Notes-System-895f335ed949406cba6e6a5ddca26a69 | Cornell Notes System
https://sspai.com/post/54148 | 用好这些实用模板，把 Notion 打造成全能助理 - 少数派
https://www.notion.so/Cornell-Notes-System-895f335ed949406cba6e6a5ddca26a69 | Cornell Notes System
https://www.notion.so/Notion-Template-Gallery-181e961aeb5c4ee6915307c0dfd5156d#98c8862810454bf694d9fb611ba26edc | Notion Template Gallery
https://www.notion.so/Copy-of-ff27c383193d4023b9099e2ae9c70d88?showMoveTo=true | Copy of 康奈尔笔记
https://cn.bing.com/search?q=worktile&cvid=b3b9d52b3925439892ed5ed3664e044d&aqs=edge.0.0l9.3233j0j4&FORM=ANAB01&PC=HCTS | worktile - 搜索
https://cn.bing.com/search?q=worktile+%E9%A1%B9%E7%9B%AE%E5%9B%9E%E6%94%B6%E7%AB%99+%E6%B8%85%E7%A9%BA&qs=n&form=QBRE&sp=-1&pq=worktile+%E9%A1%B9%E7%9B%AE%E5%9B%9E%E6%94%B6%E7%AB%99+%E6%B8%85%E7%A9%BA&sc=0-17&sk=&cvid=3F7F3D1F298E4A21BD9143F007EA7572 | worktile 项目回收站清空 - 搜索
https://zhuanlan.zhihu.com/p/59101453 | 做好项目管理、任务追踪、日程管理，Worktile成公司内部管理工作的重要载体！ - 知乎
https://blog.csdn.net/wuzhongchao/article/details/42435189 | (33条消息) 我所认识的Worktile—Worktile图解_wuzhongchao的专栏-CSDN博客
https://trello.com/hu1104sworkspace | hu1104’s Workspace (hu1104sworkspace) | Trello
https://www.jianshu.com/p/d9ca818cf732 | Trello使用：看板、列表和卡片的归档及删除方法 - 简书
https://www.zhihu.com/question/24920474 | Worktile 与 Teambition 的优劣对比？ - 知乎
https://blog.csdn.net/jackzhang11/article/details/105770223 | (33条消息) Pytorch中的神坑：关于model.eval的问题_jackzhang11的博客-CSDN博客
https://blog.csdn.net/yanxiangtianji/article/details/113798888 | (33条消息) PyTorch里eval和no_grad的关系_yanxiangtianji的专栏-CSDN博客
https://blog.csdn.net/star_xxzeng/article/details/108351557 | (33条消息) model.eval()与with torch.no_grad()_star_xxzeng的博客-CSDN博客
https://www.w3cschool.cn/article/36979814.html | pytorch测试的时候为何要加上model.eval()？ | w3c笔记
https://blog.csdn.net/confusingbird/article/details/108799908 | (33条消息) pytorch预测之多次预测结果不一致问题_confusingbird的博客-CSDN博客
https://www.cnpython.com/qa/359315 | 为什么sklearn在CPU上比在GPU上更快？ - 问答 - Python中文网
https://blog.csdn.net/weixin_39574943/article/details/111843465 | (33条消息) sklearn支持gpu_常见问题-scikit-learn (sklearn) 官方文档中文版_weixin_39574943的博客-CSDN博客
https://www.jianshu.com/p/0fe45d4e9542 | GBDT、XGBoost、LightGBM 的使用及参数调优 - 简书

https://cn.bing.com/search?q=pytorch+%e6%a8%a1%e5%9e%8b%e5%90%84%e5%b1%82+%e5%91%bd%e5%90%8d&qs=n&sp=-1&pq=pytorch+%e6%a8%a1%e5%9e%8b%e5%90%84%e5%b1%82+%e5%91%bd%e5%90%8d&sc=0-15&sk=&cvid=A99265B30150406C95CC64CD8E95C1EF&first=11&FORM=PORE | pytorch 模型各层命名 - 搜索
https://www.cnblogs.com/denny402/p/7593301.html | pytorch学习：构建网络模型的几种方法 - denny402 - 博客园
https://zhuanlan.zhihu.com/p/342492242 | pytorch model层各种骚操作 - 知乎
https://blog.csdn.net/u014261408/article/details/120996966 | (32条消息) Pytorch查看，输出，打印模型各个网络层的名字和参数_亲亲Friends的博客-CSDN博客
https://blog.csdn.net/u014734886/article/details/106230535 | (32条消息) pytorch中存储各层权重参数时的命名规则，为什么有些层的名字中带module._南国那片枫叶的博客-CSDN博客_dist.get_rank()
https://www.jianshu.com/p/55927007b98a | Pytorch冻结网络部分参数 - 简书
https://blog.csdn.net/qq_21997625/article/details/90369838 | (32条消息) pytorch冻结部分参数训练另一部分_别说话写代码的博客-CSDN博客_pytorch冻结参数
https://blog.csdn.net/jningwei/article/details/120300014 | (32条消息) 【pytorch】固定(freeze)住部分网络_JNing-CSDN博客
https://zhuanlan.zhihu.com/p/68748778 | 【炼丹技巧】指数移动平均（EMA）的原理及PyTorch实现 - 知乎
https://www.cnblogs.com/wzyuan/p/9880342.html | 解读 pytorch对resnet的官方实现 - wuzeyuan - 博客园

https://www.sdk.cn/details/EN5yDkYQWj9mkx1RVY | PyTorch分布式DPP涉及的基本概念与问题 - sdk社区 | 技术至上
https://murphypei.github.io/blog/2020/09/pytorch-distributed | PyTorch 多进程分布式训练实战 | 拾荒志
https://blog.csdn.net/qq_34914551/article/details/110576421 | (31条消息) PyTorch多进程分布式训练最简单最好用的实施办法_我的博客有点东西-CSDN博客
https://www.cnblogs.com/yh-blog/p/12877922.html#autoid-2-0-0 | pytorch(分布式)数据并行个人实践总结——DataParallel/DistributedDataParallel - fnangle - 博客园
https://www.telesens.co/2019/02/16/efficient-data-transfer-from-paged-memory-to-gpu-using-multi-threading/ | Pipelining data processing and host-to-device data transfer – Telesens
https://github.com/fnangle/text_classfication-with-bert-pytorch/blob/master/textCNN_parallel.py | text_classfication-with-bert-pytorch/textCNN_parallel.py at master · fnangle/text_classfication-with-bert-pytorch
https://blog.csdn.net/github_37320188/article/details/100519346 | (31条消息) Deep Learning:PyTorch 基于docker 容器的分布式训练实践_lo_ong的博客-CSDN博客
https://www.cnblogs.com/wzyuan/p/9880342.html | 解读 pytorch对resnet的官方实现 - wuzeyuan - 博客园
https://blog.csdn.net/hxxjxw/article/details/119606518 | (31条消息) PyTorch分布式DPP的基本概念(node&rank&local_rank&nnodes&node_rank&nproc_per_node&world_size)_hxxjxw的博客-CSDN博客
https://blog.csdn.net/u013700358/article/details/93135254?spm=1001.2101.3001.6650.10&utm_medium=distribute.pc_relevant.none-task-blog-2%7Edefault%7ECTRLIST%7Edefault-10.no_search_link&depth_1-utm_source=distribute.pc_relevant.none-task-blog-2%7Edefault%7ECTRLIST%7Edefault-10.no_search_link | (31条消息) Pytorch多机多卡训练_MaloFleur-CSDN博客
https://blog.csdn.net/junqing_wu/article/details/112732338#:~:text=dist.broadcast%20%28tensor%2C%20src%2C%20group%29%3A%20Copies%20tensor%20from%20src,but%20the%20result%20is%20stored%20in%20all%20processes. | (31条消息) torch 多进程训练(详细例程)_junqing_wu的博客-CSDN博客_torch 多进程
https://blog.csdn.net/weixin_38739735/article/details/110944658 | (31条消息) 使用 PyTorch 进行分布式训练_Warmer_Sweeter-CSDN博客
https://blog.csdn.net/weixin_41041772/article/details/109820870 | (31条消息) 通俗理解torch.distributed.barrier()工作原理_视觉弘毅的博客-CSDN博客_torch.distributed.barrier()
https://zhuanlan.zhihu.com/p/68748778 | 【炼丹技巧】指数移动平均（EMA）的原理及PyTorch实现 - 知乎
https://heary.cn/posts/DistributedDataParallel-DDP-PyTorch%E5%A4%9A%E8%BF%9B%E7%A8%8B%E5%B9%B6%E8%A1%8C%E8%AE%A1%E7%AE%97/ | DistributedDataParallel(DDP) - PyTorch多进程并行计算 | Heary’s Blog
https://zhuanlan.zhihu.com/p/178402798 | [原创][深度][PyTorch] DDP系列第一篇：入门教程 - 知乎
https://cn.bing.com/search?q=pytorch+%E5%8D%95%E6%9C%BA%E5%A4%9A%E5%8D%A1&qs=n&form=QBRE&sp=-1&pq=pytorch+%E5%8D%95%E6%9C%BA%E5%A4%9A%E5%8D%A1&sc=0-12&sk=&cvid=7F897332D2F24D82B071A944F8E33491 | pytorch 单机多卡 - 搜索
https://blog.csdn.net/sinat_22510827/article/details/121183868 | (32条消息) PyTorch 单机多卡操作总结：分布式DataParallel，混合精度，Horovod)_sinat_22510827的博客-CSDN博客
https://blog.csdn.net/weixin_38424903/article/details/107625821 | (32条消息) 关于PyTorch单机多卡训练_Karl_G的博客-CSDN博客_pytorch单机多卡训练
https://zhuanlan.zhihu.com/p/68717029 | Pytorch多机多卡分布式训练 - 知乎
https://blog.csdn.net/weixin_38424903/article/details/107625821 | (31条消息) 关于PyTorch单机多卡训练_Karl_G的博客-CSDN博客_pytorch单机多卡训练
https://www.cnblogs.com/douzujun/p/13426548.html | pytorch利用多个GPU并行计算 - douzujun - 博客园
https://blog.csdn.net/weixin_41990278/article/details/116480859 | (31条消息) 【Pytorch】单机多卡数据并行DataParallel_小苏打的学习博客-CSDN博客
https://zhuanlan.zhihu.com/p/158375254 | PyTorch 20.GPU训练 - 知乎
https://blog.csdn.net/qazwsxrx/article/details/89672578 | (31条消息) Pytorch 多块GPU的使用_NLOS的博客-CSDN博客_torch 多gpu
https://zhuanlan.zhihu.com/p/267157806 | “最全PyTorch分布式教程”来了！ - 知乎
https://bbs.cvmart.net/articles/1424 | 了解 Pytorch 分布式训练，这一篇足够了! - 极市社区
https://zhuanlan.zhihu.com/p/206467852 | DataParallel & DistributedDataParallel分布式训练 - 知乎
https://cloud.tencent.com/developer/article/1770678 | PyTorch 多机多卡训练：DDP 实战与技巧 - 云+社区 - 腾讯云
https://zhuanlan.zhihu.com/p/76638962#:~:text=Pytorch%20%E4%B8%AD%E9%80%9A%E8%BF%87%20torch.distributed%20%E5%8C%85%E6%8F%90%E4%BE%9B%E5%88%86%E5%B8%83%E5%BC%8F%E6%94%AF%E6%8C%81%EF%BC%8C%E5%8C%85%E6%8B%AC%20GPU%20%E5%92%8C%20CPU%20%E7%9A%84%E5%88%86%E5%B8%83%E5%BC%8F%E8%AE%AD%E7%BB%83%E6%94%AF%E6%8C%81%E3%80%82,%E5%88%86%E5%B8%83%E5%BC%8F%E7%9B%AE%E5%89%8D%E5%8F%AA%E6%94%AF%E6%8C%81%20Linux%20%E3%80%82%20%E5%9C%A8%E6%AD%A4%E4%B9%8B%E5%89%8D%EF%BC%8C%20torch.nn.DataParallel%20%E5%B7%B2%E7%BB%8F%E6%8F%90%E4%BE%9B%E6%95%B0%E6%8D%AE%E5%B9%B6%E8%A1%8C%E7%9A%84%E6%94%AF%E6%8C%81%EF%BC%8C%E4%BD%86%E6%98%AF%E5%85%B6%E4%B8%8D%E6%94%AF%E6%8C%81%E5%A4%9A%E6%9C%BA%E5%88%86%E5%B8%83%E5%BC%8F%E8%AE%AD%E7%BB%83%EF%BC%8C%E4%B8%94%E5%BA%95%E5%B1%82%E5%AE%9E%E7%8E%B0%E7%9B%B8%E8%BE%83%E4%BA%8E%20distributed%20%E7%9A%84%E6%8E%A5%E5%8F%A3%EF%BC%8C%E6%9C%89%E4%BA%9B%E8%AE%B8%E4%B8%8D%E8%B6%B3%E3%80%82 | Pytorch 分布式训练 - 知乎
https://zhuanlan.zhihu.com/p/68717029 | Pytorch多机多卡分布式训练 - 知乎
https://blog.csdn.net/comway_Li/article/details/107531165 | (31条消息) pytorch 分布式多卡训练DistributedDataParallel 踩坑记_comway_Li的博客-CSDN博客
http://nicethemes.cn/news/txtlist_i195148v.html | Pytorch 分布式模式介绍
https://blog.csdn.net/PanYHHH/article/details/111750248 | (31条消息) Pytorch torch.distributed 实现单机多卡分布式训练_PanYHHH的博客-CSDN博客
https://zhuanlan.zhihu.com/p/68717029 | Pytorch多机多卡分布式训练 - 知乎
https://cn.bing.com/search?q=torch+dist+group%E5%90%AB%E4%B9%89&qs=n&form=QBRE&sp=-1&pq=torch+dist+group%E5%90%AB%E4%B9%89&sc=0-18&sk=&cvid=CB9A327E12154BDBA92FF9254E4700A1 | torch dist group含义 - 搜索
https://blog.csdn.net/lzkzls/article/details/107014702 | (31条消息) torch.nn.parallel.DistributedDataParallel 使用笔记_lzkzls的博客-CSDN博客
https://www.jianshu.com/p/5f6cd6b50140 | 分布式通信包 - torch.distributed - 简书
https://vimsky.com/examples/detail/python-method-torch.distributed.new_group.html | Python distributed.new_group方法代码示例 - 纯净天空
https://blog.csdn.net/qq_34243930/article/details/106695877 | (31条消息) pytorch之多GPU使用——#CUDA_VISIBLE_DEVICES使用 #torch.nn.DataParallel() #报错解决_夏普通-CSDN博客
https://blog.csdn.net/weixin_43402775/article/details/109223794 | (31条消息) pytorh .to(device) 和.cuda()的区别_Golden-sun的博客-CSDN博客_to(device)和cuda()
https://www.cnblogs.com/tingtin/p/13544565.html | pytorch 读取和保存模型参数 - cltt - 博客园
https://blog.csdn.net/jokerxsy/article/details/108745606 | (31条消息) torch保存、读取模型_jokerxsy的博客-CSDN博客_torch 读取模型
https://blog.csdn.net/Code_Mart/article/details/88254444 | (31条消息) [日常] PyTorch 预训练模型，保存，读取和更新模型参数以及多 GPU 训练模型_不进则退-CSDN博客
https://blog.csdn.net/qq_45128278/article/details/114882460 | (31条消息) pytorch保存和读取模型，定义save_checkpoint函数_乱搭巴士的博客-CSDN博客
https://zhuanlan.zhihu.com/p/136902153 | PyTorch–lr_scheduler.step()和optimizer.step()的先后顺序 - 知乎
https://blog.csdn.net/Carlsummer/article/details/119749770 | (31条消息) pytorch scheduler汇总_曾小辉的博客-CSDN博客
https://zhuanlan.zhihu.com/p/87209990 | 【PyTorch】优化器 torch.optim.Optimizer - 知乎
https://blog.csdn.net/weixin_39228381/article/details/108548413 | (31条消息) pytorch优化器详解：Adam_拿铁大侠的博客-CSDN博客
https://cn.bing.com/search?q=pytorch+%e5%a4%9a%e6%9c%ba%e5%a4%9a%e5%8d%a1+%e4%bb%a3%e7%a0%81&qs=n&sp=-1&pq=pytorch+%e5%a4%9a%e6%9c%ba%e5%a4%9a%e5%8d%a1+dai%27m&sc=0-18&sk=&cvid=F18CB9E9296F4856922D357D82456358&first=21&FORM=PORE | pytorch 多机多卡代码 - 搜索
https://github.com/hu1104/distributed_train_pytorch/blob/main/train.sh | distributed_train_pytorch/train.sh at main · hu1104/distributed_train_pytorch
https://zhuanlan.zhihu.com/p/373395654 | Pytorch多机多卡分布式训练 - 知乎
https://blog.csdn.net/love1005lin/article/details/116456422 | (32条消息) 分布式训练 - 多机多卡 (DDP)_love1005lin的博客-CSDN博客
https://www.zhihu.com/question/453920336 | pytorch分布式多机多卡训练，希望从例子解释，以下代码中参数是什么意思？ - 知乎
https://blog.csdn.net/zandaoguang/article/details/111714014 | (32条消息) PyTorch 多机多卡训练：DDP 实战与技巧_zandaoguang的博客-CSDN博客
https://zhuanlan.zhihu.com/p/373395654 | Pytorch多机多卡分布式训练 - 知乎
https://blog.csdn.net/searobbers_duck/article/details/115209880 | (32条消息) pytorch分布式系列1——搞清torch.distributed.launch相关的环境变量_searobbers_duck的博客-CSDN博客
https://blog.csdn.net/c9Yv2cf9I06K2A9E/article/details/121154421 | (31条消息) PyTorch多卡分布式训练：DistributedDataParallel (DDP) 简要分析_Paper weekly-CSDN博客
https://zhuanlan.zhihu.com/p/419833524 | Pytorch多卡分布式训练|DistributedDataParaell（DDP）简要分析 - 知乎
https://github.com/tczhangzhi/pytorch-distributed/blob/master/distributed.py | pytorch-distributed/distributed.py at master · tczhangzhi/pytorch-distributed
https://blog.csdn.net/weixin_38424903/article/details/107625821 | (32条消息) 关于PyTorch单机多卡训练_Karl_G的博客-CSDN博客_pytorch单机多卡训练
edge://newtab/ | 新建标签页
https://cn.bing.com/search?q=pytorch+%E5%88%86%E5%B8%83%E5%BC%8F++%E5%90%8C%E6%AD%A5+BN&qs=n&form=QBRE&sp=-1&pq=pytorch+%E5%88%86%E5%B8%83%E5%BC%8F+%E5%90%8C%E6%AD%A5+bn&sc=0-17&sk=&cvid=854696C5A29E4DC4A51FB47CAECF7333 | pytorch 分布式同步 BN - 搜索
https://blog.csdn.net/lgzlgz3102/article/details/107054314 | (32条消息) PyTorch并行训练指南：单机多卡并行、混合精度、同步BN训练_idol24的博客-CSDN博客
https://zhuanlan.zhihu.com/p/40496177 | 跨卡同步 Batch Normalization - 知乎
https://www.zhihu.com/question/421369328 | pytorch中当使用nn.DataParallel进行单机多卡分布式训练时，有必要使用同步bn吗？ - 知乎
https://zhuanlan.zhihu.com/p/353587472 | PyTorch 77.结合apex支持的分布式训练，同步bn，以及混合精度 - 知乎
https://zhuanlan.zhihu.com/p/337732517 | PyTorch 源码解读之 BN & SyncBN：BN 与多卡同步 BN 详解 - 知乎
https://zhuanlan.zhihu.com/p/98535650 | 当代研究生应当掌握的并行训练方法（单机多卡） - 知乎
https://cn.bing.com/search?q=torch.nn.SyncBatchNorm.convert_sync_batchnorm&cvid=c3224e0e504b4ec2b4e7d970908e871e&aqs=edge..69i57.3002j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | torch.nn.SyncBatchNorm.convert_sync_batchnorm - 搜索
https://blog.csdn.net/qq_31768873/article/details/115294744 | (32条消息) PyTorch训练加速技巧_好名字可以让朋友..更容易记住你的博客-CSDN博客
https://zhuanlan.zhihu.com/p/158375055 | PyTorch 21.单机多卡操作(分布式DataParallel，混合精度，Horovod) - 知乎
https://github.com/BIGBALLON/distribuuuu/blob/master/tutorial/snmc_dp.py | distribuuuu/snmc_dp.py at master · BIGBALLON/distribuuuu
https://github.com/tczhangzhi/pytorch-distributed/blob/master/distributed.py | pytorch-distributed/distributed.py at master · tczhangzhi/pytorch-distributed
https://zhuanlan.zhihu.com/p/113694038 | PyTorch分布式训练简明教程 - 知乎
https://cn.bing.com/search?q=.set_epoch(epoch)&cvid=4da67c676eac41f486b08dd99b74486e&aqs=edge..69i57.1270j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | .set_epoch(epoch) - 搜索
https://zhuanlan.zhihu.com/p/113694038 | PyTorch分布式训练简明教程 - 知乎
https://cn.bing.com/search?q=pytorch+autocast&qs=n&form=QBRE&sp=-1&pq=pytorch+autocast&sc=1-16&sk=&cvid=9D81383A86594FAD9E6676E2CA5051B7 | pytorch autocast - 搜索
https://pytorch.org/docs/1.8.1/notes/amp_examples.html | Automatic Mixed Precision examples — PyTorch 1.8.1 documentation
https://blog.csdn.net/qq_38253797/article/details/116210911 | (32条消息) 【Trick2】【PyTorch】torch.cuda.amp自动混合精度训练 —— 节省显存并加快推理速度_满船清梦压星河HK的博客-CSDN博客
https://www.cnblogs.com/jimchen1218/p/14315008.html | Pytorch自动混合精度(AMP)介绍与使用 - jimchen1218 - 博客园

https://blog.csdn.net/GrayOnDream/article/details/102955297 | (29条消息) 23种激活函数_梦的灰色边沿…-CSDN博客_silu激活函数
https://jishuin.proginn.com/p/763bfbd5c12f | 一文看尽深度学习中的15种损失函数-技术圈
https://cn.bing.com/search?q=%E5%B8%B8%E7%94%A8%E6%8D%9F%E5%A4%B1%E5%87%BD%E6%95%B0+%E4%BA%BA%E8%84%B8&qs=n&form=QBRE&sp=-1&pq=%E5%B8%B8%E7%94%A8%E6%8D%9F%E5%A4%B1%E5%87%BD%E6%95%B0+%E4%BA%BA%E8%84%B8&sc=0-9&sk=&cvid=DEE5F2590DEC4360856A3C789497AA2D | 常用损失函数人脸 - 国内版 Bing
https://zhuanlan.zhihu.com/p/59440497 | 人脸识别：损失函数总结 - 知乎
https://blog.csdn.net/neil3611244?type=blog | (29条消息) 爱做菜的炼丹师_思绪零乱成海_CSDN博客-Tensorflow,无人驾驶,深度学习领域博主
https://zhuanlan.zhihu.com/p/171501111 | 人脸识别系统之损失函数篇 - 知乎
https://blog.csdn.net/u012505617/article/details/89355690 | (29条消息) 浅谈人脸识别中的loss 损失函数_liguiyuan的博客-CSDN博客_人脸识别损失函数
https://blog.csdn.net/AugustMe/article/details/110560863 | (29条消息) 【CV】常用的损失函数_学无止境、积少成多、厚积薄发-CSDN博客
https://zhuanlan.zhihu.com/p/58883095 | 常见的损失函数(loss function)总结 - 知乎
https://blog.csdn.net/junruitian/article/details/105689571 | (29条消息) 常见损失函数和评价指标总结（附公式&代码）_junruit的博客-CSDN博客
https://www.cnblogs.com/shine-lee/p/13435621.html | 人脸识别损失函数疏理与分析 - shine-lee - 博客园
https://zhuanlan.zhihu.com/p/109992475 | 新手入门 Kaggle NLP类比赛总结 - 知乎
https://zhuanlan.zhihu.com/p/266967590 | 必须收藏：20个开发技巧教你开发高性能计算代码 - 知乎
https://zhuanlan.zhihu.com/p/24495603 | Python优化第一步: 性能分析实践 - 知乎
https://www.jianshu.com/p/2e2b8701ae5c | Python性能优化-Profile - 简书
http://c.biancheng.net/view/2415.html | MySQL删除数据库（DROP DATABASE语句）
https://www.begtut.com/mysql/mysql-case-statement.html | MySQL CASE | 新手教程
https://zhuanlan.zhihu.com/p/362111660 | MySQL高级——触发器 - 知乎
https://blog.csdn.net/liudian_cz/article/details/88850773 | (30条消息) 使用docker部署flask项目_liudian_cz的博客-CSDN博客
https://www.runoob.com/python3/python3-mysql.html | Python3 MySQL 数据库连接 – PyMySQL 驱动 | 菜鸟教程
https://zhuanlan.zhihu.com/p/139763027 | Python进阶知识全篇-MySQL（PyMySQL） - 知乎
https://zhuanlan.zhihu.com/p/112050716 | Python三神器之virtualenv、virtualenvwrapper - 知乎
https://www.zhihu.com/people/na-ke-xin-zhu-na-duan-qing/posts?page=1 | 那颗心丶那段情 - 知乎
https://zhuanlan.zhihu.com/p/127175514 | python sqlalchemy基础讲解 - 知乎
https://www.jianshu.com/p/ccf098024eee | Ubuntu安装virtualenvwrapper - 简书
https://blog.csdn.net/haeasringnar/article/details/82079943 | (30条消息) Ubuntu安装anaconda 介绍、安装、配置_haeasringnar的博客-CSDN博客_ubuntu安装anaconda
https://zhuanlan.zhihu.com/p/60647332 | 最全的Python虚拟环境使用方法 - 知乎
https://blog.csdn.net/u011250186/article/details/111181765 | (30条消息) 如何在virtualenv环境中安装指定的python版本_u011250186的博客-CSDN博客
https://blog.csdn.net/weixin_43174717/article/details/106411704 | (30条消息) 创建python虚拟环境vitrualenv & pip install使用国内镜像_正在上路的猎兔犬的博客-CSDN博客
https://www.cnblogs.com/qmfsun/p/4838032.html | MySQL命令执行sql文件的两种方法 - Agoly - 博客园
https://blog.csdn.net/original_recipe/article/details/91958663 | (30条消息) Mysql经典练习题50题_original_recipe的博客-CSDN博客_mysql练习题
https://zhuanlan.zhihu.com/p/407751334 | TPH-YOLOv5：基于Transformer的改进YOLOv5的无人机目标检测 - 知乎
https://cn.bing.com/search?q=VisDrone+%E6%8C%91%E6%88%98%E8%B5%9B+2021&qs=n&form=QBRE&sp=-1&pq=visdrone+%E6%8C%91%E6%88%98%E8%B5%9B+2021&sc=0-17&sk=&cvid=AEB9B25B8F9048EC8CEED57E1BC279D5 | VisDrone 挑战赛 2021 - 搜索
http://aiskyeye.com/download/multi-object-tracking_2021/ | Multi-Object Tracking – VisDrone
https://github.com/hu1104/YOLOv5-Multibackbone-Compression | hu1104/YOLOv5-Multibackbone-Compression: YOLOv5 Series Multi-backbone(TPH-YOLOv5, Ghostnet, ShuffleNetv2,Mobilenetv3Small), Pruning and Quantization(8bit) Compression Tool Box.
https://cn.bing.com/search?q=ubuntu+%e6%98%be%e7%a4%ba%e4%b8%ad%e6%96%87+%e7%bb%88%e7%ab%af&qs=AS&pq=ubuntu+%e6%98%be%e7%a4%ba%e4%b8%ad%e6%96%87&sk=AS3&sc=4-11&cvid=E2A3408803974819A1946F721A22C3B2&FORM=QBRE&sp=4 | ubuntu 显示中文终端 - 搜索
https://blog.csdn.net/BobYuan888/article/details/88662779 | (31条消息) Ubuntu修改终端下的语言（中文或英文）_随波一落叶-CSDN博客_ubuntu命令行修改语言中文
https://blog.csdn.net/vivian_ll/article/details/80417661 | (31条消息) ubuntu server显示并输入中文_vivian_ll的博客-CSDN博客
https://blog.csdn.net/defrag257/article/details/104029924 | (31条消息) Ubuntu/WSL显示中文的方法_defrag257的博客-CSDN博客_wsl 中文
https://www.jianshu.com/p/efb23838264d | DBNet 算法阅读笔记 - 简书
https://zhuanlan.zhihu.com/p/94677957 | DBNet阅读笔记 - 知乎
https://cn.bing.com/search?q=select+from+%E4%B8%A4%E5%BC%A0%E8%A1%A8&qs=n&form=QBRE&sp=-1&pq=select+from+%E4%B8%A4%E5%BC%A0biao&sc=0-18&sk=&cvid=F083937EC0DD49D2A58D89B57A8CBA7A | select from 两张表 - 搜索
https://blog.csdn.net/zhangyj_315/article/details/2249209 | (31条消息) select * from 后有多个表的使用方法_张亚京-CSDN博客
https://cn.bing.com/search?q=mysql+exists&qs=AS&pq=mysql+ex&sk=AS1&sc=8-8&cvid=5ACDD17B0D354A22A9114DBC57A36805&FORM=QBRE&sp=2 | mysql exists - 搜索
https://www.cnblogs.com/qlqwjy/p/8598091.html | MySQL中EXISTS的用法 - QiaoZhi - 博客园
https://www.runoob.com/linux/linux-command-manual.html | Linux 命令大全 | 菜鸟教程
https://blog.csdn.net/weixin_43064185/article/details/97300443 | (31条消息) python中多个文件共用logger，重复打印问题的解决方案_Ch3nnn的博客-CSDN博客
https://www.jianshu.com/p/81a2a7fb14da | python：logging模块四大组件 - 简书
edge://newtab/ | 新建标签页

https://www.bing.com/search?q=pathlib+python&qs=n&form=QBRE&sp=-1&pq=pathlib+python&sc=0-14&sk=&cvid=9DD90F30C9F84DC8877226C02F848E5C | Search - pathlib python
https://www.jianshu.com/p/a820038e65c3 | 超好用python库(Pathlib) - 简书
https://www.cnblogs.com/heniu/p/12872604.html | python路径操作新标准：pathlib 模块 - 和牛 - 博客园
https://blog.csdn.net/u011119817/article/details/107357364 | (28条消息) Pathlib使用简单教程_dddeee的专栏-CSDN博客
https://zhuanlan.zhihu.com/p/139783331 | python路径操作新标准：pathlib 模块 - 知乎
https://zhuanlan.zhihu.com/p/87940289 | 你应该使用pathlib替代os.path - 知乎
https://blog.csdn.net/weixin_41261833/article/details/108050152 | (30条消息) Python模块——shutil模块详解_数据分析与统计学之美-CSDN博客_python shutil

https://texample.net/tikz/examples/neural-network/ | Neural network | TikZ example
https://github.com/HarisIqbal88/PlotNeuralNet/issues/24 | ResNet50 Tex · Issue #24 · HarisIqbal88/PlotNeuralNet
https://tex.stackexchange.com/questions/398805/how-to-draw-a-network-like-this-using-tikz | matrices - How to draw a network like this using tikz? - TeX - LaTeX Stack Exchange
https://www.jianshu.com/p/263ea2e7b9fd | 快速入门使用tikz绘制深度学习网络图 - 简书
https://github.com/pprp/SimpleCVReproduction/blob/master/tikz_cnn/main_modified.tex | SimpleCVReproduction/main_modified.tex at master · pprp/SimpleCVReproduction
https://www.bing.com/search?q=wandb&qs=n&form=QBRE&sp=-1&pq=wandb&sc=7-5&sk=&cvid=6A2F2ED0A82948E1B0D29353710C15C3 | Search - wandb
https://blog.csdn.net/qq_40507857/article/details/112791111 | (28条消息) wandb: 深度学习轻量级可视化工具入门教程_紫芝的博客-CSDN博客
https://zhuanlan.zhihu.com/p/266337608 | PyTorch 62.只需10分钟带你完美入门轻量级可视化工具wandb - 知乎
https://zhuanlan.zhihu.com/p/342300434 | Wandb：模型训练最强辅助 - 知乎
https://zhuanlan.zhihu.com/p/84883517 | 深度学习网络和特征图可视化的工具介绍 - 知乎
https://cloud.tencent.com/developer/article/1065135 | 7个深度神经网络可视化工具，不可错过！ - 云+社区 - 腾讯云
https://zhuanlan.zhihu.com/p/220403674 | PyTorch下的可视化工具 - 知乎
https://github.com/lorne-luo/DeepLearning | lorne-luo/DeepLearning: 深度学习入门教程&&优秀文章&&Deep Learning Tutorial

https://www.bing.com/search?q=tikz+resnet&qs=n&form=QBRE&sp=-1&pq=tikz+re%27s%27net&sc=0-13&sk=&cvid=F33C7F1F4E614698830EE86FD80D01D9 | Search - tikz resnet
https://texample.net/tikz/examples/neural-network/ | Neural network | TikZ example
https://github.com/HarisIqbal88/PlotNeuralNet/issues/24 | ResNet50 Tex · Issue #24 · HarisIqbal88/PlotNeuralNet
https://tex.stackexchange.com/questions/398805/how-to-draw-a-network-like-this-using-tikz | matrices - How to draw a network like this using tikz? - TeX - LaTeX Stack Exchange
https://www.jianshu.com/p/263ea2e7b9fd | 快速入门使用tikz绘制深度学习网络图 - 简书
https://github.com/pprp/SimpleCVReproduction/blob/master/tikz_cnn/main.tex | SimpleCVReproduction/main.tex at master · pprp/SimpleCVReproduction
https://www.bing.com/search?q=wandb&qs=n&form=QBRE&sp=-1&pq=wandb&sc=7-5&sk=&cvid=6A2F2ED0A82948E1B0D29353710C15C3 | Search - wandb
https://blog.csdn.net/qq_40507857/article/details/112791111 | (28条消息) wandb: 深度学习轻量级可视化工具入门教程_紫芝的博客-CSDN博客
https://zhuanlan.zhihu.com/p/266337608 | PyTorch 62.只需10分钟带你完美入门轻量级可视化工具wandb - 知乎
https://zhuanlan.zhihu.com/p/342300434 | Wandb：模型训练最强辅助 - 知乎
https://zhuanlan.zhihu.com/p/84883517 | 深度学习网络和特征图可视化的工具介绍 - 知乎
https://cloud.tencent.com/developer/article/1065135 | 7个深度神经网络可视化工具，不可错过！ - 云+社区 - 腾讯云
https://zhuanlan.zhihu.com/p/220403674 | PyTorch下的可视化工具 - 知乎

https://cn.bing.com/search?q=multi-task+loss&qs=n&form=QBRE&sp=-1&pq=multi-task+loss&sc=0-15&sk=&cvid=DD78A8E7DF2F4CE2B23AFDD0F0DAEE46 | multi-task loss - 国内版 Bing
https://www.zhihu.com/question/375794498 | 深度学习的多个loss如何平衡？ - 知乎
https://blog.csdn.net/laolu1573/article/details/83626555 | (27条消息) Multi-task learning 中的Loss设计_laolu1573的专栏-CSDN博客
https://blog.csdn.net/qq_34527082/article/details/100048864 | (27条消息) Multi-task中的多任务loss平衡问题_Multi-task的博客-CSDN博客
https://zhuanlan.zhihu.com/p/59413549 | Multi-task Learning(Review)多任务学习概述 - 知乎
https://cn.bing.com/search?q=python+zip%28*%29&qs=n&form=QBRE&sp=-1&pq=python+zip%28*%29&sc=0-13&sk=&cvid=61887352026F4FEC8A6C8D7FAD08E1F5 | python zip() - 国内版 Bing
https://blog.csdn.net/csdn15698845876/article/details/73411541 | (27条消息) python中zip()函数的用法_张康的博客-CSDN博客_python zip
https://www.cnblogs.com/waltsmith/p/8029539.html | Python中的zip()与zip()函数详解 - 树上的waltsmith - 博客园
https://cn.bing.com/search?q=html5&qs=AS&pq=html&sk=AS1&sc=8-4&cvid=2285FA1BA5A441379E7C419C56463FDA&FORM=QBRE&sp=2 | html5 - 国内版 Bing
https://www.runoob.com/html/html-layouts.html | HTML 布局 | 菜鸟教程
https://blog.csdn.net/u012349679/article/details/103815049 | (27条消息) Markdown 添加文献引用_Murphy-CSDN博客_markdown 参考文献
https://markdown.com.cn/intro.html | Markdown 入门基础 | Markdown 官方教程
https://sspai.com/post/43873 | 简单又好看，你的 Markdown 文稿也能加上个性化主题 - 少数派
https://cn.bing.com/search?q=markdown+min%E4%B8%8B%E6%A0%87&cvid=424116b13fce48ebb40353f5d1a9044b&aqs=edge..69i57.8941j0j4&FORM=ANAB01&PC=HCTS | markdown min下标 - 国内版 Bing
https://cn.bing.com/search?q=%5Cmathrm+markdown&cvid=9c32d9046b3447d996cb1fbb71b654d0&aqs=edge..69i57.4326j0j4&FORM=ANAB01&PC=HCTS | \mathrm markdown - 国内版 Bing
https://blog.csdn.net/huangque1234/article/details/73920666 | (27条消息) Markdown－Math语法_huangque1234的博客-CSDN博客
https://www.bing.com/search?q=markdown+%e5%85%ac%e5%bc%8f%e7%bc%96%e5%8f%b7&qs=SC&pq=markdown+%e5%85%ac%e5%bc%8f+&sc=8-12&cvid=7585DD0246E444738504B747B311118F&FORM=QBRE&sp=1 | markdown 公式编号 - 国内版 Bing
https://blog.csdn.net/KitKat09/article/details/89162216 | (27条消息) Markdown中LaTeX公式编号_KitKat09的博客-CSDN博客_markdown公式编号
https://www.zybuluo.com/fyywy520/note/82980 | Markdown下LaTeX公式、编号、对齐 - 作业部落 Cmd Markdown 编辑阅读器
https://www.cnblogs.com/1024th/p/11623258.html | LaTeX公式手册(全网最全) - 樱花赞 - 博客园
https://zhuanlan.zhihu.com/p/110756681 | LaTeX 公式篇 - 知乎
http://blog.sina.com.cn/s/blog_5e16f1770100gror.html | LaTeX技巧207：使用align环境输入多行公式的技巧_LaTeX_Fun_新浪博客

https://www.cnblogs.com/pinard/p/6140514.html | 梯度提升树(GBDT)原理小结 - 刘建平Pinard - 博客园
https://blog.csdn.net/zpalyq110/article/details/79527653 | (27条消息) GBDT算法原理以及实例理解_Freemanzxp-CSDN博客_gbdt算法
https://blog.csdn.net/program_developer/article/details/103060416 | (27条消息) 深入理解GBDT二分类算法_Microstrong-CSDN博客_gbdt分类
https://www.bing.com/search?q=GBDT+csdn&qs=n&sp=-1&pq=gbdt+csdn&sc=0-9&sk=&cvid=D7EA219C1D554154A87BBC99585A0688&first=30&FORM=PORE | GBDT csdn - 国内版 Bing
https://www.cnblogs.com/ModifyRong/p/7744987.html | 机器学习算法GBDT的面试要点总结-上篇 - ModifyBlog - 博客园
https://blog.csdn.net/qq_24519677/article/details/82020863 | (27条消息) GBDT基本原理及算法描述_Y学习使我快乐V的博客-CSDN博客_gbdt
edge://newtab/ | 新建标签页
https://cn.bing.com/search?q=latex+%E7%B4%AF%E5%8A%A0%E4%B8%8B%E6%A0%87&cvid=378c07529925456f85d2bdfbb9693e97&aqs=edge..69i57.11560j0j4&FORM=ANAB01&PC=HCTS | latex 累加下标 - 国内版 Bing
https://zhuanlan.zhihu.com/p/262973963 | 如何用LaTeX编写不同格式的求和、极限、积分式？ - 知乎
https://blog.csdn.net/beifangyu/article/details/90409924 | (27条消息) Latex 求和积分的上下标位置_beifangyu的博客-CSDN博客_latex求和符号上下标
https://cn.bing.com/search?q=l1+l2+%E6%AD%A3%E5%88%99%E5%8C%96+%E4%B8%8D%E5%90%8C%E8%A7%92%E5%BA%A6&qs=n&form=QBRE&sp=-1&pq=l1+l2+%E6%AD%A3%E5%88%99%E5%8C%96+%E4%B8%8D%E5%90%8Cjiao%27du&sc=0-19&sk=&cvid=8900FCDC90B54E13A8D079F62C3BBFAD | l1 l2 正则化不同角度 - 国内版 Bing
https://zhuanlan.zhihu.com/p/376282535 | 【面试看这篇就够了】L1、L2正则化理解 - 知乎
https://www.cnblogs.com/XDU-Lakers/p/11773552.html | L1与L2正则化的对比及多角度阐述为什么正则化可以解决过拟合问题 - 控球强迫症 - 博客园
https://zhuanlan.zhihu.com/p/376000306 | 理解L1, L2正则化的正确姿势 - 知乎
https://zhuanlan.zhihu.com/p/35356992 | L1正则化与L2正则化 - 知乎
https://zhuanlan.zhihu.com/p/29360425 | 深入理解L1、L2正则化 - 知乎
https://www.cnblogs.com/zingp/p/10375691.html#:~:text=L1%E6%AD%A3%E5%88%99%E5%8C%96%E5%92%8CL2%E6%AD%A3%E5%88%99%E5%8C%96%E7%9A%84%E8%AF%B4%E6%98%8E%E5%A6%82%E4%B8%8B%EF%BC%9A%20L1%E6%AD%A3%E5%88%99%E5%8C%96%E6%98%AF%E6%8C%87%E6%9D%83%E5%80%BC%E5%90%91%E9%87%8F%20w%20%E4%B8%AD%E5%90%84%E4%B8%AA%E5%85%83%E7%B4%A0%E7%9A%84%E7%BB%9D%E5%AF%B9%E5%80%BC%E4%B9%8B%E5%92%8C%EF%BC%8C%E9%80%9A%E5%B8%B8%E8%A1%A8%E7%A4%BA%E4%B8%BA%20%E2%80%96%20w%20%E2%80%96%201,w%20%E4%B8%AD%E5%90%84%E4%B8%AA%E5%85%83%E7%B4%A0%E7%9A%84%E5%B9%B3%E6%96%B9%E5%92%8C%E7%84%B6%E5%90%8E%E5%86%8D%E6%B1%82%E5%B9%B3%E6%96%B9%E6%A0%B9%EF%BC%88%E5%8F%AF%E4%BB%A5%E7%9C%8B%E5%88%B0Ridge%E5%9B%9E%E5%BD%92%E7%9A%84L2%E6%AD%A3%E5%88%99%E5%8C%96%E9%A1%B9%E6%9C%89%E5%B9%B3%E6%96%B9%E7%AC%A6%E5%8F%B7%EF%BC%89%EF%BC%8C%E9%80%9A%E5%B8%B8%E8%A1%A8%E7%A4%BA%E4%B8%BA%20%E2%80%96%20w%20%E2%80%96%202%202%20%E3%80%82 | 深入理解L1、L2正则化 - ZingpLiu - 博客园
https://www.miracledave.com/ | 云流剑心
https://cn.bing.com/search?q=embedding%E7%BC%96%E7%A0%81+%E5%9F%BA%E4%BA%8E%E7%94%A8%E6%88%B7&qs=n&form=QBRE&sp=-1&pq=embedding%E7%BC%96%E7%A0%81+%E5%9F%BA%E4%BA%8Ey%27h&sc=0-17&sk=&cvid=F08323595985492885CE84F1A1B7F888 | embedding编码基于用户 - 国内版 Bing
https://zhuanlan.zhihu.com/p/53194407 | 万物皆Embedding，从经典的word2vec到深度学习基本操作item2vec - 知乎
https://blog.csdn.net/XD_Cauthy/article/details/106485798 | (27条消息) embedding表示方法及原理_盏茶AI-CSDN博客_embedding
https://zhuanlan.zhihu.com/p/143763320 | 推荐系统 embedding 技术实践总结 - 知乎
https://blog.csdn.net/qq_46006468/article/details/119718868 | (27条消息) Embedding 编码方法_甘霖那的博客-CSDN博客
https://blog.csdn.net/guoyuhaoaaa/article/details/84539971 | (27条消息) 几种推荐搜索场景下的用户Embedding策略_guoyuhaoaaa的博客-CSDN博客
https://www.zhihu.com/question/336110178 | 推荐系统中如何做 User Embedding？ - 知乎
https://www.cnblogs.com/gczr/p/14350567.html | 推荐系统中的ID特征为什么有用？ - 光彩照人 - 博客园
https://github.com/gczr/FM | gczr/FM: 实现FM的基本demo和pytorch版本的FM
https://www.jianshu.com/p/5aea25e80520 | 推荐场景下的ID类特征处理方式 - 简书
https://zhuanlan.zhihu.com/p/104182282 | 浅谈电商搜索推荐中ID类特征的统一建模：Hema Embedding解读 - 知乎
https://blog.csdn.net/weixin_42720875/article/details/106106951 | (27条消息) 推荐问题中ID类特征的处理办法_軔鈳的博客-CSDN博客
https://blog.csdn.net/shine19930820/article/details/71713680 | (27条消息) GBDT原理及利用GBDT构造新的特征-Python实现_YoungChow的博客-CSDN博客
https://blog.csdn.net/Yasin0/article/details/100737222 | (27条消息) GBDT+LR_Yasin的博客-CSDN博客_gbdt lr
https://www.bing.com/search?q=gensim.models+import+Word2Vec&cvid=03204ed583744e37a04dc08cebde345c&aqs=edge..69i57.1534j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | gensim.models import Word2Vec - 国内版 Bing
https://www.cnblogs.com/pinard/p/7278324.html | 用gensim学习word2vec - 刘建平Pinard - 博客园
https://cn.bing.com/search?q=tensorflow1+%E6%95%99%E7%A8%8B&qs=n&form=QBRE&sp=-1&pq=tensorflow1+jiao&sc=0-16&sk=&cvid=DB363202DA19415A93F6FB202D20EF77 | tensorflow1 教程 - 国内版 Bing
http://c.biancheng.net/tensorflow/ | TensorFlow教程：TensorFlow快速入门教程（非常详细）
https://www.cnblogs.com/bclshuai/p/13978984.html | 2.6 tensorflow2.3学习–占位符placeholder - 一字千金 - 博客园
https://www.zhihu.com/question/326362038 | tensorflow2.0成了这样子，连placeholder都没了，你们用吗？ - 知乎
https://link.zhihu.com/?target=http%3A//nooverfit.com/wp/tensorflow-2-0%25E5%2585%25A5%25E5%259D%2591%25E5%25BF%2585%25E5%25A4%2587%25E4%25B9%258Bno-1-%25E7%25A7%25AF%25E6%259E%2581%25E6%25A8%25A1%25E5%25BC%258Feager-execution-david-9%25E7%259A%2584%25E6%258C%2596%25E5%259D%2591%25E5%25AE%259E%25E6%2588%25981/ | 知乎 - 安全中心
https://tensorflow.google.cn/tutorials/quickstart/beginner?hl=zh-cn | 初学者的 TensorFlow 2.0 教程 | TensorFlow Core

https://www.bing.com/search?q=adaboost&qs=AS&pq=ada&sk=SC1AS1&sc=8-3&cvid=D5AA0C885BD24D2B952E89588A86E490&FORM=QBRE&sp=3 | adaboost - 国内版 Bing
https://blog.csdn.net/guyuealian/article/details/70995333 | (27条消息) Adaboost算法原理分析和实例+代码（简明易懂）_pan_jinquan的博客-CSDN博客_adaboost代码
https://blog.csdn.net/quiet_girl/article/details/88756843 | (27条消息) GBDT和Xgboost：原理、推导、比较_小太阳~-CSDN博客_xgboost和gbdt区别
https://blog.csdn.net/zpalyq110/article/details/79527653 | (27条消息) GBDT算法原理以及实例理解_Freemanzxp-CSDN博客_gbdt算法
https://www.cnblogs.com/pinard/p/6140514.html | 梯度提升树(GBDT)原理小结 - 刘建平Pinard - 博客园
https://zhuanlan.zhihu.com/p/39972832 | AdaBoost算法 - 知乎
https://cloud.tencent.com/developer/article/1528767 | GBDT的回归、二分类以及多分类教程 - 云+社区 - 腾讯云
https://murphypei.github.io/blog/2019/04/gbdt | 梯度提升决策树(GBDT) | 拾荒志
https://blog.csdn.net/program_developer/article/details/103060416 | (27条消息) 深入理解GBDT二分类算法_Microstrong-CSDN博客_gbdt分类

https://www.bing.com/search?q=.spec+%E6%96%87%E4%BB%B6&form=ANNTH1&refig=ec2995a20cfc4ddea5b8f698481f01ad | .spec 文件 - 国内版 Bing
https://cloud.tencent.com/developer/news/397399 | 深入Pyinstaller的使用，用spec添加资源文件 - 云+社区 - 腾讯云
https://blog.csdn.net/tangfreeze/article/details/112240342 | (25条消息) Pyinstaller的Spec文件用法_tangfreeze的博客-CSDN博客_pyinstaller spec文件
https://cn.bing.com/search?q=docker+copy+%e5%a4%9a%e4%b8%aa%e6%96%87%e4%bb%b6&qs=HS&pq=docker+copy&sc=8-11&cvid=BB15ECB5DA8D4743AB6CF5C01C288D7E&FORM=QBRE&sp=1 | docker copy 多个文件 - 国内版 Bing
https://www.cnblogs.com/sparkdev/p/9573248.html | Dockerfile 中的 COPY 与 ADD 命令 - sparkdev - 博客园
https://blog.csdn.net/chent86/article/details/78607319 | (25条消息) dockerfile COPY如何同时拷贝多个文件夹_chent86的博客-CSDN博客
https://www.cnblogs.com/YatHo/p/8260777.html | Dockerfile ，ADD详细解读 - YatHo - 博客园
https://www.cnblogs.com/panpanwelcome/p/12604281.html | COPY 复制文件 & ADD 更高级的复制文件—— Dockerfile - PanPan003 - 博客园
http://aiuxian.com/article/p-uszknlrm-bsr.html | Dockerfile：如何使用一个COPY指令复制多个文件？ - 程序园
https://chegva.com/4232.html | Dockerfile COPY 和 ADD 使用规则 | CHEGVA
https://zhuanlan.zhihu.com/p/79949030 | 如何编写优雅的Dockerfile - 知乎
https://qastack.cn/programming/30256386/how-to-copy-multiple-files-in-one-layer-using-a-dockerfile | 如何使用Dockerfile在一层中复制多个文件？
https://cn.bing.com/search?q=%E6%9F%A5%E7%9C%8B%E6%96%87%E4%BB%B6%E5%A4%B9%E5%A4%A7%E5%B0%8F+linux&cvid=9a47aa4fe8c144dc8066a4f91571fe86&aqs=edge.2.69i57j0l8.10350j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | 查看文件夹大小 linux - 国内版 Bing
https://cn.bing.com/search?q=RUN+DEBIAN_FRONTEND%3Dnoninteractive&cvid=6224c353193a4812a8a7946b27425f74&aqs=edge..69i57.971j0j4&FORM=ANAB01&PC=HCTS | RUN DEBIAN_FRONTEND=noninteractive - 国内版 Bing
https://blog.csdn.net/oguro/article/details/102840215 | (25条消息) DEBIAN_FRONTEND noninteractive参数_oguro的博客-CSDN博客
edge://newtab/ | 新建标签页
https://cloud.tencent.com/developer/article/1626811 | Docker 时区调整方案 - 云+社区 - 腾讯云
https://cn.bing.com/search?q=dockerfile+ENV&qs=n&form=QBRE&sp=-1&pq=dockerfile+env&sc=8-14&sk=&cvid=D373AE1252D3436AAE6FCA3582F12F9D | dockerfile ENV - 国内版 Bing
https://zhuanlan.zhihu.com/p/346665702 | Dockerfile ENV 使用指南 - 知乎
https://blog.csdn.net/takujo/article/details/108293616 | (25条消息) linux /etc/profile PATH 设置多个路径_takujo的博客-CSDN博客
https://cn.bing.com/search?q=ubuntu+%e6%b8%85%e7%90%86apt%e7%bc%93%e5%ad%98&qs=AS&pq=ubuntu+%e6%b8%85%e7%90%86apt&sc=2-12&cvid=7813D39F45F04217B285CA2972D02302&FORM=QBRE&sp=1 | ubuntu 清理apt缓存 - 国内版 Bing
https://linux.cn/article-12787-1.html#:~:text=%E4%BD%A0%E5%8F%AA%E9%9C%80%E4%BD%BF%E7%94%A8%E8%BF%99%E4%B8%AA%20apt-get%20%E5%91%BD%E4%BB%A4%20%E9%80%89%E9%A1%B9%EF%BC%9A.%20sudo%20apt-get%20clean.%20%E4%BD%86%E6%98%AF%EF%BC%8C%E6%B8%85%E7%90%86,Ubuntu%20%E7%9A%84%E5%8F%91%E8%A1%8C%E7%89%88%EF%BC%8C%E6%AF%94%E5%A6%82%20Linux%20Mint%E3%80%81Deepin%20%E7%AD%89%E7%AD%89%E3%80%82.%20%E4%BB%80%E4%B9%88%E6%98%AF%20APT%20%E7%BC%93%E5%AD%98%EF%BC%9F. | 技术|如何清除 APT 缓存来回收宝贵的磁盘空间
https://cloud.tencent.com/developer/ask/88117 | 在Dockerfile中，如何更新PATH环境变量？ - 问答 - 云+社区 - 腾讯云
https://zhuanlan.zhihu.com/p/70607853 | 在docker镜像中加入环境变量 - 知乎
https://www.codenong.com/27093612/ | 关于docker：在Dockerfile中，如何更新PATH环境变量？ | 码农家园
https://cn.bing.com/search?q=docker+attach&cvid=3999b472323845758ac35c54bba2555e&aqs=edge.1.69i57j0l8.4716j0j4&FORM=ANAB01&PC=HCTS | docker attach - 国内版 Bing
https://blog.csdn.net/halcyonbaby/article/details/46884605 | (25条消息) Docker exec与Docker attach_追寻神迹-CSDN博客
https://www.runoob.com/docker/docker-attach-command.html | Docker attach 命令 | 菜鸟教程
https://www.cnblogs.com/yikoulinux/p/14401487.html#:~:text=apt%20%E5%92%8C%20apt-get%20%E4%B9%8B%E9%97%B4%E7%9A%84%E7%AC%AC%E4%B8%80%E4%B8%AA%E5%8C%BA%E5%88%AB%E6%98%AF%E5%91%BD%E4%BB%A4%E6%9C%AC%E8%BA%AB%E3%80%82%20apt%20update%20%E5%91%BD%E4%BB%A4%E4%B8%8D%E4%BB%85%E6%9B%B4%E6%96%B0%E5%AD%98%E5%82%A8%E5%BA%93%E7%B4%A2%E5%BC%95%EF%BC%8C%E8%BF%98%E5%91%8A%E7%9F%A5%E5%AD%98%E5%82%A8%E5%BA%93%E4%B8%AD%E6%98%AF%E5%90%A6%E5%8F%AF%E7%94%A8%E8%BD%AF%E4%BB%B6%E4%BB%A5%E5%8F%8A%E6%9C%89%E5%A4%9A%E5%B0%91%E6%96%B0%E7%89%88%E6%9C%AC%E5%8F%AF%E7%94%A8%E3%80%82%201.,%E5%91%BD%E4%BB%A4%E6%9B%B4%E6%96%B0%E7%B3%BB%E7%BB%9F%E4%B8%AD%E5%BD%93%E5%89%8D%E5%AD%98%E5%9C%A8%E7%9A%84%E6%89%80%E6%9C%89%E8%BD%AF%E4%BB%B6%E5%8C%85%EF%BC%8C%E5%AE%83%E4%B8%8D%E4%BC%9A%E5%9C%A8%E7%B3%BB%E7%BB%9F%E4%B8%8A%E5%AE%89%E8%A3%85%E6%88%96%E5%88%A0%E9%99%A4%E7%8E%B0%E6%9C%89%E8%BD%AF%E4%BB%B6%E5%8C%85%E3%80%82%20%E6%96%B0%E7%9A%84%20apt%20upgrade%20%E5%91%BD%E4%BB%A4%E5%B0%86%E5%AE%89%E8%A3%85%E4%BD%9C%E4%B8%BA%E5%8F%AF%E5%8D%87%E7%BA%A7%E8%BD%AF%E4%BB%B6%E5%8C%85%E7%9A%84%E4%BE%9D%E8%B5%96%E9%A1%B9%E6%B7%BB%E5%8A%A0%E7%9A%84%E8%BD%AF%E4%BB%B6%E5%8C%85%E3%80%82%20%E5%B0%BD%E7%AE%A1%E7%B1%BB%E4%BC%BC%E4%BA%8E%20apt-get%20%E5%8D%87%E7%BA%A7%EF%BC%8C%E4%BD%86%E5%AE%83%E4%B9%9F%E4%B8%8D%E4%BC%9A%E5%88%A0%E9%99%A4%E4%BB%A5%E5%89%8D%E5%AE%89%E8%A3%85%E7%9A%84%E8%BD%AF%E4%BB%B6%E5%8C%85%E3%80%82 | apt 和 apt-get 之间有什么区别？ - 一口Linux - 博客园
https://hu1104.github.io/article/hadoop-spark-hbase-hive/#more | hadoop_spark_hbase_hive | Zongxing
https://cn.bing.com/search?q=docker+ssh+%E4%B8%8D%E8%83%BD%E8%87%AA%E5%90%AF%E5%8A%A8+dockerfile&qs=n&form=QBRE&sp=-1&pq=docker+ssh+%E4%B8%8D%E8%83%BD%E8%87%AA%E5%90%AF%E5%8A%A8+dockerfile&sc=0-27&sk=&cvid=39EFB73E9BE442C5B94695A6C1C8178F | docker ssh 不能自启动 dockerfile - 国内版 Bing
https://www.thinbug.com/q/22886470 | 使用docker容器自动启动sshd - Thinbug
https://www.codenong.com/25135897/ | 运行Docker容器时如何自动启动服务？ | 码农家园
https://www.zhangshengrong.com/p/boNwxzMXwZ/ | Docker使用Dockerfile创建支持ssh服务自启动的容器镜像 / 张生荣
https://blog.csdn.net/kongxx/article/details/38412119 | (25条消息) Docker使用Dockerfile创建支持ssh服务自启动的容器镜像_kongxx的专栏-CSDN博客
https://zhuanlan.zhihu.com/p/75575339 | 零基础Docker学习笔记5：docker ubuntu sshd开机启动 - 知乎
https://www.cnblogs.com/ievjai/p/12715632.html | Linux常用命令05 - 文本追加到文件中 - 叉叉敌 - 博客园
https://www.cnblogs.com/dream4567/p/9690850.html | Ubuntu 18.04 国内的 apt 源 - 如梦dream - 博客园
https://cn.bing.com/search?q=debconf%3A+delaying+package+configuration%2C+since+apt-utils+is+not+installed&cvid=e9f17d8743254135a096783fe0fd165d&aqs=edge..69i57j69i58.1175j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | debconf: delaying package configuration, since apt-utils is not installed - 国内版 Bing
https://cn.bing.com/search?q=ssh%3A+connect+to+host+localhost+port+22%3A+Cannot+assign+requested+address&cvid=35683c9c96994b8ba3ea2a48ebbfdfe4&aqs=edge..69i57j69i58.1053j0j4&FORM=ANAB01&PC=HCTS | ssh: connect to host localhost port 22: Cannot assign requested address - 国内版 Bing
https://www.jianshu.com/p/ecfb791dcb3b | localhost: ssh: connect to host localhost port 22: Cannot assign requested address - 简书

https://cn.bing.com/search?q=idea+maven+%e9%85%8d%e7%bd%ae&qs=AS&pq=idea+maven&sk=AS1&sc=8-10&cvid=21EE31D205294417AB05314BBCC87506&FORM=QBRE&sp=2 | idea maven 配置 - 国内版 Bing
https://blog.csdn.net/m0_46491549/article/details/119594094 | (24条消息) Intellij IDEA 2021 Maven 配置指南_月见希子-CSDN博客
https://archive.apache.org/dist/maven/maven-3/3.8.2/binaries/ | Index of /dist/maven/maven-3/3.8.2/binaries
https://maven.apache.org/download.cgi | Maven – Download Apache Maven
https://blog.csdn.net/qq_42057154/article/details/106114515 | (24条消息) IDEA配置Maven（详细版）_Dawn的博客-CSDN博客_idea配置maven
https://blog.csdn.net/yiluohan0307/article/details/79568363 | (24条消息) 利用intellij idea 搭建spark开发环境（windows）_yiluohan0307的专栏-CSDN博客
https://cn.bing.com/search?q=idea+scala+log+%E4%B8%8D%E8%BE%93%E5%87%BA&qs=n&form=QBRE&sp=-1&pq=idea+scala+log+bu%27shu%27c&sc=0-23&sk=&cvid=643CEDB253FB4E8F8F7A641375F60047 | idea scala log 不输出 - 国内版 Bing
https://blog.csdn.net/dabokele/article/details/52599768 | (24条消息) IntelliJ IDEA在Local模式下Spark程序消除日志中INFO输出_dabokele的博客-CSDN博客
https://cn.bing.com/search?q=sc.textFile&cvid=3bf1bbdf636241ec9aa0ffd25a02b7e2&aqs=edge..69i57.5180j0j4&FORM=ANAB01&PC=HCTS | sc.textFile - 国内版 Bing
https://blog.csdn.net/legotime/article/details/51871724 | (24条消息) Spark从外部读取数据之textFile_legotime的博客-CSDN博客_sc.textfile
https://www.datafountain.cn/competitions/542 | 图书推荐系统竞赛 - DataFountain
https://www.bing.com/search?q=log+spark+%E8%BE%93%E5%87%BA&qs=n&form=QBRE&sp=-1&pq=log+spark+shu&sc=0-13&sk=&cvid=8CBA3ED3E1194EFF8746E370E258F42D | log spark 输出 - 国内版 Bing
https://blog.csdn.net/an1090239782/article/details/99579535 | (24条消息) Spark日志，及设置日志输出级别_爱是与世界平行-CSDN博客_spark设置日志打印级别
https://www.bing.com/search?q=docker+pull+%E4%BF%9D%E5%AD%98%E5%9C%B0%E5%9D%80+windows&qs=n&form=QBRE&sp=-1&pq=docker+pull+%E4%BF%9D%E5%AD%98%E5%9C%B0%E5%9D%80+windows&sc=0-24&sk=&cvid=1831411A37144F099563A94D700DB68E | docker pull 保存地址 windows - 国内版 Bing
https://blog.csdn.net/luomin369433047/article/details/105110161 | (24条消息) Windows安装docker后的相关设置设置docker镜像文件位置设置docker目录挂载设置国内镜像源 2375端口查看容器进入容器_罗山仔的博客-CSDN博客
https://blog.csdn.net/u013948858/article/details/80811986 | (24条消息) 【Docker】win10上修改docker的镜像文件存储位置（九）_2021 真实-CSDN博客_docker 镜像位置
https://blog.csdn.net/yiwanxingchena/article/details/78709297 | (24条消息) windows10+docker:镜像文件的存储位置_yiwanxingchena的专栏-CSDN博客
https://blog.csdn.net/qq_41081026/article/details/112314283 | (24条消息) docker pull 镜像存储位置以及默认存储路径的修改_小菜鸟啦啦啦啦的博客-CSDN博客_docker pull 位置
https://www.cnblogs.com/guliang-jieguo/p/14766536.html | Win10 中 Docker 的安装与镜像加速（Docker Desktop for Windows） - 菇凉、借过 - 博客园
https://phpor.net/blog/post/4009 | docker资源限制之IO – PHPor 的Blog
https://docs.docker.com/engine/reference/commandline/dockerd/ | dockerd | Docker Documentation
https://www.bing.com/search?q=Windows+docker+service&cvid=a9b10f6f3f8c439c88c2d1af5d6dc546&aqs=edge..69i57.6830j0j4&FORM=ANAB01&PC=HCTS | Windows docker service - 国内版 Bing
https://www.bing.com/search?q=Docker+Root+Dir+windows&qs=n&form=QBRE&sp=-1&pq=docker+root+dir+&sc=0-16&sk=&cvid=ED120AB0405F45C6A25D68A7A8EC2538 | Docker Root Dir windows - 国内版 Bing
https://blog.csdn.net/u013948858/article/details/80811986 | (24条消息) 【Docker】win10上修改docker的镜像文件存储位置（九）_2021 真实-CSDN博客_docker 镜像位置
https://blog.csdn.net/u013948858/article/details/111464534 | (24条消息) 【Docker】win10上修改docker的镜像文件存储位置（九）- 通过WSL2修改_2021 真实-CSDN博客
https://blog.csdn.net/i6101206007/article/details/113173712 | (24条消息) 修改 docker 默认存储路径（Docker Root Dir）_77wpa的博客-CSDN博客
https://www.bing.com/search?q=Goodbooks-10k&cvid=859a1283204a4e59841b37454687a4b4&aqs=edge.0.69i59j69i60.1493j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | Goodbooks-10k - 国内版 Bing
https://github.com/zygmuntz/goodbooks-10k | zygmuntz/goodbooks-10k: Ten thousand books, six million ratings
https://www.bing.com/search?q=%E6%8A%97%E9%85%B8%E5%88%86%E6%9E%9D%E6%9D%86%E8%8F%8C&cvid=d261ecb3513543059557d9e20c3ef561&aqs=edge.0.69i59j0j69i60l3.1502j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | 抗酸分枝杆菌 - 国内版 Bing
https://baike.baidu.com/item/%E5%88%86%E6%9E%9D%E6%9D%86%E8%8F%8C/540225 | 分枝杆菌_百度百科

edge://newtab/ | 新建标签页
https://blog.csdn.net/yx868yx/article/details/107249792 | (24条消息) 后处理方法NMS、Soft-NMS、 Softer-NMS、WBC、DIoUNMS 、NMS替代算法Confluence_yx868yx的博客-CSDN博客_diounms
https://github.com/Huangdebo/Confluence/blob/master/confluence.py | Confluence/confluence.py at master · Huangdebo/Confluence
https://cn.bing.com/search?q=%E5%87%BD%E6%95%B0%E5%BC%8F%E7%BC%96%E7%A8%8B&cvid=b24d18333cab4478bad7531622c06664&aqs=edge.0.69i59j69i57j0l5j69i61j69i65.2238j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | 函数式编程 - 国内版 Bing
https://zhuanlan.zhihu.com/p/363757919 | 什么是函数式编程？ - 知乎
https://cn.bing.com/search?q=iou+ciou+diou+giou&cvid=169494e353a3424c9dd4476080efb50b&aqs=edge.2.0l4.5799j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | iou ciou diou giou - 国内版 Bing
https://zhuanlan.zhihu.com/p/94799295 | IoU、GIoU、DIoU、CIoU损失函数的那点事儿 - 知乎
https://zhuanlan.zhihu.com/p/270663039 | 目标检测回归损失函数——IOU、GIOU、DIOU、CIOU、EIOU - 知乎
https://www.zhihu.com/people/si-xu-ling-luan-cheng-hai/posts?page=3 | 陈伟 - 知乎
https://zhuanlan.zhihu.com/p/407283625 | IoU、GIoU、DIoU、CIoU - 知乎
https://zhuanlan.zhihu.com/p/142755610 | IoU、GIoU、DIoU、CIoU - 知乎
https://blog.csdn.net/neil3611244/article/details/113794197 | (24条消息) 目标检测回归损失函数——IOU、GIOU、DIOU、CIOU、EIOU_爱做菜的炼丹师-CSDN博客
https://blog.csdn.net/neil3611244?type=blog | (24条消息) 爱做菜的炼丹师_思绪零乱成海_CSDN博客-Tensorflow,无人驾驶,深度学习领域博主
https://blog.csdn.net/neil3611244/article/details/113355028?spm=1001.2014.3001.5502 | (24条消息) 目标检测分类损失函数——Cross entropy、Focal loss_爱做菜的炼丹师-CSDN博客
https://blog.csdn.net/neil3611244/article/details/113355032?spm=1001.2014.3001.5502 | (24条消息) 目标检测回归损失函数——L1、L2、smooth L1_爱做菜的炼丹师-CSDN博客
https://blog.csdn.net/neil3611244/article/details/113355031?spm=1001.2014.3001.5502 | (24条消息) 信息量、信息熵、交叉熵、KL散度、JS散度、Wasserstein距离_爱做菜的炼丹师-CSDN博客
https://blog.csdn.net/leonardohaig/article/details/103394369 | (24条消息) IOU & GIOU & DIOU 介绍及其代码实现_leonardohaig的博客-CSDN博客
https://www.jianshu.com/p/b8d48c99a47c | 卷积神经网络中参数量的计算与模型占用显存的计算 - 简书
https://blog.csdn.net/NODIECANFLY/article/details/102854842 | (24条消息) CNN模型的计算量、参数、显存占用_NODIECANFLY-CSDN博客
https://cn.bing.com/search?q=EIOU+torch%E5%AE%9E%E7%8E%B0&qs=n&form=QBRE&sp=-1&pq=eiou+torchshi%27x&sc=0-15&sk=&cvid=AD24733FFC984EA180E783D248DB678C | EIOU torch实现 - 国内版 Bing
https://jishuin.proginn.com/p/763bfbd5c12f | 一文看尽深度学习中的15种损失函数-技术圈
https://blog.csdn.net/amusi1994/article/details/112966895 | (24条消息) Focal-EIOU Loss：用于精确边界框回归的高效IOU损失_阿木寺的博客-CSDN博客_eiou
https://blog.csdn.net/jining11/article/details/89947541 | (24条消息) pytorch轻松计算模型的参数总量、占用显存量以及flops_羊城迷鹿的博客-CSDN博客
https://blog.csdn.net/qq_42597373/article/details/118977378 | (24条消息) 用PyTorch实现CIoU NMS,DIoU NMS和GIoU NMS_五指峰的博客-CSDN博客
https://blog.csdn.net/qq_41917697/article/details/117471604 | (24条消息) Focal Loss及代码_球场书生的博客-CSDN博客

https://zhuanlan.zhihu.com/p/29534841 | 如何一步一步提高图像分类准确率？ - 知乎
https://cn.bing.com/search?q=c%2B%2B+unique&qs=n&form=QBRE&sp=-1&pq=c%2B%2B+uniqu&sc=0-9&sk=&cvid=4CD78B598AF847619554962997A37B1E | c++ unique - 国内版 Bing
https://www.cnblogs.com/wangkundentisy/p/9033782.html | C++STL中的unique函数解析 - Excaliburer - 博客园
http://c.biancheng.net/view/7255.html | C++ STL迭代器适配器是什么？
http://c.biancheng.net/ | C语言中文网：C语言程序设计门户网站(入门教程、编程软件)
http://c.biancheng.net/stl/ | STL教程：C++ STL快速入门（非常详细）
http://c.biancheng.net/cplus/ | C++入门教程，C++基础教程（更新完毕）
https://cn.bing.com/search?q=c%2b%2b+%e8%bf%ad%e4%bb%a3%e5%99%a8&qs=SC&pq=c%2b%2b+diedai&sc=8-10&cvid=037146448C8D4260BB66DB5FC77EA9DB&FORM=QBRE&sp=1 | c++ 迭代器 - 国内版 Bing
http://c.biancheng.net/view/6675.html | 迭代器是什么，C++ STL迭代器（iterator）用法详解
https://en.cppreference.com/w/cpp/iterator | Iterator library - cppreference.com
https://cn.bing.com/search?q=pytorch+load&cvid=62bb9f56edac4499b8d4d64d96708294&aqs=edge..69i57j0l8.5006j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | pytorch load - 国内版 Bing
https://blog.csdn.net/pearl8899/article/details/109566084 | (23条消息) pytorch(一)模型加载函数torch.load()_凝眸伏笔的博客-CSDN博客
https://pytorch.org/docs/stable/generated/torch.load.html#:~:text=Loads%20an%20object%20saved%20with%20torch.save%20%28%29%20from,moved%20to%20the%20device%20they%20were%20saved%20from. | torch.load — PyTorch 1.9.1 documentation

http://www.pythondoc.com/flask-mega-tutorial/userlogin.html | 用户登录 — flask mega-tutorial 1.1 documentation
https://github.com/miguelgrinberg | miguelgrinberg (Miguel Grinberg)
https://github.com/miguelgrinberg/microblog | miguelgrinberg/microblog: A microblogging web application written in Python and Flask that I developed as part of my Flask Mega-Tutorial series.
https://github.com/miguelgrinberg/flasky/tree/master/app | flasky/app at master · miguelgrinberg/flasky
https://cn.bing.com/search?q=flask%E6%95%B0%E6%8D%AE%E5%BA%93%E8%BF%81%E7%A7%BB&cvid=31eb4903ffb0442b8ce34e7368ce7e6c&aqs=edge..69i57.12570j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | flask数据库迁移 - 国内版 Bing
https://cloud.tencent.com/developer/article/1585940 | 25. Flask 数据库迁移 flask-migrate - 云+社区 - 腾讯云
https://blog.csdn.net/weixin_43790276/article/details/103554632 | (21条消息) Flask 中的数据库迁移_Python碎片的博客-CSDN博客_flask 数据库迁移
https://www.cnblogs.com/itpython/p/10440786.html#:~:text=%E4%BD%BF%E7%94%A8%E6%95%B0%E6%8D%AE%E5%BA%93%E8%BF%81%E7%A7%BB%EF%BC%8C%E5%8F%AF%E4%BB%A5%E8%BF%BD%E8%B8%AA%E6%95%B0%E6%8D%AE%E5%BA%93%E6%A8%A1%E5%BC%8F%E7%9A%84%E5%8F%98%E5%8C%96%EF%BC%8C%E7%84%B6%E5%90%8E%E6%8A%8A%E5%8F%98%E5%8A%A8%E5%BA%94%E7%94%A8%E5%88%B0%E6%95%B0%E6%8D%AE%E5%BA%93%E4%B8%AD%E3%80%82,%E5%9C%A8flask%E4%B8%AD%E4%BD%BF%E7%94%A8Flask-Migrate%E6%9D%A5%E5%AE%9E%E7%8E%B0%E6%95%B0%E6%8D%AE%E5%BA%93%E8%BF%81%E7%A7%BB%EF%BC%8C%E5%B9%B6%E4%B8%94%E9%9B%86%E6%88%90%E5%88%B0Flask-Script%E4%B8%AD%EF%BC%8C%E6%89%80%E6%9C%89%E7%9A%84%E6%93%8D%E4%BD%9C%E9%80%9A%E8%BF%87%E5%91%BD%E4%BB%A4%E6%9D%A5%E5%AE%8C%E6%88%90%E3%80%82%20%E4%B8%BA%E4%BA%86%E5%AF%BC%E5%87%BA%E6%95%B0%E6%8D%AE%E5%BA%93%E8%BF%81%E7%A7%BB%E5%91%BD%E4%BB%A4%EF%BC%8CFlask-Migrate%E4%BD%BF%E7%94%A8%E4%BA%86%E4%B8%80%E4%B8%AAMigrateCommand%E7%B1%BB%EF%BC%8C%E5%8F%AF%E4%BB%A5%E9%99%84%E5%8A%A0%E5%88%B0Flask-Script%E7%9A%84manager%E5%AF%B9%E8%B1%A1%E4%B8%8A%E3%80%82 | flask数据库迁移理解及命令 - 飞鸟的梦想 - 博客园
https://cn.bing.com/search?q=sqlalchemy-migrate+&qs=n&form=QBRE&sp=-1&pq=sqlalchemy-migrate+&sc=2-19&sk=&cvid=C6E8BB3575C34B889FAF12CA75EAF4A7 | sqlalchemy-migrate - 国内版 Bing
https://www.programminghunter.com/article/15331095130/ | Flask实践——Flask 中的数据库 (4) - 编程猎人
https://sqlalchemy-migrate.readthedocs.io/en/latest/versioning.html | Database schema versioning workflow — SQLAlchemy Migrate 0.7.3.dev documentation
https://blog.csdn.net/ljb0077/article/details/96120267 | (22条消息) 【python-flask学习笔记】3.6 SQLAlchemy-flask-migrate_Aboil的博客-CSDN博客
https://blog.csdn.net/saltysoda/article/details/117334183 | (21条消息) sqlalchemy-migrate数据库迁移教学_saltysoda的博客-CSDN博客
https://www.jianshu.com/p/5fd8c2cbad3b | Flask - Migrate - 简书
https://www.jianshu.com/nb/18001835 | Flask 入门 - 文集 - 简书
https://github.com/SingleDiego/Flask-Migrate-Demo | SingleDiego/Flask-Migrate-Demo: Flask-Migrate 一个实例
https://github.com/SingleDiego/Flask-Notes/blob/main/docs/08.Flask-Migrate.md | Flask-Notes/08.Flask-Migrate.md at main · SingleDiego/Flask-Notes
https://www.jianshu.com/p/5fd8c2cbad3b | Flask - Migrate - 简书
https://www.jianshu.com/u/be4ef284317c | SingleDiego - 简书
https://www.jianshu.com/p/d0751de31a0a | 03.Web 表单 - 简书
https://www.jianshu.com/p/017c242fc0e3 | 05.用户登录 - 简书
https://www.jianshu.com/p/1e36cc402173 | 02.模板 - 简书
https://www.jianshu.com/p/8e33f24b8bbc | 01.Hello, World! - 简书
https://github.com/SingleDiego/Flask-Tutorial-Source-Code/tree/main | SingleDiego/Flask-Tutorial-Source-Code
https://www.jianshu.com/p/40f96a8a660b | 04.数据库 - 简书
https://cn.bing.com/search?q=exec+python&qs=n&form=QBRE&sp=-1&pq=exec+python&sc=4-11&sk=&cvid=C9E24EB7401741FD98A0096B8E20BB4D | exec python - 国内版 Bing
https://www.runoob.com/python3/python3-func-exec.html | Python3 exec 函数 | 菜鸟教程
https://cn.bing.com/search?q=torch+view+reshape&qs=n&form=QBRE&sp=-1&pq=torch+view+reshape&sc=0-18&sk=&cvid=EFD5C63F5ADD4C799F386B738AFBF87A | torch view reshape - 国内版 Bing
https://blog.csdn.net/Flag_ing/article/details/109129752 | (22条消息) PyTorch：view() 与 reshape() 区别详解_Flag_ing的博客-CSDN博客
https://blog.csdn.net/chinesehuazhou2/article/details/109759400 | (22条消息) Python调试神器：PySnooper详细使用指南！_Python猫-CSDN博客
edge://newtab/ | 新建标签页
https://www.baidu.com/ | 百度一下，你就知道
https://cn.bing.com/search?q=np+logical_and+%E7%BB%B4%E5%BA%A6&qs=n&form=QBRE&sp=-1&pq=np+logical_and+%E7%BB%B4%E5%BA%A6&sc=0-17&sk=&cvid=3C458B9225AD4ADF8FF9FA32EC541009 | np logical_and 维度 - 国内版 Bing
https://cloud.tencent.com/developer/article/1770089 | python numpy logic_and - 云+社区 - 腾讯云
https://cn.bing.com/search?q=numpy+logic_and&qs=n&form=QBRE&sp=-1&pq=numpy+logic_and&sc=0-15&sk=&cvid=951F6FCA3BAD435AB696C68C7F28E32B | numpy logic_and - 国内版 Bing
https://numpy.org/doc/stable/reference/generated/numpy.logical_and.html | numpy.logical_and — NumPy v1.21 Manual
https://blog.csdn.net/marukoheart/article/details/88091409 | (22条消息) python之Numpy的行列维度、axis=0/1/2…便捷记忆方法_marukoheart的博客-CSDN博客_numpy 行列
https://cn.bing.com/search?q=numpy+%E4%BA%8C%E7%BB%B4%E6%95%B0%E7%BB%84+%E9%80%BB%E8%BE%91%E8%BF%90%E7%AE%97&qs=n&form=QBRE&sp=-1&pq=numpy+%E4%BA%8C%E7%BB%B4%E6%95%B0%E7%BB%84+%E9%80%BB%E8%BE%91yun%27s&sc=0-18&sk=&cvid=F9BA3C510FBF4B98A09882BE61CBE2AD | numpy 二维数组逻辑运算 - 国内版 Bing
https://blog.csdn.net/qq_33356616/article/details/83448709 | (23条消息) Numpy学习笔记（二）：数组的逻辑运算_润物细无声-CSDN博客_numpy 逻辑运算
https://blog.csdn.net/qq_27825451/article/details/82838402 | (23条消息) numpy高级教程之np.where和np.piecewise_MIss-Y的博客-CSDN博客

https://cn.bing.com/search?q=resnet+pytorch+%e5%88%86%e7%b1%bb&qs=n&sp=-1&pq=resnet+pytorch+%e5%88%86%e7%b1%bb&sc=0-17&sk=&cvid=9780935804C445C29FFDC75F1DE35427&first=10&FORM=PORE | resnet pytorch 分类 - 国内版 Bing
https://zhuanlan.zhihu.com/p/383238930 | 在pytorch上导入ResNet模型进行图片分类 - 知乎
https://www.jianshu.com/p/2c37965a06ae | Pytorch_基于预训练的ResNet模型训练自己的分类器 - 简书
https://www.jianshu.com/p/b935e108ba7d | 【pytorch】利用resnet50进行汽车图像分类 - 简书
https://github1s.com/wangzhebufangqi/MachineLearning/blob/master/lab1/models/resnet.py | resnet.py - wangzhebufangqi/MachineLearning - GitHub1s
https://www.cnblogs.com/wzyuan/p/9880342.html | 解读 pytorch对resnet的官方实现 - wuzeyuan - 博客园
https://zhuanlan.zhihu.com/p/350009257 | ResNet网络详解及Pytorch代码实现 - 知乎
https://cn.bing.com/search?q=resnet+pytorch+%E4%B8%8B%E8%BD%BD&qs=n&form=QBRE&sp=-1&pq=resnet+pytorch+%E4%B8%8B%E8%BD%BD&sc=0-17&sk=&cvid=0D7040EAE3D34055A0FF989054041082 | resnet pytorch 下载 - 国内版 Bing
https://blog.csdn.net/zzy_pphz/article/details/105290022 | (21条消息) Pytorch离线下载预训练模型方法（VGG，RESNET等）_zzy_pphz的博客-CSDN博客
https://cn.bing.com/search?q=model+zoo&cvid=fcdcf94e7aec494a802fe1e8de53dc2f&aqs=edge..69i57j69i60.3344j0j4&FORM=ANAB01&PC=HCTS | model zoo - 国内版 Bing
https://pytorch.org/tutorials/beginner/basics/quickstart_tutorial.html | Quickstart — PyTorch Tutorials 1.9.1+cu102 documentation
https://zhuanlan.zhihu.com/p/38088256 | GitHub深度学习模型大大大集合的Model Zoo，支持按框架、按类别筛选 - 知乎
https://modelzoo.co/ | Model Zoo - Deep learning code and pretrained models for transfer learning, educational purposes, and more
https://cn.bing.com/search?q=resnet+%E7%BD%91%E7%9B%98&form=ANNTH1&refig=d8ef6ceb4c024d3587660d22f9d9e277 | resnet 网盘 - 国内版 Bing
https://zhuanlan.zhihu.com/p/353235794 | ResNet50网络结构图及结构详解 - 知乎
https://cn.bing.com/search?q=UserWarning%3A+Named+tensors+and+all+their+associated+APIs+are+an+experimental+feature+and+subject+to+change.+Please+do+not+use+them+for+anything+important+until+they+are+released+as+stable.&cvid=45a42832d7f24d35aac8d478b3a2b31b&aqs=edge..69i57j69i59j69i58.1158j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | UserWarning: Named tensors and all their associated APIs are an experimental feature and subject to change. Please do not use them for anything important until they are released as stable. - 国内版 Bing
https://blog.csdn.net/The_dream1/article/details/118066252 | (21条消息) 9.PyTorch实现MNIST（手写数字识别）（2卷积1全连接）_YJ语-CSDN博客
https://github.com/pytorch/pytorch/issues/54846
https://github.com/pytorch/pytorch/issues/60053 | Annoying warning with nn.MaxPool2d · Issue #60053 · pytorch/pytorch
edge://newtab/ | 新建标签页
https://www.baidu.com/ | 百度一下，你就知道
https://www.baidu.com/s?ie=utf-8&f=8&rsv_bp=1&rsv_idx=2&tn=baiduhome_pg&wd=%E5%A4%9A%E4%BB%BB%E5%8A%A1%E5%AD%A6%E4%B9%A0%20loss&oq=pytorch%2520%25E5%2588%2586%25E7%25B1%25BB%25E7%25BD%2591%25E7%25BB%259C&rsv_pq=865c397c00002ec6&rsv_t=94cbGRRDaJFuanzCODgd5uGm8nyNNAqcBVHX7l1rQ9y%2FyKJMoTTVYl4yfNuCCJiBgD0z&rqlang=cn&rsv_enter=1&rsv_dl=tb&rsv_btype=t&inputT=4340&rsv_sug3=29&rsv_sug1=23&rsv_sug7=100&rsv_sug2=0&rsv_sug4=4340 | 多任务学习 loss_百度搜索
https://www.cnblogs.com/douzujun/p/14633524.html | 多任务学习的多个loss应该如何权衡 - douzujun - 博客园
https://blog.csdn.net/litt1e/article/details/89316705 | (21条消息) pytorch实践(改造属于自己的resnet网络结构并训练二分类网络)_litt1e的博客-CSDN博客
https://www.aiuai.cn/aifarm636.html | Focal Loss 论文理解及公式推导 - AI备忘录
https://zhuanlan.zhihu.com/p/75542467 | [日常] 关于 Focal Loss（附实现代码） - 知乎
https://wqw547243068.github.io/2020/07/28/multi-task/ | 多任务学习-Multi-Task-Learning
https://blog.csdn.net/u011583927/article/details/90716942 | (21条消息) focal loss的几种实现版本(Keras/Tensorflow)_随煜而安的专栏-CSDN博客
https://blog.csdn.net/W1995S/article/details/114687437 | (21条消息) OpenPCDet的多标签分类损失函数使用的是focal loss_W1995S的博客-CSDN博客
https://github.com/zheng-yuwei/multi-label-classification
https://zhuanlan.zhihu.com/p/126161750 | Focal loss tf+pytorch实现（带ohem和label smoothing） - 知乎
https://www.codeleading.com/article/22884940342/ | pytorch多标签分类类别不平衡损失函数 - 代码先锋网
https://zhuanlan.zhihu.com/p/80594704 | 5分钟理解Focal Loss与GHM——解决样本不平衡利器 - 知乎
https://blog.csdn.net/qq_33278884/article/details/91572173 | (21条消息) focal_loss 多类别和二分类 Pytorch代码实现_Mrhan的博客-CSDN博客_focalloss多分类代码
https://www.aiuai.cn/aifarm1344.html | Pytorch - FocalLoss的几种实现 - AI备忘录
https://cloud.tencent.com/developer/article/1669261 | 10分钟理解Focal loss数学原理与Pytorch代码（翻译） - 云+社区 - 腾讯云
https://ptorch.com/news/253.html | 基于Pytorch实现Focal loss.(简单、易用、全中文注释、带例子) - pytorch中文网
edge://newtab/ | 新建标签页

https://www.cnblogs.com/jiangzhaowei/p/10055425.html | Docker镜像构建上下文（Context） - 江召伟 - 博客园
https://www.jianshu.com/p/592ff5ee3b62 | 深入理解 Docker 构建上下文 - 简书
http://www.pythondoc.com/flask-mega-tutorial/database.html | 数据库 — flask mega-tutorial 1.1 documentation
https://ondras.zarovi.cz/sql/demo/ | WWW SQL Designer
https://flask-wtf.readthedocs.io/en/0.15.x/install/ | Installation — Flask-WTF Documentation (0.15.x)
https://cn.bing.com/search?q=flask+openid+%E6%98%AF%E5%95%A5&qs=n&form=QBRE&sp=-1&pq=flask+openid+%E6%98%AF%E5%95%A5&sc=0-15&sk=&cvid=EFE2278F450F4B14BF73BE6118C9AEE6 | flask openid 是啥 - 国内版 Bing
https://blog.csdn.net/aizenggege/article/details/82845373 | (21条消息) flask_openid多节点下登录问题_aizenggege的博客-CSDN博客
https://www.yiibai.com/flask/flask_sending_form_data_to_template.html | Flask表单处理 - Flask教程™
https://blog.csdn.net/weixin_30483495/article/details/101782958 | (21条消息) No module named ‘flask.ext’_weixin_30483495的博客-CSDN博客
https://cn.bing.com/search?q=flask+%E6%95%B0%E6%8D%AE%E5%BA%93&qs=n&form=QBRE&sp=-1&pq=flask+shu%27ju%27k&sc=0-14&sk=&cvid=686F8CE0D53C43179D011E708034E355 | flask 数据库 - 国内版 Bing
https://wizardforcel.gitbooks.io/greyli-flask-tutorial/content/c5-database.html | 第 5 章：数据库 · Flask 入门教程
https://cloud.tencent.com/developer/article/1560357 | Flask 操作Mysql数据库 - flask-sqlalchemy扩展 - 云+社区 - 腾讯云
https://zhuanlan.zhihu.com/p/124194654#:~:text=Flask%E4%B8%AD%E7%9A%84%E6%95%B0%E6%8D%AE%E5%BA%93.,Flask%E5%B9%B6%E4%B8%8D%E5%8E%9F%E7%94%9F%E6%94%AF%E6%8C%81%E6%95%B0%E6%8D%AE%E5%BA%93%EF%BC%8C%E8%80%8C%E6%98%AF%E9%80%9A%E8%BF%87Python%E5%8C%85%E4%BB%A5%E5%8F%8AFlask%E6%95%B0%E6%8D%AE%E5%BA%93%E6%8F%92%E4%BB%B6%E3%80%82%20%E6%95%B0%E6%8D%AE%E5%BA%93%E5%88%86%E4%B8%BA%E5%85%B3%E7%B3%BB%E5%9E%8B%E6%95%B0%E6%8D%AE%E5%BA%93%E5%92%8C%E9%9D%9E%E5%85%B3%E7%B3%BB%E5%9E%8B%E6%95%B0%E6%8D%AE%E5%BA%93%EF%BC%8C%E8%BF%99%E9%87%8C%E6%88%91%E4%BB%AC%E4%BD%BF%E7%94%A8%E5%85%B3%E7%B3%BB%E5%9E%8B%E6%95%B0%E6%8D%AE%E5%BA%93%E3%80%82%E6%88%91%E4%BB%AC%E4%BD%BF%E7%94%A8Flask-SQLAlchemy%E6%8F%92%E4%BB%B6%EF%BC%8C%E5%AE%83%E6%98%AFSQLAlchemy%E7%9A%84Flask%E6%8F%92%E4%BB%B6%E5%8C%85%EF%BC%8C%E5%9F%BA%E4%BA%8E%E5%AF%B9%E8%B1%A1%E5%85%B3%E7%B3%BB%E6%98%A0%E5%B0%84ORM%E3%80%82 | Flask教程4：数据库 - 知乎
https://www.zhihu.com/people/he-tao-68-63-82/posts | 核桃 - 知乎
https://www.cnblogs.com/lnkDel/p/8952957.html | Flask实践——microblog WEB表单 (3) - lnkDel - 博客园
https://cn.bing.com/search?q=virtualenv+python3+%E6%8C%87%E5%AE%9A&qs=n&form=QBRE&sp=-1&pq=virtualenv+python3+%E6%8C%87%E5%AE%9A&sc=0-21&sk=&cvid=DC1BB41B79464EB0806D59E076286606 | virtualenv python3 指定 - 国内版 Bing
https://zhuanlan.zhihu.com/p/38457004 | virtualenv安装并指定python版本 - 知乎
https://www.jianshu.com/p/69de97987bc9 | python3 virtualenv - 简书
http://10.0.10.166:5000/index | Home - microblog
https://cn.bing.com/search?q=FlaskWTFDeprecationWarning%3A+%22flask_wtf.Form%22+has+been+renamed+to+%22FlaskForm%22&cvid=7f27612510c04aab845aaa2bd4e5362b&aqs=edge..69i57j69i58.951j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | FlaskWTFDeprecationWarning: “flask_wtf.Form” has been renamed to “FlaskForm” - 国内版 Bing
https://cn.bing.com/search?q=FSADeprecationWarning%3A+SQLALCHEMY_TRACK_MODIFICATIONS+adds+significant+overhead+and+will+be+disabled+by+default+in+the+future.+Set+it+to+True+or+False+to+suppress+this+warning.&cvid=36a0399fe76f47ce979515a41471e8e9&aqs=edge..69i57j69i58.1129j0j4&FORM=ANAB01&PC=HCTS | FSADeprecationWarning: SQLALCHEMY_TRACK_MODIFICATIONS adds significant overhead and will be disabled by default in the future. Set it to True or False to suppress this warning. - 国内版 Bing
https://cn.bing.com/search?q=importlib&qs=n&form=QBRE&sp=-1&pq=importlib&sc=8-9&sk=&cvid=66668391A5AF4126A2B51EF695919088 | importlib - 国内版 Bing
https://learnku.com/docs/pymotw/importlib-pythons-import-mechanism/3492 | 20.2. importlib — Python 的模块载入机制 |《Python 3 标准库实例教程》| Python 技术论坛
https://www.jb51.net/article/183787.htm | Python动态导入模块：__import__、importlib、动态导入的使用场景实例分析_python_脚本之家
https://www.jb51.net/article/111282.htm | Python中标准模块importlib详解_python_脚本之家
https://blog.csdn.net/edward_zcl/article/details/88809212 | (21条消息) Python中动态导入对象importlib.import_module()的使用_edward_zcl的博客-CSDN博客_importlib.import_module

https://cn.bing.com/search?q=leetcode%E5%88%B7%E9%A2%98%E9%A1%BA%E5%BA%8F&cvid=afb10eb77aac47729448fb7c36f6a841&aqs=edge.6.69i57j0l5j69i59j69i60l2.6806j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | leetcode刷题顺序 - 国内版 Bing
https://www.zhihu.com/question/36738189 | LeetCode按照怎样的顺序来刷题比较好？ - 知乎
https://zhuanlan.zhihu.com/p/339403474 | Github 疯传！史上最强！BAT 大佬「LeetCode刷题手册」电子书开放下载了！ - 知乎
https://zhuanlan.zhihu.com/p/328143744 | 看完这本谷歌师兄的Leetcode刷题笔记，可以直接秒杀 80% 的 Leetcode 题目! - 知乎
https://mp.weixin.qq.com/s?__biz=Mzg4NjUxMzg5MA==&mid=100004597&idx=1&sn=c23bbe37e2c03eea86214825a2055ab0&chksm=4f99db7878ee526e5c422a24d31983cf422b0c15cce90b346cb4c3f12efd2edb11e378fce8b2#rd | BAT大佬写的Leetcode刷题笔记，看完秒杀80%的算法题！
https://zhuanlan.zhihu.com/p/388470520 | ACM金牌选手整理的【LeetCode刷题顺序】 - 知乎
https://cn.bing.com/search?q=leetcode+%E5%B8%B8%E7%94%A8%E7%AD%96%E7%95%A5&cvid=7aa0ef1be5f14c87bae14032dfc04b32&aqs=edge..69i57.11014j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | leetcode 常用策略 - 国内版 Bing
https://www.cnblogs.com/Wu13241454771/p/15156753.html | LeetCode算法：贪心策略到动态规划 - Wu_Candy - 博客园
https://blog.csdn.net/zy450271923/article/details/105298999 | (21条消息) leetcode刷题常用算法总结_张毅的博客-CSDN博客_leetcode算法总结

edge://history/older | 历史记录
https://pillow.readthedocs.io/en/stable/reference/Image.html | Image Module — Pillow (PIL Fork) 8.3.2 documentation
https://matplotlib.org/stable/api/index.html | API Overview — Matplotlib 3.4.3 documentation
https://www.cnblogs.com/skyfsm/p/8276501.html | Python各类图像库的图片读写方式总结 - Madcola - 博客园
https://zhuanlan.zhihu.com/p/397330093 | python上opencv, pillow, matplotlib, skimage 基本读写显示区别 - 知乎
https://docs.scipy.org/doc/numpy-1.17.0/numpy-ref-1.17.0.pdf | NumPy Reference
https://scikit-image.org/docs/stable/api/skimage.io.html | Module: io — skimage v0.18.0 docs
https://www.bing.com/search?q=scipy+&qs=n&form=QBRE&sp=-1&pq=&sc=0-0&sk=&cvid=D50F50E33FDC4AE9B46D44D653523A8C | scipy - Bing
https://docs.scipy.org/doc/scipy/reference/reference/index.html#scipy-api | SciPy documentation — SciPy v1.7.1 Manual
https://docs.scipy.org/doc/scipy/reference/tutorial/io.html | File IO (scipy.io) — SciPy v1.7.1 Manual
https://docs.scipy.org/doc/scipy/reference/ | SciPy documentation — SciPy v1.7.1 Manual
http://ipython.org/ipython-doc/stable/index.html | IPython Documentation — IPython 3.2.1 documentation
https://docs.scipy.org/doc/scipy/reference/tutorial/io.html | File IO (scipy.io) — SciPy v1.7.1 Manual
https://imageio.readthedocs.io/en/stable/userapi.html | Imageio’s user API — imageio 2.9.0 documentation
https://cn.bing.com/search?q=argparse+main%E5%87%BD%E6%95%B0&qs=n&form=QBRE&sp=-1&pq=argparse+mainhan%27shu&sc=0-20&sk=&cvid=9EB1AA6ED12A433795E0B75C82C13F30 | argparse main函数 - 国内版 Bing
https://blog.csdn.net/lichji2016/article/details/120391116 | (21条消息) argparse模块用法_lichji2016的博客-CSDN博客
https://blog.csdn.net/qq_36653505/article/details/83788460 | (21条消息) python3中argparse模块详解_Never-Giveup的博客-CSDN博客_argparse模块
https://www.jianshu.com/p/ea52fdfaa4ad | Python–argparse 模块 - 简书
https://www.jianshu.com/p/a41fbd4919f8 | python argparse 用法总结 - 简书
https://zhuanlan.zhihu.com/p/56922793 | argparse模块用法实例详解 - 知乎
https://docs.python.org/3/library/argparse.html | argparse — Parser for command-line options, arguments and sub-commands — Python 3.9.7 documentation
https://www.jianshu.com/p/417738fc9960 | python：configparser模块 - 简书
https://www.jianshu.com/p/cdf5e1ca6afe | Python 函数式编程（一） - 简书
https://zhuanlan.zhihu.com/p/343553714 | Python函数式编程浅谈 - 知乎
https://www.21ic.com/article/881812.html | 5分钟掌握Python中常见的配置文件 - 21ic电子网
https://blog.csdn.net/sinat_32857543/article/details/118997403 | (21条消息) python四种配置文件_清风专栏-CSDN博客_python配置文件
https://blog.csdn.net/weixin_44409630/article/details/93074115 | (21条消息) python读取配置文件（ini、yaml、xml）_PLDPY的博客-CSDN博客_python读取配置文件
https://www.cnblogs.com/zhangyafei/p/10265072.html | Python项目读取配置的几种方式 - DreamBoy_张亚飞 - 博客园
https://www.cnblogs.com/zkfopen/p/11801706.html | python3使用ConfigParser从配置文件中获取列表 - zkfopen - 博客园
https://www.pgadmin.org/docs/pgadmin4/development/config_py.html | The config.py File — pgAdmin 4 5.2 documentation
https://www.jianshu.com/p/e1befc75a310 | python flask config.py 常量配置 - 简书
https://blog.csdn.net/qq_41924554/article/details/109955474 | (21条消息) python中使用.py配置文件_Estella1024的博客-CSDN博客
https://blog.csdn.net/weixin_40355324/article/details/89671693 | (21条消息) config.py参数解释_G换一种活法的博客-CSDN博客_config.py
https://blog.csdn.net/qq_38213612/article/details/81221772 | (21条消息) DensePose（四）：config.py_Molength的博客-CSDN博客
https://www.jianshu.com/p/417738fc9960 | python：configparser模块 - 简书
https://blog.csdn.net/weixin_42419002/article/details/99307955 | (21条消息) parse_config.py 文件的理解：用于解析网络配置和数据配置_轮子去哪儿了-CSDN博客_parse_config
https://zhuanlan.zhihu.com/p/54764686 | Python项目读取配置的正确姿势 - 知乎
https://blog.csdn.net/weixin_40437821/article/details/105810670 | (21条消息) PSENet-pytorch源码精读（一）config.py_DYF的博客-CSDN博客
https://www.jianshu.com/c/8a40c79feddf?order_by=commented_at | Python 优化–cython/numba/pypy/cppyy/pybind11 - 专题 - 简书
https://cn.bing.com/search?q=pathlib&cvid=316cb54747564b709b8a75f9a74cedb3&aqs=edge.0.69i59j69i57j0l5j69i60l2.4449j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | pathlib - 国内版 Bing
https://www.cnblogs.com/poloyy/p/12435628.html | (oﾟvﾟ)ノ Hi - Python
https://www.cnblogs.com/sigai/p/8074329.html | [pathlib]内置pathlib库的常用属性和方法 - [sigai] - 博客园
https://www.jianshu.com/p/a820038e65c3 | 超好用python库(Pathlib) - 简书
https://cn.bing.com/search?q=Path+replace&qs=n&form=QBRE&sp=-1&pq=path+replace&sc=0-12&sk=&cvid=C0AAC600943F418EB80B68357BD45461 | Path replace - 国内版 Bing
https://docs.python.org/3/library/pathlib.html#pathlib.PurePath.joinpath | pathlib — Object-oriented filesystem paths — Python 3.9.7 documentation
https://blog.csdn.net/u011119817/article/details/107357364 | (21条消息) Pathlib使用简单教程_dddeee的专栏-CSDN博客
https://blog.csdn.net/qq_37266079/article/details/104437947 | (21条消息) Python 中 os.path.abspath(file)与os.path.dirname()以及os.path.basename(file)的用法详解_wengmq的博客-CSDN博客_os.path.abspath(file)

https://www.bing.com/search?q=+libgdk_pixbuf-2.0.so.0&qs=n&form=QBRE&sp=-1&pq=libgdk_pixbuf-2.0.so.0&sc=0-22&sk=&cvid=4261F16084C040FEB415D0499FC6934E | libgdk_pixbuf-2.0.so.0 - Bing
https://www.bing.com/search?q=python+%E6%B7%B7%E5%90%88%E7%BC%96%E7%A8%8B+setup.py&cvid=6e4f4219081f4092b1030fff6f4e936f&aqs=edge..69i57.10750j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | python 混合编程 setup.py - Bing
https://blog.csdn.net/tanmx219/article/details/86665706 | (20条消息) Python与C语言混合编程：通过distutils或setuptools实现的一个简单的C扩展_tanmx219的博客-CSDN博客
https://zhuanlan.zhihu.com/p/276461821 | 花了两天，终于把 Python 的 setup.py 给整明白了 - 知乎
https://www.cnblogs.com/xueweihan/p/12030457.html | Python 库打包分发、setup.py 编写、混合 C 扩展打包的简易指南（转载） - 削微寒 - 博客园
https://www.bing.com/search?q=python+cookbook&cvid=5bb61d7ee6894bbe8c9c12fefce91359&aqs=edge.0.0l4.7671j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | python cookbook - Bing
https://www.jianshu.com/p/847586e53129 | Python cookbook 第三版高清pdf - 简书
https://pan.baidu.com/s/1pL1cI9d | 《Python Cookbook》第三版中文v3.0.0.pdf_免费高速下载|百度网盘-分享无限制
https://zhuanlan.zhihu.com/p/181333808 | 《Python Cookbook》第三版中文PDF文档免费下载 - 知乎
https://python3-cookbook.readthedocs.io/zh_CN/latest/chapters/p15_c_extensions.html | 第十五章：C语言扩展 — python3-cookbook 3.0.0 文档
https://github.com/docker/compose/releases | Releases · docker/compose
https://www.baidu.com/s?wd=docker%20volume&rsv_spt=1&rsv_iqid=0x865174850007d99a&issp=1&f=3&rsv_bp=1&rsv_idx=2&ie=utf-8&tn=baiduhome_pg&rsv_enter=1&rsv_dl=ts_0&rsv_sug3=11&rsv_sug1=11&rsv_sug7=100&rsv_sug2=0&rsv_btype=i&prefixsug=docker%2520volu&rsp=0&inputT=4888&rsv_sug4=4888 | docker volume_百度搜索
https://www.jianshu.com/p/8c22cdfc0ffd | docker volume - 简书
https://www.runoob.com/docker/docker-compose.html | Docker Compose | 菜鸟教程
https://blog.csdn.net/weixin_42028353/article/details/80018028 | (20条消息) Docker的volumes的使用_冷雨泽的博客-CSDN博客_volumes:
https://cloud.tencent.com/developer/article/1352216 | 如何在Ubuntu 18.04上安装Docker Compose - 云+社区 - 腾讯云
https://cn.bing.com/search?q=yaml+python&cvid=d539dbff4d2749b8b82239c1fa7bc7f0&aqs=edge.2.69i57j0l6j69i61j69i60.2438j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | yaml python - 国内版 Bing
https://zhuanlan.zhihu.com/p/82839713 | 使用Python处理yaml格式的数据简单到爆！ - 知乎
https://github.com/ultralytics/yolov5/blob/master/models/common.py | yolov5/common.py at master · ultralytics/yolov5

https://zhuanlan.zhihu.com/p/138816612 | 通俗易懂之YOLOV3-ASFF(自适应特征融合) - 知乎
https://github.com/ruinmessi/ASFF/blob/master/models/yolov3_asff.py | ASFF/yolov3_asff.py at master · ruinmessi/ASFF
https://cn.bing.com/search?q=cbam+pytorch&qs=AS&pq=cbam+&sk=SC2&sc=8-5&cvid=8DCFB9962A9A456AADFE46EF37421664&FORM=QBRE&sp=3 | cbam pytorch - 国内版 Bing
https://zhuanlan.zhihu.com/p/99261200 | pytorch中加入注意力机制（CBAM），以ResNet为例。解析到底要不要用ImageNet预训练？如何加预训练参数？ - 知乎
https://zhuanlan.zhihu.com/p/136308008 | 注意力机制BAM和CBAM详细解析（附代码） - 知乎
https://blog.csdn.net/baidu_32885165/article/details/109226784 | (21条消息) CBAM实现(pytorch)_He_YI的博客-CSDN博客_cbam pytorch
https://github.com/ultralytics/yolov5/blob/master/models/yolo.py | yolov5/yolo.py at master · ultralytics/yolov5
https://blog.csdn.net/qq_40179595/article/details/106320420 | (21条消息) YOLOV3+ASFF论文复现笔记_寒江雪的博客-CSDN博客
https://github.com/WangZhaoxue000/yolov3-asff | WangZhaoxue000/yolov3-asff: Yolov3-asff
https://github.com/Jongchan/attention-module | Jongchan/attention-module: Official PyTorch code for “BAM: Bottleneck Attention Module (BMVC2018)” and “CBAM: Convolutional Block Attention Module (ECCV2018)”
https://blog.csdn.net/SIGAI_CSDN/article/details/88064353 | (21条消息) 论文解读 Receptive Field Block Net for Accurate and Fast Object Detection_SIGAI_CSDN的博客-CSDN博客
https://zhuanlan.zhihu.com/p/57633073 | Receptive Field Block Net - 知乎
https://zhuanlan.zhihu.com/p/43970251 | Receptive Field Block论文笔记 - 知乎
https://blog.csdn.net/u011046017/article/details/109327356 | (21条消息) RFBNet_寂寞的小乞丐-CSDN博客_rfbnet
https://cn.bing.com/search?q=%e7%89%b9%e5%be%81%e8%9e%8d%e5%90%88+%e6%b7%b1%e5%ba%a6%e5%ad%a6%e4%b9%a0&qs=AS&pq=%e7%89%b9%e5%be%81%e8%9e%8d%e5%90%88&sk=AS3&sc=8-4&cvid=D82EEFE5CF124BC6857C40DD67D4211D&FORM=QBRE&sp=4 | 特征融合深度学习 - 国内版 Bing
https://blog.csdn.net/xys430381_1/article/details/88370733 | (21条消息) 深度特征融合—高低层（多尺度）特征融合_xys430381_1的专栏-CSDN博客_特征融合
https://zhuanlan.zhihu.com/p/141685352 | 盘点目标检测中的特征融合技巧（根据YOLO v4总结） - 知乎
https://zhuanlan.zhihu.com/p/158738464 | 深度学习的特征融合方式的思考—论pointwise addition和concatenate的异同 - 知乎
https://blog.csdn.net/qq_43703185/article/details/108201970 | (21条消息) 深度学习之特征融合（高低层特征融合）_薛定谔的炼丹炉！的博客-CSDN博客_深度学习特征融合
https://www.cnblogs.com/hansjorn/p/14295889.html | SPP，PPM、ASPP和FPN结构理解和总结 - Sanny.Liu-CV&&ML - 博客园
https://blog.csdn.net/qq_41174201/article/details/88787770 | (21条消息) 网络模型配件 ASPP 学习笔记_这个军姬不太冷的博客-CSDN博客_aspp网络
https://blog.csdn.net/weixin_38241876/article/details/109993781 | (21条消息) ASPP的pytorch实现_Alphapeople的博客-CSDN博客_aspp pytorch
https://blog.csdn.net/weixin_38145317/article/details/104917218 | (21条消息) pytorch 中register_buffer（）_shuijinghua的博客-CSDN博客_register_buffer
https://cn.bing.com/search?q=faster+rcnn&qs=AS&pq=faster&sc=8-6&cvid=7DBD7B3AB8774FD38C2FC77C98D5A2B9&FORM=QBRE&sp=1 | faster rcnn - 国内版 Bing
https://zhuanlan.zhihu.com/p/31426458 | 一文读懂Faster RCNN - 知乎
https://blog.csdn.net/WZZ18191171661/article/details/79444217 | (21条消息) SSD详解_WZZ18191171661的博客-CSDN博客
https://blog.csdn.net/qq_38675397/article/details/106440638 | (21条消息) SSD网络原理解析_追逐AI的蜗牛的博客-CSDN博客_ssd网络
https://zhuanlan.zhihu.com/p/33544892 | 目标检测|SSD原理与实现 - 知乎
https://www.sohu.com/a/368108696_100007727 | 目标检测 | 清晰易懂的SSD算法原理综述_矩形框
https://zhuanlan.zhihu.com/p/30195134 | 全卷积网络 FCN 详解 - 知乎
https://blog.csdn.net/TYtangyan/article/details/106592461 | (21条消息) resnet50、FPN、Panet结构及代码_TYtangyan的博客-CSDN博客_resnet50+fpn
https://cn.bing.com/search?q=BIFPN++&qs=n&form=QBRE&sp=-1&pq=bifpn+&sc=1-6&sk=&cvid=B4EBB9043A244B01A4982A7EFC8F76DF | BIFPN - 国内版 Bing
https://www.cnblogs.com/xuanyuyt/p/14843532.html | 深度学习笔记（二十二）Structure-Preserving Neural Style Transfer - xuanyuyt - 博客园
https://cloud.tencent.com/developer/article/1820939 | 基于深度学习的目标检测算法面试必备（RCNNYOLOv5） - 云+社区 - 腾讯云
https://www.cnblogs.com/wj-1314/p/12015278.html#::text=%E6%BF%80%E6%B4%BB%E5%87%BD%E6%95%B0%EF%BC%88%E5%8F%88%E5%8F%AB%E6%BF%80,%E6%9C%89%E4%B8%80%E4%B8%AA%E6%BF%80%E6%B4%BB%E5%87%BD%E6%95%B0%E3%80%82 | 深度学习笔记——常用的激活（激励）函数 - 战争热诚 - 博客园
https://zhuanlan.zhihu.com/p/46990010 | 论文解读:Attention is All you need - 知乎
https://zhuanlan.zhihu.com/p/49465950 | SqueezeNet详解 - 知乎
https://zhuanlan.zhihu.com/p/31558773 | CNN模型之SqueezeNet - 知乎
https://blog.csdn.net/csdnldp/article/details/78648543 | (20条消息) SqueezeNet详细解读_花有重开日人无再少年-CSDN博客_squeezenet
https://zhuanlan.zhihu.com/p/32702350 | 解读Squeeze-and-Excitation Networks（SENet） - 知乎

https://blog.csdn.net/weixin_44119362/category_10582826.html | (20条消息) 一点就分享系列_啥都会一点的小程的博客-CSDN博客
https://blog.csdn.net/weixin_44119362/article/details/114289607 | (20条消息) 一点就分享系列（实践篇3-上篇）— 修改YOLOV5 之”魔刀小试“+ Trick心得分享+V5精髓部分源码解读_啥都会一点的小程的博客-CSDN博客
https://blog.csdn.net/weixin_44119362/article/details/115087632 | (21条消息) 一点就分享系列（理解篇3）—Cv任务“新世代”之Transformer（下篇）提前“cv领域展开”——快速学习“视觉transformer的理解”+“一些吐槽”_啥都会一点的小程的博客-CSDN博客
https://blog.csdn.net/weixin_44119362/article/details/113886600 | (21条消息) 一点就分享系列（理解篇3）—Cv任务“新世代”之Transformer系列（中篇-视觉模型篇DETR初代版本）_啥都会一点的小程的博客-CSDN博客
https://cn.bing.com/search?q=%E5%8F%AF%E5%8F%98%E5%BD%A2%E5%8D%B7%E7%A7%AF+pytorch&cvid=67e0b7cb08e0435e84fd5802a0f28871&aqs=edge.2.69i57j0l8.7413j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | 可变形卷积 pytorch - 国内版 Bing
https://blog.csdn.net/LEEANG121/article/details/104234927 | (21条消息) 可变形卷积从概念到实现过程_LEEANG121的博客-CSDN博客_可变形卷积
https://zhuanlan.zhihu.com/p/102707081 | 可变形卷积（4uiiurz1-pytorch版）源码分析 - 知乎
https://cn.bing.com/search?q=asff+yolo&cvid=8434cf029e1e4fb0a4364658ed290540&aqs=edge.5.69i57j0l8.5581j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | asff yolo - 国内版 Bing
https://blog.csdn.net/weixin_42096202/article/details/103293579 | (21条消息) [目标检测]–YOLOV3+ASFF-Learning Spatial Fusion for Single-Shot Object Detection_chenzy_hust的博客-CSDN博客
https://zhuanlan.zhihu.com/p/110205719 | 【CV中的特征金字塔】一，工程价值极大的YOLOv3-ASFF - 知乎
https://github.com/ruinmessi/ASFF | ruinmessi/ASFF: yolov3 with mobilenet v2 and ASFF
https://www.cnblogs.com/wujianming-110117/p/12921279.html | 目标检测中特征融合技术（YOLO v4）（上） - 吴建明wujianming - 博客园
https://cn.bing.com/search?q=%E5%B0%8F%E7%9B%AE%E6%A0%87%E6%A3%80%E6%B5%8B&cvid=4b6b49004de54062a496c8af01f6e476&aqs=edge.0.0l9.6436j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | 小目标检测 - 国内版 Bing
https://blog.csdn.net/qq_42013574/article/details/110221360 | (19条消息) 《小目标目标检测的解决方法及方式》_LeoLee-CSDN博客_小目标检测的解决方案
https://cn.bing.com/search?q=anchor+free+anchor+base&cvid=052ddd076d6c46e9834ba151ef6d54ae&aqs=edge.2.69i57j0l7j69i61.10271j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | anchor free anchor base - 国内版 Bing
https://blog.csdn.net/ytusdc/article/details/107864527 | (19条消息) Anchor-based 与 Anchor-free 优缺点_ytusdc的博客-CSDN博客
https://zhuanlan.zhihu.com/p/122221992 | 目标检测中的anchor-based 和anchor free - 知乎
https://blog.csdn.net/weixin_43750626/article/details/109722034 | (19条消息) 一文读懂anchor-base和anchor-free_文冲的笔记的博客-CSDN博客
https://www.zhihu.com/question/428972054#:~:text=Anchor%20Based%E6%96%B9%E6%B3%95%E6%98%AF%E5%9C%A8Anchor%20Free%E7%9A%84%E5%9F%BA%E7%A1%80%E4%B8%8A%EF%BC%8C%E4%B8%BA%E7%9B%AE%E6%A0%87%E7%9A%84%E5%AE%BD%E9%AB%98%E5%BC%95%E5%85%A5%E4%BA%86%E8%8B%A5%E5%B9%B2%E4%B8%AA%E5%85%88%E9%AA%8C%E5%80%BC%EF%BC%9A%20%E5%AF%B9%E4%BA%8EHeatmap,Label%20Feature%20Map%E4%B8%AD%E7%9A%84%E4%B8%80%E4%B8%AA%E7%82%B9%EF%BC%8C%E5%8F%AA%E6%9C%89%E5%BD%93%E4%BB%A5%E8%BF%99%E4%B8%AA%E7%82%B9%E4%B8%BA%E4%B8%AD%E5%BF%83%EF%BC%8C%E5%85%88%E9%AA%8C%E5%80%BC%E7%9A%84%E5%AE%BD%E9%AB%98%E6%89%80%E7%BB%84%E6%88%90%E7%9A%84%E2%80%9C%E5%81%87%E8%AE%BE%E6%A1%86%22%E4%B8%8E%E7%9C%9F%E5%80%BC%E5%BE%88%E7%9B%B8%E8%BF%91%20%EF%BC%88%E6%AF%94%E5%A6%82IOU%E8%B6%B3%E5%A4%9F%E5%A4%A7%EF%BC%89%EF%BC%8C%E6%89%8D%E4%BC%9A%E5%B0%86%E6%A3%80%E6%B5%8B%E6%A1%86%E7%BC%96%E7%A0%81%E5%88%B0%E8%BF%99%E4%B8%AA%E7%82%B9%E4%B8%8A%E3%80%82 | 单阶段、双阶段、anchor-based、anchor-free这四者之间有什么联系吗？ - 知乎
https://cn.bing.com/search?q=Naiyan+Wang%EF%BC%9ACVPR18+Detection%E6%96%87%E7%AB%A0%E9%80%89%E4%BB%8B%EF%BC%88%E4%B8%8A%EF%BC%89&cvid=3f8347ea11f243fc8b224517ff36a90f&aqs=edge..69i57.1048j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | Naiyan Wang：CVPR18 Detection文章选介（上） - 国内版 Bing
https://zhuanlan.zhihu.com/p/35882192 | CVPR18 Detection文章选介（上） - 知乎
https://www.zhihu.com/people/naiyan-wang | Naiyan Wang - 知乎
https://blog.csdn.net/Mr_Clutch/article/details/119912926 | (20条消息) Yolov5–从模块解析到网络结构修改（添加注意力机制）_Mr_Clutch的博客-CSDN博客
https://blog.csdn.net/hk121/article/details/71439975 | (19条消息) 论文-MS-CNN_hk121的博客-CSDN博客
https://www.cnblogs.com/zhang-yd/p/6568126.html | 论文-MS-CNN - zhang–yd - 博客园
https://blog.csdn.net/App_12062011/article/details/77945816 | (19条消息) 目标检测（十六）–MS-CNN_工作笔记-CSDN博客
https://blog.csdn.net/sinat_39307513/article/details/82839408?utm_medium=distribute.pc_relevant.none-task-blog-2%7Edefault%7ECTRLIST%7Edefault-1.control&depth_1-utm_source=distribute.pc_relevant.none-task-blog-2%7Edefault%7ECTRLIST%7Edefault-1.control | (19条消息) MSCNN论文解读-A Unified Multi-scale Deep Convolutional Neural Network for Fast Object Detection_sinat_39307513的博客-CSDN博客
https://blog.csdn.net/weixin_41560402/article/details/109182138 | (19条消息) MS-CNN_dekiang-CSDN博客
https://cn.bing.com/search?q=torch.meshgrid&cvid=566133af8b4e4c67ad4b7503693c6178&aqs=edge..69i57.1259j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | torch.meshgrid - 国内版 Bing
https://blog.csdn.net/weixin_39504171/article/details/106356977 | (19条消息) torch.meshgrid（）函数解析_小娜美要努力努力的博客-CSDN博客_torch.meshgrid
https://cn.bing.com/search?q=vscode+task.json%E5%92%8Claunch.json&qs=n&form=QBRE&sp=-1&pq=vscode+task.json%E5%92%8Claunch.json&sc=0-28&sk=&cvid=2F45D5436E654D0D8D91B75FEF8A41C5 | vscode task.json和launch.json - 国内版 Bing
https://www.cnblogs.com/z1141000271/p/14963333.html | vscode tasks.json以及launch.json配置详解（待整理） - 猪突猛进！！！ - 博客园
https://zhuanlan.zhihu.com/p/101590167 | CBAM：卷积注意力机制模块 - 知乎

https://docs.neptune.ai/how-to-guides/experiment-tracking/organize-ml-experiments | Organize ML experiments - Documentation
https://neptune.ai/blog/best-tools-to-log-and-manage-ml-model-building-metadata | Best Tools to Log and Manage ML Model Building Metadata - neptune.ai
https://vimsky.com/examples/list/python-all-page-1.html | Python常用类和方法 - 纯净天空
https://cn.bing.com/search?q=mAP%40.5%3A.95&cvid=3396462e22f6427b8a0283c1953decfc&aqs=edge.1.69i57j69i59.12685j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | mAP@.5:.95 - 国内版 Bing
https://blog.csdn.net/ruyingcai666666/article/details/109670567 | (19条消息) mAP@0.5与mAP@0.5:0.95的含义，YOLO_ruyingcai666666的博客-CSDN博客_map@0.5:0.95
https://blog.csdn.net/xys430381_1/article/details/90770520 | (19条消息) 机器学习基础—分类与检测的评价指标—AP，mAP，PRC曲线_xys430381_1的专栏-CSDN博客
https://cn.bing.com/search?q=git+%E6%95%99%E7%A8%8B&cvid=8a7e49cbf00b44dfa46313385fa45034&aqs=edge.0.0j69i64l2j69i60l3j69i65.9534j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | git 教程 - 国内版 Bing
https://zhuanlan.zhihu.com/p/30044692 | Git使用教程,最详细，最傻瓜，最浅显，真正手把手教 - 知乎
http://www.ruanyifeng.com/blog/2015/12/git-cheat-sheet.html | 常用 Git 命令清单 - 阮一峰的网络日志
https://www.liaoxuefeng.com/wiki/896043488029600 | Git教程 - 廖雪峰的官方网站
https://blog.nowcoder.net/n/8903a296456d43a498940a29a9993f4c?from=nowcoder_improve | Git最全教程—听说你精通git操作_牛客博客
https://cloud.tencent.com/developer/article/1662720 | Git实用教程（九）| 使用VS Code进行Git可视化操作 - 云+社区 - 腾讯云
https://gitlab.com/hu1104/git_learn | zongxing / git_learn · GitLab
https://cn.bing.com/search?q=git+reset+--hard+HEAD+%7E&qs=n&form=QBRE&sp=-1&pq=git+reset+--hard+head+%7E&sc=0-23&sk=&cvid=784AC5F66E7045E3A8D3A51704748A42 | git reset –hard HEAD ~ - 国内版 Bing
https://devconnected.com/how-to-git-reset-to-head/#:~:text=To%20hard%20reset%20files%20to%20HEAD%20on%20Git%2C,--hard%20HEAD~2%20%28going%20back%20two%20commits%20before%20HEAD%29 | How To Git Reset to HEAD – devconnected
https://cn.bing.com/search?q=github+%E5%A4%9A%E4%BA%BA%E5%90%88%E4%BD%9C+segmentfault&qs=n&form=QBRE&sp=-1&pq=github+%E5%A4%9A%E4%BA%BA%E5%90%88%E4%BD%9C+segmentfault&sc=0-24&sk=&cvid=740812ED1128425C8CA8EE7F29C2D8E4 | github 多人合作 segmentfault - 国内版 Bing
https://zhuanlan.zhihu.com/p/96518956 | 多人合作使用git，推送代码、和并分支 - 知乎
https://segmentfault.com/a/1190000013059664 | Git 多人协作开发流程 - SegmentFault 思否
https://cn.bing.com/search?q=gitlab+%E5%88%A0%E9%99%A4%E5%88%86%E6%94%AF&cvid=88d581adbb8c4af290b53b6f9ac9ac0e&aqs=edge..69i57j69i64.10715j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | gitlab 删除分支 - 国内版 Bing
https://cn.bing.com/search?q=padding+%E8%AE%A1%E7%AE%97&qs=n&form=QBRE&sp=-1&pq=padding+ji%27suan&sc=0-15&sk=&cvid=F42D4BB647BA4D9BA8840090B17A77A1 | padding 计算 - 国内版 Bing
https://blog.csdn.net/weixin_44903147/article/details/104828982 | (19条消息) 简单理解CNN的padding如何计算_Penn Li的博客-CSDN博客
edge://newtab/ | 新建标签页
edge://newtab/ | 新建标签页
https://www.cnblogs.com/wj-1314/p/12015278.html#:~:text=%E6%BF%80%E6%B4%BB%E5%87%BD%E6%95%B0%EF%BC%88%E5%8F%88%E5%8F%AB%E6%BF%80,%E6%9C%89%E4%B8%80%E4%B8%AA%E6%BF%80%E6%B4%BB%E5%87%BD%E6%95%B0%E3%80%82 | 深度学习笔记——常用的激活（激励）函数 - 战争热诚 - 博客园
https://blog.csdn.net/GrayOnDream/article/details/102955297 | (19条消息) 23种激活函数_梦的灰色边沿…-CSDN博客_silu激活函数
https://zhuanlan.zhihu.com/p/139696588 | Sigmiod、Tanh、Relu、Mish…等激活函数的那点事儿 - 知乎
https://cn.bing.com/search?q=git+%e5%a4%9a%e4%ba%ba%e5%8d%8f%e4%bd%9c&cvid=dc626fee1ae74c4e908c15954190310d&aqs=edge..69i57j69i64l2j69i61l3.18029j0j4&PC=HCTS&first=11&FORM=PORE | git 多人协作 - 国内版 Bing
https://zhuanlan.zhihu.com/p/104757080 | Git内幕系列三：Git多人协同 - 知乎
https://zhuanlan.zhihu.com/p/23478654 | 图文详解如何利用Git+Github进行团队协作开发 - 知乎
https://blog.csdn.net/sculpta/article/details/104448310 | (19条消息) GitHub 多人队伍合作详细教程_深刻的博客-CSDN博客_github 合作
https://www.cnblogs.com/onelikeone/p/6857910.html | Git 多人协作开发的过程 - 行走的思想 - 博客园
https://blog.csdn.net/haoaiqian/article/details/76586895 | (19条消息) Git本地分支与远程分支关联(git branch –set-upstream-to=origin/ )_逸尘的专栏-CSDN博客
https://www.freesion.com/article/24961405152/ | git新分支的创建和合并、冲突解决、分支管理、bug分支 - 灰信网（软件开发博客聚合）
https://www.cnblogs.com/windok/p/12842121.html | Git学习系列（四）创建与合并分支、解决冲突、分支管理策略、Bug分支、Feature分支、多人协作、Rebase - 世界因你而小 - 博客园
https://blog.csdn.net/yytkkn/article/details/109244887 | (19条消息) git分支管理（分支创建、合并、冲突、bug、多人协作、git标签）_yytkkn的博客-CSDN博客
https://cloud.tencent.com/developer/article/1409346 | Git教程5(bug分支和多人协作及标签管理) - 云+社区 - 腾讯云

https://www.cnblogs.com/lulipro/p/6661763.html | g++编译器的使用 - lulipro - 博客园
https://cloud.tencent.com/developer/article/1394309 | g++入门教程 - 云+社区 - 腾讯云
https://zhuanlan.zhihu.com/p/358846624 | 小麦检测trick2 - 知乎
https://zhuanlan.zhihu.com/p/365638511 | Kaggle竞赛中使用YOLOv5将目标检测的性能翻倍的心路历程 - 知乎
https://www.zhihu.com/people/cver-38/posts?page=14 | CVer计算机视觉 - 知乎
https://zhuanlan.zhihu.com/p/340863330 | 旷视提出Self-EMD：无需ImageNet的自监督目标检测 - 知乎
https://zhuanlan.zhihu.com/p/342417254 | 浙江大学提出DyFPN：用于目标检测的动态特征金字塔网络 - 知乎
https://zhuanlan.zhihu.com/p/343481363 | 涨点神器！SoftPool：一种新的池化方法，带你起飞，代码已开源！ - 知乎
https://zhuanlan.zhihu.com/p/344915799 | AAAI 2021 目标检测论文大盘点（YOLObile/R3Det/StarNet等） - 知乎
https://zhuanlan.zhihu.com/p/344988350 | 南航提出ALCNet：用于红外小目标检测的注意力局部对比网络 - 知乎
https://zhuanlan.zhihu.com/p/346006319 | Focal-EIOU Loss：用于精确边界框回归的高效IOU损失 - 知乎
https://zhuanlan.zhihu.com/p/335770407 | 这三篇目标检测论文刚刚开源了！AutoAssign/可变形DETR/DeFCN - 知乎
https://zhuanlan.zhihu.com/p/336681585 | 你的NMS该换了！Confluence：实现更准、更强的目标检测 - 知乎
https://zhuanlan.zhihu.com/p/337737656 | 细长目标检测：分析与改进 - 知乎
https://zhuanlan.zhihu.com/p/357350922 | 56.4 AP！超越YOLOv4，更快更强的CenterNet2来了！ - 知乎
https://zhuanlan.zhihu.com/p/357631998 | CVPR 2021 | 全新Backbone！ReXNet：助力CV任务全面涨点 - 知乎
https://github.com/mostafaibrahim17/VinBigData-Chest-Xrays-Object-detection-/blob/main/yolov5-vinbigdata.ipynb | VinBigData-Chest-Xrays-Object-detection-/yolov5-vinbigdata.ipynb at main · mostafaibrahim17/VinBigData-Chest-Xrays-Object-detection-
https://mostafaibrahim18.medium.com/ | Mostafa Ibrahim – Medium
https://github.com/ZFTurbo/Weighted-Boxes-Fusion/tree/master | ZFTurbo/Weighted-Boxes-Fusion: Set of methods to ensemble boxes from different object detection models, including implementation of “Weighted boxes fusion (WBF)” method.
https://pub.towardsai.net/state-of-the-art-models-in-every-machine-learning-field-2021-c7cf074da8b2 | State of the Art Models in Every Machine Learning Field 2021 | by Mostafa Ibrahim | Towards AI
https://arxiv.org/pdf/2008.13367.pdf | 2008.13367.pdf
https://blog.csdn.net/u011984148/article/details/115713505 | (23条消息) Kaggle竞赛中使用YoloV5将物体检测的性能翻倍的心路历程_AI公园-CSDN博客
https://blog.csdn.net/qq_39056987/article/details/119002910 | (23条消息) YOLOX自定义数据集训练（抢先踩坑）_小菜的博客-CSDN博客
https://hyper.ai/15198 | 用 YOLO v5+DeepSORT ，打造实时多目标跟踪模型 - 人工智能资讯 - 超神经
https://cloud.tencent.com/developer/article/1845383 | 使用YOLOv5模型进行目标检测！ - 云+社区 - 腾讯云
https://cloud.tencent.com/developer/article/1744371 | 如何有效增强数据集，yolov5 mAP从0.46提升到了0.79？ - 云+社区 - 腾讯云
https://jishuin.proginn.com/p/763bfbd5b3ab | （附代码）高级YoloV5指南，使用WBF来提升目标检测性能-技术圈
https://blog.csdn.net/flyfish1986/article/details/117133648 | (23条消息) 目标检测 YOLOv5 常见的边框（bounding box ）坐标表示方法_flyfish-CSDN博客
https://zhuanlan.zhihu.com/p/128640817 | 【目标检测】SNPIER（有效的多尺度训练算法） - 知乎
https://towardsdatascience.com/advanced-yolov5-tutorial-enhancing-yolov5-with-weighted-boxes-fusion-3bead5b71688 | Advanced YoloV5 tutorial — Enhancing YoloV5 with Weighted Boxes Fusion | by Mostafa Ibrahim | Towards Data Science
https://blog.csdn.net/JianqiuChen/article/details/106639429 | (23条消息) 目标检测-提升方案-目标框加权融合-Weighted Boxes Fusion笔记及源码_JianqiuChen的博客-CSDN博客
https://cn.bing.com/search?q=yolo+wbf&qs=n&sp=-1&pq=yolo+wbf&sc=0-8&sk=&cvid=C4B6E27E548E4567B5B89C900B879DC1&first=11&FORM=PORE | yolo wbf - 国内版 Bing
https://my.oschina.net/u/1416903/blog/5024093 | 融合EfficientNet和YoloV5，非常实用的物体检测二阶段pipeline - ronghuaiyang的个人空间 - OSCHINA - 中文开源技术交流社区
https://mp.weixin.qq.com/s?__biz=Mzg5ODAzMTkyMg==&mid=2247494826&idx=1&sn=cad73b0086b8671d058f725baa352f18&chksm=c06a64f7f71dede15960ea4206df4e2826477e899ae612b6b546545dfd9fdc865aa38076fc00&scene=21#wechat_redirect | WBF：优化目标检测，融合过滤预测框
https://github.com/AlexeyAB/darknet | AlexeyAB/darknet: YOLOv4 / Scaled-YOLOv4 / YOLO - Neural Networks for Object Detection (Windows and Linux version of Darknet )
https://aijishu.com/a/1060000000198774 | 高级YoloV5指南，使用WBF来提升目标检测性能 - 极术社区 - 连接 AIoT 开发者与生态服务
https://cn.bing.com/search?q=efficientnet+yolov5+%E7%BB%93%E5%90%88&qs=n&form=QBRE&sp=-1&pq=efficientnet+yolov5+jie%27he&sc=0-26&sk=&cvid=C4B6E27E548E4567B5B89C900B879DC1 | efficientnet yolov5 结合 - 国内版 Bing
https://blog.csdn.net/u011984148/article/details/115774794 | (23条消息) 融合EfficientNet和YoloV5，非常实用的物体检测二阶段pipeline_AI公园-CSDN博客
https://blog.csdn.net/qq_35470686/article/details/116245197 | (23条消息) YOLOv5结合BiFPN_曾星的博客-CSDN博客
https://cn.bing.com/search?q=wbf+yolov5&qs=n&form=QBRE&sp=-1&pq=wbf+yolov5&sc=0-10&sk=&cvid=1BF390F444BC4566825951505ED6D46A | wbf yolov5 - 国内版 Bing
https://blog.csdn.net/shyjhyp11/article/details/109016610 | (23条消息) 随手记_vscode怎样加命令行参数进行调试(python)惊鸿一博-CSDN博客_vscode命令行参数
http://localhost:50544/a1cea3b8588568ef9e54b5f90f6222ea/?utm_campaign=startup&utm_content=&utm_medium=lantern&utm_source=windows#/ | 蓝灯
https://www.bing.com/search?q=markdown+%E5%BC%BA%E8%B0%83&cvid=ce5e1e9e53d24b4d80c8b25ec26ff103&aqs=edge..69i57.9096j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | markdown 强调 - Bing
https://www.jianshu.com/p/59b2e251be2c | Markdown文字加粗与斜体以及换行 - 简书
https://www.bing.com/search?q=markdown+%E6%B3%A8%E9%87%8A&qs=n&form=QBRE&sp=-1&pq=markdown+%E6%B3%A8%E9%87%8A&sc=0-11&sk=&cvid=5A91AE75E63743FD9F08FE4A1261C253 | markdown 注释 - Bing
https://blog.csdn.net/weixin_51917840/article/details/114416273 | (21条消息) YOLOv5代码阅读笔记 - 模型解读_ChiruZy的博客-CSDN博客
https://blog.csdn.net/kobayashi/article/details/109624901 | (29条消息) YOLOv5：代码阅读（一）：train.py；yaml文件_kobayashi_的博客-CSDN博客
https://blog.csdn.net/weixin_42410915/article/details/118785807 | (29条消息) YOLOv5.yaml文件 & 超参详细介绍_王吉吉丫的博客-CSDN博客
https://www.cnblogs.com/E-Dreamer-Blogs/p/13297580.html | yolov5的yaml文件解析 - E-Dreamer - 博客园
https://blog.csdn.net/mary_0830/article/details/107124459 | (29条消息) YOLOv5代码详解（yolov5l.yaml部分）Liaojiajia-Blog-CSDN博客_yolov5代码
https://blog.csdn.net/weixin_42410915/article/details/118785807 | (29条消息) YOLOv5.yaml文件 & 超参详细介绍_王吉吉丫的博客-CSDN博客
https://blog.csdn.net/kobayashi/article/details/109624901 | (21条消息) YOLOv5：代码阅读（一）：train.py；yaml文件_kobayashi_的博客-CSDN博客

edge://newtab/ | 新建标签页

https://blog.csdn.net/ruyingcai666666/article/details/109670567 | (19条消息) mAP@0.5与mAP@0.5:0.95的含义，YOLO_ruyingcai666666的博客-CSDN博客_map@0.5:0.95
https://zhuanlan.zhihu.com/p/93107394#:~:text=%E6%9C%BA%E5%99%A8%E5%AD%A6%E4%B9%A0%E6%9C%89%E5%BE%88%E5%A4%9A%E8%AF%84%E4%BC%B0%E7%9A%84%E6%8C%87%E6%A0%87%E3%80%82.%20%E6%9C%89%E4%BA%86%E8%BF%99%E4%BA%9B%E6%8C%87%E6%A0%87%E6%88%91%E4%BB%AC%E5%B0%B1%E6%A8%AA%E5%90%91%E7%9A%84%E6%AF%94%E8%BE%83%E5%93%AA%E4%BA%9B%E6%A8%A1%E5%9E%8B%E7%9A%84%E8%A1%A8%E7%8E%B0%E6%9B%B4%E5%A5%BD%E3%80%82.%20%E6%88%91%E4%BB%AC%E5%85%88%E4%BB%8E%E6%95%B4%E4%BD%93%E4%B8%8A%E6%9D%A5%E7%9C%8B%E7%9C%8B%E4%B8%BB%E6%B5%81%E7%9A%84%E8%AF%84%E4%BC%B0%E6%8C%87%E6%A0%87%E9%83%BD%E6%9C%89%E5%93%AA%E4%BA%9B%EF%BC%9A.%20%E5%88%86%E7%B1%BB%E9%97%AE%E9%A2%98%E8%AF%84%E4%BC%B0%E6%8C%87%E6%A0%87%EF%BC%9A.%20%E5%87%86%E7%A1%AE%E7%8E%87%20%E2%80%93,Accuracy.%20%E7%B2%BE%E7%A1%AE%E7%8E%87%EF%BC%88%E5%B7%AE%E5%87%86%E7%8E%87%EF%BC%89-%20Precision.%20%E5%8F%AC%E5%9B%9E%E7%8E%87%EF%BC%88%E6%9F%A5%E5%85%A8%E7%8E%87%EF%BC%89-%20Recall.%20F1%E5%88%86%E6%95%B0.%20ROC%E6%9B%B2%E7%BA%BF. | 一文看懂机器学习指标：准确率、精准率、召回率、F1、ROC曲线、AUC曲线 - 知乎
https://blog.csdn.net/xys430381_1/article/details/90770520 | (19条消息) 机器学习基础—分类与检测的评价指标—AP，mAP，PRC曲线_xys430381_1的专栏-CSDN博客
https://blog.csdn.net/qq_29598161/article/details/105081369 | (19条消息) 推荐的准确度评价指标AP，MAP，AUC，NDCG_linkstack的博客-CSDN博客
https://www.cnblogs.com/laozhanghahaha/p/12499979.html | 机器学习–Micro Average,Macro Average, Weighted Average - 老张哈哈哈 - 博客园

https://www.cnblogs.com/xuanmanstein/p/7630047.html | vscode python3 配置生成任务 - 永远的幻想 - 博客园
https://www.php.cn/python-tutorials-391709.html | VSCode下配置python调试运行环境的方法_python-Python教程-PHP中文网
https://blog.csdn.net/qq_36501295/article/details/81558697 | (19条消息) vscode python task.json2.0.0_富贵的博客-CSDN博客
https://blog.csdn.net/jzj_c_love/article/details/101443895?utm_medium=distribute.pc_relevant.none-task-blog-2~default~baidujs_title~default-0.control&spm=1001.2101.3001.4242 | (19条消息) VScode launch.json和tasks.json文件的配置_jzj_c_love的博客-CSDN博客
https://zhuanlan.zhihu.com/p/92175757 | VScode tasks.json和launch.json的设置 - 知乎

https://cn.bing.com/search?q=python+%E5%A4%9A%E7%BA%BF%E7%A8%8B%E5%AE%9E%E7%8E%B0&cvid=e0eb5e6ef6bd43edad6635f438a8d9a8&aqs=edge..69i57.4769j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | python 多线程实现 - 国内版 Bing
https://zhuanlan.zhihu.com/p/402198819 | GPU多卡并行训练总结（以pytorch为例） - 知乎
https://www.cnblogs.com/o-v-o/p/9975355.html | Pytorch Distributed 初始化 - overfitover - 博客园
https://zhuanlan.zhihu.com/p/113694038 | PyTorch分布式训练简明教程 - 知乎
https://zhuanlan.zhihu.com/p/267157806 | “最全PyTorch分布式教程”来了！ - 知乎
https://murphypei.github.io/blog/2020/09/pytorch-distributed | PyTorch 多进程分布式训练实战 | 拾荒志
https://zhuanlan.zhihu.com/p/360405558 | 「新生手册」：PyTorch分布式训练 - 知乎
https://www.cnblogs.com/yh-blog/p/12877922.html | pytorch(分布式)数据并行个人实践总结——DataParallel/DistributedDataParallel - fnangle - 博客园
https://blog.csdn.net/qq_34914551/article/details/110576421 | (28条消息) PyTorch多进程分布式训练最简单最好用的实施办法_我的博客有点东西-CSDN博客
https://www.sdk.cn/details/EN5yDkYQWj9mkx1RVY | PyTorch分布式DPP涉及的基本概念与问题 - sdk社区 | 技术至上
https://blog.csdn.net/github_37320188/article/details/100519346 | (28条消息) Deep Learning:PyTorch 基于docker 容器的分布式训练实践_lo_ong的博客-CSDN博客
https://cn.bing.com/search?q=yolov5+l%E7%82%BC%E4%B8%B9&cvid=6d70f8807ba84fd0a448d6c0a3f0e074&aqs=edge..69i57.10080j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | yolov5 l炼丹 - 国内版 Bing
https://zhuanlan.zhihu.com/p/107230545 | Pytorch 分布式、多进程模块测试 - 知乎

https://cn.bing.com/search?q=subprocess+%E4%BD%BF%E7%94%A8tqdm&qs=n&form=QBRE&sp=-1&pq=subprocess+shi%27ytqdm&sc=0-20&sk=&cvid=A35FB0191BFB47D19DC5F86228CF86CE | subprocess 使用tqdm - 国内版 Bing
https://www.codenong.com/56437081/ | 关于python：在子进程中使用tqdm | 码农家园
https://github.com/codingo/Interlace/blob/master/Interlace/lib/threader.py | Interlace/threader.py at master · codingo/Interlace
https://cn.bing.com/search?q=python+setup&qs=AS&pq=python+set&sk=AS2&sc=8-10&cvid=F6917CDDD7394E088532343D6E37E163&FORM=QBRE&sp=3 | python setup - 国内版 Bing
https://cn.bing.com/search?q=yolov5+test+voc%E6%95%B0%E6%8D%AE%E9%9B%86&qs=n&form=QBRE&sp=-1&pq=yolov5+test+vocshu%27ju%27j&sc=0-23&sk=&cvid=FC1E58B5B4A7472D8E55F9E0DE54F66F | yolov5 test voc数据集 - 国内版 Bing
https://blog.csdn.net/qq_36756866/article/details/109111065 | (28条消息) YOLOv5训练自己的数据集（超详细完整版）_深度学习菜鸟的博客-CSDN博客_yolov5训练自己的数据
https://blog.csdn.net/Chnyac/article/details/79324059 | (28条消息) Python：Python的实用库大全_Chnyac的博客-CSDN博客_python实用库
https://www.zhihu.com/question/24590883 | 哪些 Python 库让你相见恨晚？ - 知乎
https://zhuanlan.zhihu.com/p/27350980 | Python库资源大全 - 知乎
https://baijiahao.baidu.com/s?id=1640618883712993123&wfr=spider&for=pc | 掌握了这24个顶级Python库，你就是大神！
https://cn.bing.com/search?q=kaggle&go=%E6%90%9C%E7%B4%A2&qs=ds&form=QBRE | kaggle - 国内版 Bing
https://www.kaggle.com/c/carvana-image-masking-challenge/overview | Carvana Image Masking Challenge | Kaggle
https://cn.bing.com/search?q=%e5%b1%b1%e4%b8%9c+ai+%e6%af%94%e8%b5%9b&qs=n&sp=-1&pq=shan%27d+ai+%e6%af%94%e8%b5%9b&sc=0-12&sk=&cvid=B25E714BF40941FFBDB751D42C45C491&first=11&FORM=PORE | 山东 ai 比赛 - 国内版 Bing
https://zhuanlan.zhihu.com/p/51869254 | 【资源总结】国内AI领域的赛事全集 - 知乎
https://www.zhihu.com/question/399674482 | 国内有哪些不错的AI大数据竞赛平台？ - 知乎
https://gitee.com/coggle/DIGIX2021-BASELINE | Coggle数据科学/2021-DIGIX-BASELINE
https://gitee.com/Liam1104/competition-baseline | competition-baseline: 数据竞赛Baseline & Topline分享
https://tianchi.aliyun.com/competition/entrance/531925/introduction | 第三届 Apache Flink 极客挑战赛暨AAIG CUP——电商推荐“抱大腿”攻击识别-天池大赛-阿里云天池
https://www.pianshen.com/article/25901721139/ | Recording︱有价值的各类AI、机器学习比赛心得、经验抄录 - 程序员大本营
https://cloud.tencent.com/developer/article/1499385 | 机器学习实践套路总结 - 云+社区 - 腾讯云
https://blog.csdn.net/AugustMe/article/details/108702014 | (28条消息) 收集机器学习/深度学习比赛网站（持续更新）_学无止境、积少成多、厚积薄发-CSDN博客_深度学习比赛
https://js.dclab.run/v2/cmptlist.html?cmptName=&order=desc&orderType=&page=1&pageSize=10&state=&type= | DC竞赛-大数据竞赛平台
edge://newtab/ | 新建标签页
https://www.baidu.com/ | 百度一下，你就知道

https://www.bing.com/search?q=MemoryEfficientSwish&cvid=790ddf3016c048d1a4c6fe7ad526377a&aqs=edge..69i57.625j0j1&pglt=41&FORM=ANNTA1&PC=HCTS&mkt=zh-CN | MemoryEfficientSwish - 国内版 Bing
https://zhuanlan.zhihu.com/p/96773680 | 令人拍案叫绝的EfficientNet和EfficientDet - 知乎
https://www.bing.com/search?q=%E6%A8%A1%E5%9E%8B%E8%9E%8D%E5%90%88+%E7%9B%AE%E6%A0%87%E6%A3%80%E6%B5%8B+kfold&qs=n&sp=-1&pq=%E6%A8%A1%E5%9E%8B%E8%9E%8D%E5%90%88+%E7%9B%AE%E6%A0%87%E6%A3%80%E6%B5%8B+k%27fold&sc=0-16&sk=&cvid=92FDBA00CC414118A10639FCD899E9B2&first=11&FORM=PORE&mkt=zh-CN | 模型融合目标检测 kfold - 国内版 Bing
https://zhuanlan.zhihu.com/p/102817180 | 目标检测比赛中的tricks（已更新更多代码解析） - 知乎
https://blog.csdn.net/moxibingdao/article/details/112598106 | (23条消息) 目标检测多模型集成方法总结_我爱计算机视觉-CSDN博客
https://cloud.tencent.com/developer/article/1631590 | 目标检测 | 盘点目标检测中的特征融合技巧（根据YOLO v4总结） - 云+社区 - 腾讯云
https://zhuanlan.zhihu.com/p/61705517 | 模型融合及python实现 - 知乎
https://blog.csdn.net/xiaosongshine/article/details/88658274 | (23条消息) [深度概念]·K-Fold 交叉验证 (Cross-Validation)的理解与应用_简明AI工作室-CSDN博客_kfold交叉验证
https://blog.csdn.net/mdjxy63/article/details/88604526 | (23条消息) 目标检测当中的模型融合_mdjxy63的博客-CSDN博客_目标检测模型融合
https://www.bing.com/search?q=DyFPN&cvid=ebe3f33a110449ddb08229a85755334e&aqs=edge..69i57j69i61.1063j0j4&FORM=ANAB01&PC=HCTS&mkt=zh-CN | DyFPN - 国内版 Bing
https://github.com/Mingjian-Zhu/DyFPN/tree/master/configs | DyFPN/configs at master · Mingjian-Zhu/DyFPN
https://www.jianshu.com/p/6e12283379f7 | 阅读笔记-DyFPN Dynamic Feature Pyramid Networks for Object Detection - 简书
https://blog.csdn.net/qq_41375609/article/details/94833928 | (23条消息) 【干货】用mmdetection，在COCO数据集上跑通faster R-CNN（测试、训练）_Activewaste-CSDN博客
https://www.bing.com/search?q=%E8%9E%8D%E5%90%88EfficientNet%E5%92%8CYoloV5&cvid=20b7ea9ba82c40ac85b1924c8f029510&aqs=edge..69i57j69i61l2.1569j0j1&pglt=41&PC=HCTS&first=10&FORM=PORE&mkt=zh-CN | 融合EfficientNet和YoloV5 - 国内版 Bing
https://www.cnblogs.com/wujianming-110117/p/12921308.html | 目标检测中特征融合技术（YOLO v4）（下） - 吴建明wujianming - 博客园
https://github.com/Thinklab-SJTU/R3Det_Tensorflow | Thinklab-SJTU/R3Det_Tensorflow: Code for AAAI 2021 paper: R3Det: Refined Single-Stage Detector with Feature Refinement for Rotating Object

https://www.bing.com/search?q=image-fmix%2F&qs=n&form=QBRE&sp=-1&pq=image-fmix%2F&sc=0-11&sk=&cvid=702179031F7F46FCAF7FEA983D778810 | image-fmix/ - Bing
https://blog.csdn.net/weixin_42621901/article/details/111185217 | (23条消息) 深度学习中的数据增强方法之FMix_沐风大大的博客-CSDN博客
https://github.com/ecs-vlc/FMix | ecs-vlc/FMix: Official implementation of ‘FMix: Enhancing Mixed Sample Data Augmentation’
https://zhuanlan.zhihu.com/p/191558572 | 全网最全:盘点那些图像数据增广方式Mosiac,MixUp,CutMix等. - 知乎
https://www.zhihu.com/people/liu-xuan-ye-6 | 林大佬 - 知乎
https://zhuanlan.zhihu.com/p/373480745 | 超越PP-YOLOv2-YOLOR发布，mAP高达55，速度比ScaledYOLOv4还快！ - 知乎
https://zhuanlan.zhihu.com/p/337374574 | 新的SOTA!谷歌团队通过复制粘贴实现目标检测新突破! - 知乎
https://zhuanlan.zhihu.com/p/333957627 | 最近目标检测新范式汇总SparseRCNN,OneNet,DeFCN等 - 知乎
https://www.zhihu.com/question/375794498 | 深度学习的多个loss如何平衡？ - 知乎
https://www.zhihu.com/question/391195715/answer/1421770287 | 如何理解结构化剪枝与非结构化剪枝? - 知乎
https://zhuanlan.zhihu.com/p/191558095 | Pytorch网络压缩系列教程一:Prune你的模型 - 知乎
https://zhuanlan.zhihu.com/p/152180258 | DETR使用Transformer做目标检测的模型迎来了0.2版本的发布 - 知乎
https://github.com/rwightman/pytorch-image-models#:~:text=PyTorch%20Image%20Models%20%28timm%29%20is%20a%20collection%20of,to%20pull%20together%20a%20wide%20variety%20of%20SOTA | rwightman/pytorch-image-models: PyTorch image models, scripts, pretrained weights – ResNet, ResNeXT, EfficientNet, EfficientNetV2, NFNet, Vision Transformer, MixNet, MobileNet-V3/V2, RegNet, DPN, CSPNet, and more
https://www.bing.com/search?q=%E6%8D%9F%E5%A4%B1%E5%87%BD%E6%95%B0&qs=n&form=QBRE&sp=-1&pq=%E6%8D%9F%E5%A4%B1h&sc=0-3&sk=&cvid=F215B593BC0A41F48406EB00E3427781 | 损失函数 - Bing
https://blog.csdn.net/junruitian/article/details/105689571 | (23条消息) 常见损失函数和评价指标总结（附公式&代码）_junruit的博客-CSDN博客
https://www.cnblogs.com/wj-1314/p/12015278.html#:~:text=%E6%BF%80%E6%B4%BB%E5%87%BD%E6%95%B0%EF%BC%88%E5%8F%88%E5%8F%AB%E6%BF%80,%E6%9C%89%E4%B8%80%E4%B8%AA%E6%BF%80%E6%B4%BB%E5%87%BD%E6%95%B0%E3%80%82 | 深度学习笔记——常用的激活（激励）函数 - 战争热诚 - 博客园
https://zhuanlan.zhihu.com/p/58883095 | 常见的损失函数(loss function)总结 - 知乎
https://cn.bing.com/search?q=FMix+%E5%92%8C+mosanic&cvid=44c593a7f9ff494aac4078e51ea30727&aqs=edge.0.69i59j0j69i57j0l3j69i60l2j69i61.2759j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | FMix 和 mosanic - 国内版 Bing
https://blog.csdn.net/hahabeibei123456789/article/details/104605601 | (23条消息) 开源算法FMix：用于深度学习中增强混合样本数据增强_图像算法Imalg的博客-CSDN博客
https://cn.bing.com/search?q=image+HSV-Hue+augmentation&cvid=7a7ef326369e4c1e879af63b6189d446&aqs=edge..69i57.1006j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | image HSV-Hue augmentation - 国内版 Bing
https://www.jianshu.com/p/3edc15cf0e31 | Image Augmentation - 简书
https://cn.bing.com/search?q=faster+rcnn+%E6%8E%A8%E7%90%86%E9%80%9F%E5%BA%A6&qs=n&form=QBRE&sp=-1&pq=faster+rcnn+%E6%8E%A8%E7%90%86su%27du&sc=0-19&sk=&cvid=DA830D2600AE4589BDC05CECBB02BC90 | faster rcnn 推理速度 - 国内版 Bing
https://cloud.tencent.com/developer/news/666055 | YOLOv5 对决 Faster RCNN，谁赢谁输？ - 云+社区 - 腾讯云
https://blog.csdn.net/weixin_40659838/article/details/95174513 | (23条消息) pandas中关于apply、lambda的应用。_挨踢天才的博客-CSDN博客
https://www.jianshu.com/p/4fdd6eee1b06 | pandas apply() 函数用法 - 简书
https://cn.bing.com/search?q=yolo+%E8%BF%94%E5%9B%9E%E7%9A%84%E5%9D%90%E6%A0%87+%E5%BD%92%E4%B8%80%E5%8C%96&qs=n&form=QBRE&sp=-1&pq=yolo+%E8%BF%94%E5%9B%9E%E7%9A%84%E5%9D%90%E6%A0%87+gui%27yi%27h&sc=0-19&sk=&cvid=EE4A0F738B8B418EA0E7B192E505B743 | yolo 返回的坐标归一化 - 国内版 Bing
https://blog.csdn.net/m0_38139098/article/details/107387057 | (23条消息) YOLO　边框预测中的坐标系转换详解_m0_38139098的博客-CSDN博客
https://blog.csdn.net/sinat_26811377/article/details/101055557 | (23条消息) Pandas 报错ValueError: The truth value of a Series is ambiguous. Use a.empty, a.bool(), a.item(), a.an_给我一点温度-CSDN博客
https://cn.bing.com/search?q=%E6%AD%A3%E5%88%99%E8%A1%A8%E8%BE%BE%E5%BC%8F&cvid=f42b4c346a4e4a5f8643120fe1dfed0d&aqs=edge.6.69i57j0l8.4095j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | 正则表达式 - 国内版 Bing
https://cn.bing.com/search?q=numpy+%e6%8e%92%e5%ba%8f+key&qs=AS&pq=numpy+%e6%8e%92%e5%ba%8f&sk=AS6&sc=8-8&cvid=4A4C2A0B38C0475DA9A8BD3D03134586&FORM=QBRE&sp=7 | numpy 排序 key - 国内版 Bing
https://blog.csdn.net/authorized_keys/article/details/111684383 | (23条消息) Python Numpy 按指定行或列排序_authorized_keys的博客-CSDN博客_python按行排序
https://blog.csdn.net/qq_35531549/article/details/88405224 | (23条消息) python sort、sorted高级排序技巧(key的使用)_寸草心的博客-CSDN博客
https://blog.csdn.net/Haiyang_Duan/article/details/79221458 | (23条消息) Python——numpy排序（sort、argsort、lexsort、partition、sorted）_Haiyang_Duan-CSDN博客_numpy排序

https://blog.csdn.net/qq_29598161/article/details/105081369 | (23条消息) 推荐的准确度评价指标AP，MAP，AUC，NDCG_linkstack的博客-CSDN博客
https://blog.csdn.net/Vodka_Lou/article/details/117793101 | (23条消息) 目标检测评价指标（一文看懂）_Vodka_Lou的博客-CSDN博客
https://cloud.tencent.com/developer/article/1624811 | 【必备】目标检测中的评价指标有哪些？ - 云+社区 - 腾讯云
https://www.bing.com/search?q=torchtext+%E8%AE%AF%E9%A3%9E&qs=n&form=QBRE&sp=-1&pq=torchtext+%E8%AE%AF%E9%A3%9E&sc=0-12&sk=&cvid=D69253A9535940689806D20AF61EF6D0&mkt=zh-CN&mkt=zh-CN | torchtext 讯飞 - 国内版 Bing
https://zhuanlan.zhihu.com/p/387630350 | [算法赛-文本分类] 基于用户画像的商品推荐挑战赛-baselines - 知乎
https://blog.csdn.net/qq_33039859/article/details/108232289 | (19条消息) 200825-JupyterLab插件安装后无法正常显示的解决方法_专注机器学习之路-CSDN博客_jupyterlab插件用不了
https://cn.bing.com/search?q=textcnn+pytorch&qs=AS&pq=textcnn&sk=AS1&sc=8-7&cvid=886B9FCB763E40C586C0B8ABF919EB09&FORM=QBRE&sp=2 | textcnn pytorch - 国内版 Bing
https://blog.csdn.net/qq_25037903/article/details/85058217 | (23条消息) pytorch实现textCNN_无所知的博客-CSDN博客_pytorch textcnn
https://www.jianshu.com/p/db400a569730 | 用tf的VocabularyProcessor创建词汇表vocab - 简书
https://blog.csdn.net/weixin_42241438/article/details/116144547 | (21条消息) 使用TorchText库进行文本分类_汐baby go 的博客-CSDN博客
https://www.jianshu.com/p/f44046fdd317 | 基于torchtext的文本分类 - 简书
https://blog.csdn.net/xys430381_1/article/details/90770520 | (23条消息) 机器学习基础—分类与检测的评价指标—AP，mAP，PRC曲线_xys430381_1的专栏-CSDN博客_机器学习ap
https://cloud.tencent.com/developer/article/1624811 | 【必备】目标检测中的评价指标有哪些？ - 云+社区 - 腾讯云
https://blog.csdn.net/Vodka_Lou/article/details/117793101 | (23条消息) 目标检测评价指标（一文看懂）_Vodka_Lou的博客-CSDN博客
https://zhuanlan.zhihu.com/p/91511706 | 常用损失函数和评价指标总结 - 知乎

https://cn.bing.com/search?q=oserror%3A+%5Bwinerror+1455%5D+%E9%A1%B5%E9%9D%A2%E6%96%87%E4%BB%B6%E5%A4%AA%E5%B0%8F%EF%BC%8C%E6%97%A0%E6%B3%95%E5%AE%8C%E6%88%90%E6%93%8D%E4%BD%9C&form=ANSPH1&refig=116f7b6bb19a44bcf065cc3bf380fa26&pc=U531 | oserror: [winerror 1455] 页面文件太小，无法完成操作 - 国内版 Bing
https://www.bing.com/search?q=%E8%AE%AF%E9%A3%9E+%E5%9F%BA%E4%BA%8E%E7%94%A8%E6%88%B7%E7%94%BB%E5%83%8F%E7%9A%84%E5%95%86%E5%93%81%E6%8E%A8%E8%8D%90%E6%8C%91%E6%88%98%E8%B5%9B&qs=n&form=QBRE&sp=-1&pq=%E8%AE%AF%E9%A3%9E+%E5%9F%BA%E4%BA%8E%E7%94%A8%E6%88%B7%E7%94%BB%E5%83%8F%E7%9A%84%E5%95%86%E5%93%81%E6%8E%A8%E8%8D%90%E6%8C%91%E6%88%98%E8%B5%9B&sc=0-17&sk=&cvid=9A80CBD3AEDA4C639AB04342A2ED37B7&mkt=zh-CN | 讯飞基于用户画像的商品推荐挑战赛 - 国内版 Bing
https://blog.csdn.net/sinat_37574187/article/details/119211560 | (19条消息) 基于用户画像的商品推荐挑战赛_人工智能曾小健-CSDN博客
https://blog.csdn.net/weixin_44289754/article/details/118773411 | (19条消息) 商品推荐挑战赛Baseline来袭，赶快阅读代码提交作品吧！_张小磊的博客-CSDN博客
https://blog.csdn.net/lyc44813418/article/details/118383983 | (19条消息) 基于用户画像的商品推荐挑战赛Baseline【线上0.67】_老肥码码码-CSDN博客
https://zhuanlan.zhihu.com/p/387630350 | [算法赛-文本分类] 基于用户画像的商品推荐挑战赛-baselines - 知乎
https://hu1104.github.io/ | Zongxing
https://cn.bing.com/search?q=torchtext&cvid=d8320b337fcb44bd8d80a6442f29eee7&aqs=edge..69i57j0l8.1459j0j1&pglt=41&FORM=ANNTA1&PC=HCTS | torchtext - 国内版 Bing
https://pypi.org/project/torchtext/ | torchtext · PyPI
https://pytorch.org/docs/stable/torch.html | torch — PyTorch 1.9.0 documentation
https://www.cnblogs.com/sandwichnlp/p/11698996.html#textcnn | 文本分类（TextRNN/TextCNN/TextRCNN/FastText/HAN） - 西多士NLP - 博客园
https://zhuanlan.zhihu.com/p/76003775 | 文本分类算法综述 - 知乎
https://www.bing.com/search?q=leetcode%E5%88%B7%E9%A2%98%E9%A1%BA%E5%BA%8F&cvid=75c2cb785307429dbef31dfbbece8483&aqs=edge.2.69i57j0l7j69i61.9256j0j1&pglt=41&PC=HCTS&first=41&FORM=PORE&mkt=zh-CN | leetcode刷题顺序 - 国内版 Bing
https://github.com/youngyangyang04/leetcode-master | youngyangyang04/leetcode-master: LeetCode 刷题攻略：200道经典题目刷题顺序，共60w字的详细图解，视频难点剖析，50余张思维导图，支持C++，Java，Python，Go，JavaScript等多语言版本，从此算法学习不再迷茫！🔥🔥 来看看，你会发现相见恨晚！🚀
https://leetcode-cn.com/circle/article/wGp7Y9/ | 力扣刷题最强指南（版本1.0） - 力扣（LeetCode）
https://www.bing.com/search?q=opencl+opengl+opencv&qs=AS&pq=opencl+opengl&sk=AS1&sc=5-13&cvid=82FD453C3B334E27B574F9AE9CA6ACFD&FORM=QBRE&sp=2&mkt=zh-CN | opencl opengl opencv - 国内版 Bing
https://blog.csdn.net/victory_li/article/details/70054636 | (20条消息) OpenCV/OpenCL/OpenGL区别_victory_li的专栏-CSDN博客
https://blog.csdn.net/lzp_k2/article/details/88035013?utm_medium=distribute.pc_relevant.none-task-blog-2%7Edefault%7EBlogCommendFromMachineLearnPai2%7Edefault-3.control&depth_1-utm_source=distribute.pc_relevant.none-task-blog-2%7Edefault%7EBlogCommendFromMachineLearnPai2%7Edefault-3.control | (20条消息) OpenCV、OpenGL、OpenCL和DirectX 的区别、联系、前景_lzp_k2的专栏-CSDN博客_opencv opengl区别
https://pypi.org/project/PyOpenGLng/ | PyOpenGLng · PyPI
https://blog.csdn.net/xiaoxianerqq/article/details/88707565 | (20条消息) python如何安装配置OpenGL？_xiaoxianerqq的专栏-CSDN博客_python安装opengl
https://www.lfd.uci.edu/~gohlke/pythonlibs/#pyopengl | Python Extension Packages for Windows - Christoph Gohlke
https://www.bing.com/search?q=opengl&qs=AS&pq=opengl&sc=8-6&cvid=1D602B51F06C4A63A189A4A75FB8C89C&FORM=QBRE&sp=1 | opengl - 国内版 Bing
https://www.khronos.org/opengl/wiki/Language_bindings#Python | Language bindings - OpenGL Wiki
https://github.com/moderngl/moderngl | moderngl/moderngl: Modern OpenGL binding for python
http://pyopengl.sourceforge.net/ | PyOpenGL – The Python OpenGL Binding
https://blog.csdn.net/xufive/article/details/86565130 | (20条消息) 写给 python 程序员的 OpenGL 教程_Python作业辅导员 - 天元浪子-CSDN博客
https://learnopengl-cn.github.io/01%20Getting%20started/01%20OpenGL/ | OpenGL - LearnOpenGL CN
https://cn.bing.com/search?q=%E4%B8%8A%E6%B5%B7%E8%90%BD%E6%88%B7%E5%AE%8C%E8%A6%81%E9%80%80%E5%B7%A5%EF%BC%9F&qs=n&form=QBRE&sp=-1&pq=%E4%B8%8A%E6%B5%B7%E8%90%BD%E6%88%B7%E5%AE%8C%E8%A6%81%E9%80%80%E5%B7%A5&sc=0-8&sk=&cvid=F7D01536A1C540358AE0E55C7706305F | 上海落户完要退工？ - 国内版 Bing
https://cn.bing.com/search?q=%E6%99%AE%E9%99%80%E5%B0%B1%E4%B8%9A%E4%BF%83%E8%BF%9B%E4%B8%AD%E5%BF%83&cvid=8caa0270699d4eadbe68c5a1ce1314af&aqs=edge.0.0j69i57.9937j0j4&FORM=ANAB01&PC=HCTS | 普陀就业促进中心 - 国内版 Bing
https://cn.bing.com/search?q=%e6%9e%84%e5%bb%ba%e8%af%8d%e6%b1%87%e8%a1%a8&qs=AS&pq=%e6%9e%84%e5%bb%ba%e8%af%8d%e6%b1%87%e8%a1%a8&sc=1-5&cvid=92947A9B9557444EBF60A339ABF7D242&FORM=QBRE&sp=1 | 构建词汇表 - 国内版 Bing
https://blog.csdn.net/kunAUGUST/article/details/106886070 | (22条消息) 使用Counter构建词汇表_kunAUGUST的博客-CSDN博客
https://zhuanlan.zhihu.com/p/63974131 | 在深度学习模型中词向量的 - 知乎
https://wmathor.com/index.php/archives/1445/ | TextCNN的PyTorch实现 - mathor
https://blog.csdn.net/qq_25037903/article/details/85058217 | (22条消息) pytorch实现textCNN_无所知的博客-CSDN博客_pytorch textcnn
https://github.com/orgs/datawhalechina/repositories | Datawhale
https://github.com/DLLXW/data-science-competition/blob/main/%E5%A4%A9%E6%B1%A0/%E5%8C%BB%E5%AD%A6%E5%BD%B1%E5%83%8F%E6%8A%A5%E5%91%8A%E5%BC%82%E5%B8%B8%E6%A3%80%E6%B5%8B/train.py | data-science-competition/train.py at main · DLLXW/data-science-competition
https://github.com/DLLXW/data-science-competition/blob/faf7aad6a90d843b61e9ec8a0d41702b17c53a51/%E5%A4%A9%E6%B1%A0/%E5%8C%BB%E5%AD%A6%E5%BD%B1%E5%83%8F%E6%8A%A5%E5%91%8A%E5%BC%82%E5%B8%B8%E6%A3%80%E6%B5%8B/net.py | data-science-competition/net.py at faf7aad6a90d843b61e9ec8a0d41702b17c53a51 · DLLXW/data-science-competition
https://blog.csdn.net/qq_38883844/article/details/104331382 | (23条消息) Pytorch中的nn.Embedding()_奥特曼丶毕健旗的博客-CSDN博客_nn.embedding pytorch
https://cn.bing.com/search?q=WBF+%E7%9B%AE%E6%A0%87%E6%A3%80%E6%B5%8B&cvid=44933841ca3745fc8eb800fdb4853f77&aqs=edge..69i57j69i61l2.14534j0j4&FORM=ANAB01&PC=HCTS | WBF 目标检测 - 国内版 Bing
https://cn.bing.com/search?q=%E5%8A%A0%E6%9D%83%E6%A1%86%E8%9E%8D%E5%90%88&cvid=d1ebe19e748b4c8f90740e30497bc6b3&aqs=edge..69i57.1579j0j4&FORM=ANAB01&PC=HCTS | 加权框融合 - 国内版 Bing
https://blog.csdn.net/jacke121/article/details/114769857 | (23条消息) 加权边界框融合(WBF)_jacke121的专栏-CSDN博客

https://github.com/ruotianluo/pytorch-faster-rcnn | ruotianluo/pytorch-faster-rcnn: pytorch1.0 updated. Support cpu test and demo. (Use detectron2, it’s a masterpiece)
https://cn.bing.com/search?q=yolov5+%e6%94%b9%e8%bf%9b&qs=n&sp=-1&pq=yolov5+%e6%94%b9%e8%bf%9b&sc=0-9&sk=&cvid=F5E226947E254B6BBD9E740473F2244D&first=11&FORM=PORE | yolov5 改进 - 国内版 Bing
https://blog.csdn.net/weixin_44119362/article/details/114289607 | (18条消息) 一点就分享系列（实践篇3-上篇）— 修改YOLOV5 之”魔刀小试“+ Trick心得分享+V5精髓部分源码解读_啥都会一点的小程的博客-CSDN博客
https://blog.csdn.net/weixin_44119362/article/details/116785798 | (18条消息) 一点就分享系列（实践篇3-中篇）— 虽迟但到！全网首发？yolov5之“baseline修改小结“+“CV领域展开-Involution&&transformer&&cnn”啥都会一点的小程的博客-CSDN博客
https://zhuanlan.zhihu.com/p/390301088 | YOLOV5目标检测改进策略 - 知乎
https://chowdera.com/2021/04/20210406232411097H.html | yolov5改进VariFocalNet - 文章整合
https://tianchi.aliyun.com/notebook-ai/detail?postId=116822 | yolov5加全局nms 第八名方案分享-天池实验室-实时在线的数据分析协作工具，享受免费计算资源
https://github.com/jianpingZhonggit/tianchi-ocr/blob/master/code/yolov5-master/result_merge.py | tianchi-ocr/result_merge.py at master · jianpingZhonggit/tianchi-ocr
https://ask.csdn.net/questions/7416579 | (18条消息) 如何将YOLOv5中的NMS替换成DIOU NMS?-Python-CSDN问答
https://github.com/Zzh-tju/yolov5 | Zzh-tju/yolov5: YOLOv5 in PyTorch > ONNX > CoreML > iOS
https://github.com/ultralytics/yolov3/issues/679 | INCREASING NMS SPEED · Issue #679 · ultralytics/yolov3
https://github.com/Zzh-tju/ultralytics-YOLOv3-Cluster-NMS/blob/9b3ede15c636fb44b92b6c7b547fce494b80c09d/utils/utils.py#L629 | ultralytics-YOLOv3-Cluster-NMS/utils.py at 9b3ede15c636fb44b92b6c7b547fce494b80c09d · Zzh-tju/ultralytics-YOLOv3-Cluster-NMS
https://github.com/hu1104/hu1104.github.io/tree/backup | hu1104/hu1104.github.io at backup
https://zhuanlan.zhihu.com/p/102817180 | 目标检测比赛中的tricks（已更新更多代码解析） - 知乎
https://blog.csdn.net/moxibingdao/article/details/112598106 | (18条消息) 目标检测多模型集成方法总结_我爱计算机视觉-CSDN博客
https://zhuanlan.zhihu.com/p/162730118 | MMDetection v2.0 训练自己的数据集 - 知乎
https://www.zhihu.com/people/cjsnx-shsb/posts | cjsnx shsb - 知乎
https://zhuanlan.zhihu.com/p/151002502 | MMDetection v2.0 安装配置 - 知乎
https://zhuanlan.zhihu.com/p/101225733 | MMDetection中文文档—详解 - 知乎
https://zhuanlan.zhihu.com/p/101202864 | MMDetection中文文档—1.安装 - 知乎
https://cn.bing.com/search?q=mmdetection+%E6%94%AF%E6%8C%81wsl2&qs=n&form=QBRE&sp=-1&pq=mmdetection+%E6%94%AF%E6%8C%81wsl2&sc=0-18&sk=&cvid=4C9F602A51174CF4B1F07C8DDF970C9A | mmdetection 支持wsl2 - 国内版 Bing
https://openmmlab.com/installation | OpenMMLab
https://cn.bing.com/search?q=TTA&qs=n&form=QBRE&sp=-1&pq=tta&sc=8-3&sk=&cvid=4E84028D571840EF8FF357EA9499EF72 | TTA - 国内版 Bing
https://zhuanlan.zhihu.com/p/131539596 | TTA(Test-Time Augmentation) 之Pytorch - 知乎
https://registry.hub.docker.com/r/pytorch/pytorch/tags?page=1&ordering=last_updated | pytorch/pytorch Tags | Docker Hub
edge://newtab/ | 新建标签页
https://blog.csdn.net/mzpmzk/article/details/88065416 | (18条消息) PASCAL VOC 数据集简介_Man-CSDN博客_voc数据集
https://www.bing.com/search?q=mmdetction+%E8%AE%AD%E7%BB%83%E8%87%AA%E5%B7%B1%E7%9A%84%E6%95%B0%E6%8D%AE%E9%9B%86+VOC+png&qs=n&form=QBRE&sp=-1&pq=mmdetction+%E8%AE%AD%E7%BB%83%E8%87%AA%E5%B7%B1%E7%9A%84%E6%95%B0%E6%8D%AE%E9%9B%86+vo+png&sc=0-26&sk=&cvid=38F622E850914071BFD4E699DFF157A6 | mmdetction 训练自己的数据集 VOC png - 国内版 Bing
https://www.jianshu.com/p/924db4d52213 | 将自己数据集转化成voc数据集格式并用mmdetection训练 - 简书
https://blog.csdn.net/qq_33897832/article/details/103995636 | (18条消息) （详细教程）mmdetection训练自己的模型，测试，评估_若风的博客-CSDN博客
https://zhuanlan.zhihu.com/p/162730118 | MMDetection v2.0 训练自己的数据集 - 知乎
https://blog.csdn.net/qq_41251963/article/details/112940253 | (18条消息) 【MMDetection-学习记录】训练自己的VOC数据集_乐亦亦乐的博客-CSDN博客
https://zhuanlan.zhihu.com/p/333890175 | MMDetection2（四）：训练自己的数据集 - 知乎
https://github.com/open-mmlab/mmdetection/blob/master/configs/faster_rcnn/faster_rcnn_r50_fpn_1x_coco.py | mmdetection/faster_rcnn_r50_fpn_1x_coco.py at master · open-mmlab/mmdetection
https://github.com/open-mmlab/mmdetection/blob/master/docker/Dockerfile | mmdetection/Dockerfile at master · open-mmlab/mmdetection
https://www.jianshu.com/p/42891e5a0422 | MMDetection 2.5 目标检测（4）：模型训练和测试 - 简书
https://www.jianshu.com/p/0aed4739fa92 | MMDetection 2.5 目标检测（2）：数据准备 - 简书
https://www.jianshu.com/p/38f4525246e1 | MMDetection 2.5 目标检测（3）：配置修改 - 简书
https://cn.bing.com/search?q=%e7%81%ab%e7%84%b0%e6%a3%80%e6%b5%8b+GitHub&cvid=d24d56486b614b7d9087efb5d92e0ef0&aqs=edge..69i57j0l8.7805j0j1&pglt=41&PC=HCTS&first=41&FORM=PORE | 火焰检测 GitHub - 国内版 Bing
https://zhuanlan.zhihu.com/p/367323739 | yolov3训练实现火焰检测(pytorch) - 知乎
https://blog.csdn.net/Zhangrx/article/details/108992426 | (19条消息) 基于yolov5的火焰+烟雾识别_Rena的博客-CSDN博客
https://blog.csdn.net/haoji007/article/details/119215251 | (19条消息) 火灾烟雾检测数据集和yolov4-yolov5检测模型_haoji007的博客-CSDN博客_烟火数据集
https://github.com/gengyanlei/fire-smoke-detect-yolov4 | gengyanlei/fire-smoke-detect-yolov4: fire-smoke-detect-yolov4-yolov5 and fire-smoke-detection-dataset 火灾检测，烟雾检测
https://github.com/ModelBunker/YOLOv3-PyTorch | ModelBunker/YOLOv3-PyTorch: YOLOv3: An Incremental Improvement

https://cn.bing.com/search?q=%E5%9B%BE%E5%83%8F%E6%B7%BB%E5%8A%A0%E6%96%87%E5%AD%97+%E4%B8%8D%E5%87%BA&qs=n&form=QBRE&sp=-1&pq=%E5%9B%BE%E5%83%8F%E6%B7%BB%E5%8A%A0%E6%96%87%E5%AD%97+%E4%B8%8D%E5%87%BA&sc=0-9&sk=&cvid=10A306BB89D2452A8C39BBF6275AD9E6 | 图像添加文字不出 - 国内版 Bing
https://www.geeksforgeeks.org/python-pil-imagedraw-draw-text/ | Python PIL | ImageDraw.Draw.text() - GeeksforGeeks
https://hu1104.github.io/article/yolov5_real/#more | yolov5_real | Zongxing
https://cn.bing.com/search?q=cv2.putText+%E8%B6%85%E5%87%BA%E8%BE%B9%E7%95%8C&qs=n&form=QBRE&sp=-1&pq=cv2.puttext+%E8%B6%85%E5%87%BAbian%27jie&sc=0-22&sk=&cvid=B348AEBABE9E4F5AB30F744BDB112B27 | cv2.putText 超出边界 - 国内版 Bing
https://www.geeksforgeeks.org/python-opencv-cv2-puttext-method/#:~:text=OpenCV-Python%20is%20a%20library%20of%20Python%20bindings%20designed,color%20%20%20thickness%20%5B%2C%20lineType%20%5B%2C%20bottomLeftOrigin%5D%5D%5D%29 | Python OpenCV | cv2.putText() method - GeeksforGeeks
https://cn.bing.com/search?q=markdown+%E8%A1%A8%E6%A0%BC&cvid=7fedb785afcc484db6c82ea249aa0024&aqs=edge.0.0l9.9791j0j4&FORM=ANAB01&PC=HCTS | markdown 表格 - 国内版 Bing
https://markdown.com.cn/extended-syntax/tables.html | Markdown 表格 | Markdown 语法教程
https://cn.bing.com/search?q=pip+%E5%9B%BD%E5%86%85%E6%BA%90&cvid=b8e891121dcd45f8be2ef29be7511908&aqs=edge.2.69i57j69i59j0l2j69i61j69i60l2.2612j0j1&pglt=43&FORM=ANNTA1&PC=HCTS | pip 国内源 - 国内版 Bing
https://zhuanlan.zhihu.com/p/109939711#:~:text=%E8%A7%A3%E5%86%B3%E5%8A%9E%E6%B3%95%E6%98%AF%E6%8A%8A%20pip%20%E6%BA%90%E6%8D%A2%E6%88%90%E5%9B%BD%E5%86%85%E7%9A%84%EF%BC%8C%E6%9C%80%E5%B8%B8%E7%94%A8%E7%9A%84%E5%B9%B6%E4%B8%94%E5%8F%AF%E4%BF%A1%E8%B5%96%E7%9A%84%E6%BA%90%E5%8C%85%E6%8B%AC%E6%B8%85%E5%8D%8E%E5%A4%A7%E5%AD%A6%E6%BA%90%E3%80%81%E8%B1%86%E7%93%A3%E6%BA%90%E3%80%81%E9%98%BF%E9%87%8C%E6%BA%90%EF%BC%9A.%20pypi%20%E6%B8%85%E5%8D%8E%E5%A4%A7%E5%AD%A6%E6%BA%90%EF%BC%9A.%20https%3A%2F%2F.%20pypi.tuna.tsinghua.edu.cn.,%2Fsimple.%20pypi%20%E8%B1%86%E7%93%A3%E6%BA%90%20%EF%BC%9A.%20http%3A%2F%2F.%20pypi.douban.com%2Fsimple%2F.%20pypi%20%E8%85%BE%E8%AE%AF%E6%BA%90%EF%BC%9A. | Python 修改 pip 源为国内源 - 知乎
https://cn.bing.com/search?q=yolov5+%E6%8D%A2backbone&qs=n&form=QBRE&sp=-1&pq=yolov5+%E6%8D%A2backbone&sc=0-16&sk=&cvid=912DEC1DD68845968D4BFF890BB03C0F | yolov5 换backbone - 国内版 Bing
https://blog.csdn.net/wa1tzy/article/details/114492726 | (18条消息) YOLOv5系列（3）——YOLOv5修改网络结构_wa1tzy的博客-CSDN博客_yolov5 注意力
https://github.com/search?q=lane+detection | Search · lane detection
https://arxiv.org/abs/1802.05591 | [1802.05591] Towards End-to-End Lane Detection: an Instance Segmentation Approach
https://paperswithcode.com/paper/towards-end-to-end-lane-detection-an-instance | Towards End-to-End Lane Detection: an Instance Segmentation Approach | Papers With Code
https://zhuanlan.zhihu.com/p/259778366 | UltraFastYoloV5发布!基于剪枝技术的模型压缩,3ms推理的YoloV5mAP60+! - 知乎
https://zhuanlan.zhihu.com/p/242456389 | 震惊! 手把手教你100行代码搭建YoloV5! - 知乎
https://cn.bing.com/search?q=tqdm++&qs=n&form=QBRE&sp=-1&pq=tqdm+&sc=8-5&sk=&cvid=E8728F33AD3B481CAADB3E6D82D9038B | tqdm - 国内版 Bing
https://blog.csdn.net/qq_27825451/article/details/95486373 | (18条消息) python进度条神器之tqdm包的使用_MIss-Y的博客-CSDN博客_python tqdm包
https://blog.csdn.net/qq_33472765/article/details/82940843 | (18条消息) Python的Tqdm模块——进度条配置_家有代码初写成的博客-CSDN博客
https://blog.csdn.net/cui_yonghua/article/details/106816605 | (18条消息) python中tqdm(进度条模块) 详细用法及参数解释_崔永华的博客-CSDN博客_tqdm 参数
https://www.cnblogs.com/huma/p/12198386.html | Python实现进度条的4种方式 - 胡马 - 博客园
https://zhuanlan.zhihu.com/p/391045703 | yolov5软剪枝(二)：SFP和FPGM算法解析 - 知乎
https://cn.bing.com/search?q=%e8%bd%a6%e9%81%93%e7%ba%bf%e6%a3%80%e6%b5%8bgithub+2021&qs=n&sp=-1&pq=%e8%bd%a6%e9%81%93%e7%ba%bf%e6%a3%80%e6%b5%8bgithub+2021&sc=0-16&sk=&cvid=A56A2F0D44114844B727064BB0080227&first=11&FORM=PORE | 车道线检测github 2021 - 国内版 Bing
https://www.auto-testing.net/news/show-110300.html | 用于自动驾驶的实时车道线检测和智能告警_汽车技术__汽车测试网
https://ai.51cto.com/art/202107/674149.htm | 速度最快 250fps！实时、高性能车道线检测算法 LaneATT - 51CTO.COM
https://zhuanlan.zhihu.com/p/348493677 | WACV 2021 论文大盘点-智能驾驶篇 - 知乎
https://zhuanlan.zhihu.com/p/391747139 | 车道线检测调研 - 知乎
https://zhuanlan.zhihu.com/p/53431081 | GitHub：车道线检测最全资料集锦 - 知乎
https://github.com/liuchenyang33/laneDection | liuchenyang33/laneDection: 👍车道线检测
edge://newtab/ | 新建标签页
https://zhuanlan.zhihu.com/p/208440934 | 超快的车道线检测TensorRT部署: 300fps - 知乎
https://manaai.cn/index3.html | 神力AI(MANA)-国内最大的AI代码平台
https://www.baidu.com/s?wd=UltrFastLane&rsv_spt=1&rsv_iqid=0xd1fe92cc00035e56&issp=1&f=8&rsv_bp=1&rsv_idx=2&ie=utf-8&tn=baiduhome_pg&rsv_enter=0&rsv_dl=tb&rsv_n=2&rsv_sug3=1&rsv_btype=i&inputT=1826&rsv_sug4=1826 | UltrFastLane_百度搜索
https://zhuanlan.zhihu.com/p/157530787 | 超快的车道线检测 - 知乎
https://github.com/cfzd/Ultra-Fast-Lane-Detection | cfzd/Ultra-Fast-Lane-Detection: Ultra Fast Structure-aware Deep Lane Detection (ECCV 2020)

https://cn.bing.com/search?q=gbdt%2BLR&qs=n&form=QBRE&sp=-1&pq=g%27b%27d%27t%2Blr&sc=0-10&sk=&cvid=4F20787F17014D4EA973629926168C0A | gbdt+LR - 国内版 Bing
https://zhuanlan.zhihu.com/p/269578091 | 推荐系统学习笔记-5.GBDT+LR - 知乎
https://cn.bing.com/search?q=IoU-aware&cvid=fd97ac5832bc4e51b53a3cc170db904f&aqs=edge..69i57j0l4.1547j0j1&pglt=43&FORM=ANNTA1&PC=HCTS | IoU-aware - 国内版 Bing
https://blog.csdn.net/moxibingdao/article/details/106667487 | (17条消息) IoU-aware的目标检测，显著提高定位精度_我爱计算机视觉-CSDN博客

https://github.com/datawhalechina/fun-rec/blob/master/codes/GBDT_LR.py | fun-rec/GBDT_LR.py at master · datawhalechina/fun-rec
https://cn.bing.com/search?q=%e6%b7%b1%e5%ba%a6%e5%ad%a6%e4%b9%a0+ctr+github&qs=n&sp=-1&pq=%e6%b7%b1%e5%ba%a6%e5%ad%a6%e4%b9%a0+ctr+github&sc=0-15&sk=&cvid=977F8D495F8949849B21C1C09C7A13B2&first=11&FORM=PORE | 深度学习 ctr github - 国内版 Bing
https://blog.csdn.net/shenziheng1/article/details/89737430 | (22条消息) 深度学习CTR模型最全演化图谱 [王喆观点]_沈春旭的博客-CSDN博客
https://github.com/HanXiaoyang/CTR_NN | HanXiaoyang/CTR_NN: 基于深度学习的CTR预估，从FM推演各深度学习CTR预估模型（附代码）
https://github.com/DSXiangLi/CTR | DSXiangLi/CTR: CTR模型代码和学习笔记总结
https://github.com/liulin7576/DL_CTR | liulin7576/DL_CTR: 基于科大讯飞AI营销算法比赛实现CTR深度学习方法
https://daiwk.github.io/posts/dl-dl-ctr-models.html | 基于深度学习的ctr预估模型集合（持续更新） - daiwk-github博客
https://github.com/cdj0311/deep_learning_estimator | cdj0311/deep_learning_estimator: 深度学习在推荐算法的ctr预估等场景中的应用，该项目使用tensorflow estimator进行封装，包括dnn、deepfm、din、autoint等模型
https://github.com/Eadon999/DeepCTR | Eadon999/DeepCTR: Easy-to-use,Modular and Extendible package of deep-learning based CTR models .
https://blog.csdn.net/lc013/article/details/101444406 | (23条消息) 【原创】推荐广告入门：DeepCTR-Torch，基于深度学习的CTR预测算法库_AI 算法笔记-CSDN博客
https://blog.csdn.net/weixin_45366499/article/details/113844975 | (22条消息) 推荐系统之基于用户画像推荐_蔡政洁的博客-CSDN博客
https://zhuanlan.zhihu.com/p/355768708 | 用户画像系列（5）–个性化推荐算法 - 知乎
https://cloud.tencent.com/developer/article/1461126 | 基于深度学习的推荐系统 - 云+社区 - 腾讯云
https://zhuanlan.zhihu.com/p/358183268 | GitHub 访问不了？教你几招轻松解决 - 知乎
https://cn.bing.com/search?q=pytorch+DIEN&qs=n&form=QBRE&sp=-1&pq=pytorch+dien&sc=0-12&sk=&cvid=0874DB4C5BDE47C588BE07C928C4D5E0 | pytorch DIEN - 国内版 Bing
https://blog.csdn.net/hellozhxy/article/details/102501355 | (22条消息) DeepCTR-Torch，基于深度学习的CTR预测算法库_hellozhxy的博客-CSDN博客
https://deepctr-torch.readthedocs.io/en/latest/index.html | Welcome to DeepCTR-Torch’s documentation! — DeepCTR-Torch 0.2.7 documentation
https://zhuanlan.zhihu.com/p/50758485 | 详解阿里之Deep Interest Evolution Network(AAAI 2019) - 知乎
https://github.com/mouna99/dien | mouna99/dien
https://www.cnblogs.com/gongyanzh/p/12133984.html | 深度兴趣网络DIN-DIEN-DSIN - 鱼与鱼 - 博客园
https://blog.csdn.net/weixin_35154281/article/details/112275652 | (23条消息) 推荐系统之DIN代码详解_张先生－您好的博客-CSDN博客
https://cn.bing.com/search?q=gbdt%2Blr+%E6%A8%A1%E5%9E%8B%E6%80%8E%E4%B9%88%E4%BF%9D%E5%AD%98&qs=n&form=QBRE&sp=-1&pq=gbdt%2Blr+%E6%A8%A1%E5%9E%8B%E6%80%8E%E4%B9%88%E4%BF%9D%E5%AD%98&sc=0-14&sk=&cvid=C84DF7F0C79447918D513D4959FEA7A1 | gbdt+lr 模型怎么保存 - 国内版 Bing
https://blog.csdn.net/gamer_gyt/article/details/90183215 | (23条消息) Spark排序算法系列之（MLLib、ML）GBDT+LR使用方式介绍（模型训练、保存、加载、预测）_Thinkgamer博客-CSDN博客
https://zhuanlan.zhihu.com/p/114010125 | GBDT+LR简述+实例 - 知乎
https://www.cnblogs.com/wkang/p/9657032.html | GBDT+LR算法解析及Python实现 - Bo_hemian - 博客园

https://arxiv.org/pdf/2107.08430.pdf | 2107.08430.pdf
https://github.com/Megvii-BaseDetection/YOLOX | Megvii-BaseDetection/YOLOX: YOLOX is a high-performance anchor-free YOLO. Exceeding yolov3~v5 with ONNX, TensorRT, ncnn, and OpenVINO supported.
https://cn.bing.com/search?q=yoloX&cvid=5959457b412447eaa68b8238cf25e845&aqs=edge.0.69i59j69i60l3.4735j0j1&pglt=43&PC=HCTS&first=21&FORM=PORE | yoloX - 国内版 Bing
https://pypi.org/project/yolox/ | yolox · PyPI
https://cn.bing.com/search?q=Decoupled+head&cvid=2aad474c7b2045c6a3db2e5567b3c255&aqs=edge..69i57.1015j0j1&pglt=43&FORM=ANNTA1&PC=HCTS | Decoupled head - 国内版 Bing
https://zhuanlan.zhihu.com/p/269598144 | 实例分割新思路之SOLO v1&v2深度解析 - 知乎
https://cn.bing.com/search?q=anchor-based+%E5%AD%98%E5%9C%A8%E7%9A%84%E9%97%AE%E9%A2%98&cvid=9e2a8b0c425f482ba9765b3539f3f96a&aqs=edge..69i57.10043j0j1&pglt=43&FORM=ANNTA1&PC=HCTS | anchor-based 存在的问题 - 国内版 Bing
https://cloud.tencent.com/developer/article/1600160 | Anchor Free的目标检测方法–FCOS - 云+社区 - 腾讯云
https://blog.csdn.net/ytusdc/article/details/107864527 | (17条消息) Anchor-based 与 Anchor-free 优缺点_ytusdc的博客-CSDN博客
https://zhuanlan.zhihu.com/p/76710711 | 目标检测——anchor-based 与anchor-free - 知乎
https://cn.bing.com/search?q=label+assignment&qs=n&form=QBRE&sp=-1&pq=label+assignment&sc=0-16&sk=&cvid=008DD7AF541B43BFBE66924B093C47DA | label assignment - 国内版 Bing
https://zhuanlan.zhihu.com/p/265721299 | 《图解目标检测》系列之八：目标检测中的Label Assignment - 知乎

https://blog.csdn.net/DragonGirI/article/details/107434775 | (17条消息) Pytorch：Dataloader和Dataset以及搭建数据部分的步骤_龙雪之樱的博客-CSDN博客
https://pytorch.org/tutorials/beginner/basics/data_tutorial.html#:~:text=Dataset%20stores%20the%20samples%20and%20their%20corresponding%20labels%2C,and%20implement%20functions%20specific%20to%20the%20particular%20data. | Datasets & Dataloaders — PyTorch Tutorials 1.9.0+cu102 documentation
https://blog.csdn.net/guyuealian/article/details/88343924 | (16条消息) pytorch Dataset, DataLoader产生自定义的训练数据_pan_jinquan的博客-CSDN博客_dataset pytorch
https://blog.csdn.net/zw__chen/article/details/82806900 | (16条消息) Pytorch（五）入门：DataLoader 和 Dataset_嘿芝麻的树洞-CSDN博客
https://cn.bing.com/search?q=df+%E8%BD%AC+tensor&qs=n&form=QBRE&sp=-1&pq=df+%E8%BD%AC+tensor&sc=0-11&sk=&cvid=D4F0434DA4AF4211B96554F2CE21CADB | df 转 tensor - 国内版 Bing
https://blog.csdn.net/u011253318/article/details/109682137 | (17条消息) 把pandas.DataFrame转成torch.tensor的格式_u011253318的博客-CSDN博客
https://blog.csdn.net/xys430381_1/article/details/89640699 | (17条消息) 细粒度图像分类（FGVC）—综述_xys430381_1的专栏-CSDN博客_细粒度图像分类
https://zhuanlan.zhihu.com/p/23006190?refer=xiaoleimlnote | RCNN- 将CNN引入目标检测的开山之作 - 知乎
https://www.zhihu.com/people/xiao-lei-75-81/posts?page=2 | 大雄的机器梦 - 知乎
https://www.zhihu.com/column/xiaoleimlnote | 笔记 - 知乎
https://www.zhihu.com/topic/19596960/hot | 目标检测 - 知乎
https://fanyi.youdao.com/ | 在线翻译_有道
https://blog.csdn.net/qq_39056987/article/details/104327638 | (17条消息) YOLOv3-SPP介绍_那年当上博士前的博客-CSDN博客
https://cn.bing.com/search?q=pytorch+%E5%81%9A%E6%9C%BA%E5%99%A8%E5%AD%A6%E4%B9%A0+dataset+%E9%9D%9E%E5%9B%BE%E7%89%87&qs=n&form=QBRE&sp=-1&pq=pytorch+%E5%81%9A%E6%9C%BA%E5%99%A8%E5%AD%A6%E4%B9%A0+dataset+%E9%9D%9E%E5%9B%BE%E7%89%87&sc=0-25&sk=&cvid=D8B9863164384577B5C1DF8385653D2D | pytorch 做机器学习 dataset 非图片 - 国内版 Bing
https://blog.csdn.net/qq_43790749/article/details/114983366 | (17条消息) 使用pytorch加载非图像数据集方法总结_枭志的博客-CSDN博客
https://www.cnblogs.com/PythonLearner/p/13587125.html | 【小白学PyTorch】3 浅谈Dataset和Dataloader - 忽逢桃林 - 博客园
https://cn.bing.com/search?q=+gradcam+%E5%8F%AF%E4%BB%A5%E7%94%A8%E4%BA%8E%E7%9B%AE%E6%A0%87%E6%A3%80%E6%B5%8B%E5%90%97&qs=n&form=QBRE&sp=-1&pq=gradcam+%E5%8F%AF%E4%BB%A5%E7%94%A8%E4%BA%8E%E7%9B%AE%E6%A0%87%E6%A3%80%E6%B5%8B%E5%90%97&sc=0-17&sk=&cvid=0A15B8A2E8A6409C915518DB01092280 | gradcam 可以用于目标检测吗 - 国内版 Bing
https://cloud.tencent.com/developer/article/1624811 | 【必备】目标检测中的评价指标有哪些？ - 云+社区 - 腾讯云
https://zhuanlan.zhihu.com/p/264947208 | 盘点GAN在目标检测中的应用 - 知乎
https://cn.bing.com/search?q=pytorch+%E7%9B%AE%E6%A0%87%E6%A3%80%E6%B5%8B+%E5%BE%AE%E8%B0%83&qs=n&form=QBRE&sp=-1&pq=pytorch+%E7%9B%AE%E6%A0%87%E6%A3%80%E6%B5%8B+wei%27t&sc=0-18&sk=&cvid=D399A30510C140629D1C729438902363 | pytorch 目标检测微调 - 国内版 Bing
https://blog.csdn.net/zhjm07054115/article/details/104801165 | (17条消息) PyTorch TorchVision 目标检测网络微调_海尔兄弟的博客-CSDN博客
http://tech.sina.com.cn/csj/2020-05-17/doc-iirczymk2071933.shtml | PyTorch官方教程大更新：增加标签索引，更加新手友好_创事记_新浪科技_新浪网
https://zhuanlan.zhihu.com/p/96931265 | 目标检测三大开源神器：Detectron2/mmDetection/SimpleDet - 知乎
https://blog.csdn.net/weixin_41929524/article/details/107756275 | (17条消息) 可解释性深度学习工具Captum轻度体验_Kanny-CSDN博客_captum
https://blog.csdn.net/m0_46653437/article/details/113406227 | (17条消息) grad-cam用于SSD目标检测_敲代码的小风-CSDN博客
https://zhuanlan.zhihu.com/p/375637795 | CNN可视化又添新作，南大开源Group-CAM：高效的显著图生成方法 - 知乎
https://github.com/withtimesgo1115/gradCAM-YOLOv3-pytorch/blob/main/models.py | gradCAM-YOLOv3-pytorch/models.py at main · withtimesgo1115/gradCAM-YOLOv3-pytorch
https://cloud.tencent.com/developer/article/1846991 | 详细解读SSPNet| 小目标检测该如何进行改进？ - 云+社区 - 腾讯云
https://www.zhihu.com/question/280703314 | 目标检测领域还有什么可以做的？ - 知乎
https://www.zhihu.com/question/347517663 | GAN网络在目标检测领域的应用和发展？ - 知乎
https://www.zhihu.com/question/351254577 | 想知道目标检测领域中还有哪些方向能做？ - 知乎
https://zhuanlan.zhihu.com/p/269702192 | 万字长文：特征可视化技术(CAM) - 知乎
https://www.zhihu.com/question/372889525 | 如何对训练好的目标检测模型，输入图片进行Grad-CAM可视化？ - 知乎
https://github.com/yuyijie1995/Grad-CAM.pytorch#%E7%9B%AE%E6%A0%87%E6%A3%80%E6%B5%8B-faster-r-cnn | yuyijie1995/Grad-CAM.pytorch: pytorch实现Grad-CAM和Grad-CAM++,可以可视化任意分类网络的Class Activation Map (CAM)图,包括自定义的网络;同时也实现了目标检测faster r-cnn和retinanet两个网络的CAM图;欢迎试用、关注并反馈问题…

https://cn.bing.com/search?q=yolo+%E6%B7%B7%E6%B7%86%E7%9F%A9%E9%98%B5&cvid=286c663f8dbf4877a681a2d02924fe97&aqs=edge..69i57.7735j0j1&pglt=43&FORM=ANSPA1&PC=HCTS | yolo 混淆矩阵 - 国内版 Bing
https://blog.csdn.net/weixin_43760844/article/details/115208925 | (22条消息) 混淆矩阵：用于多分类模型评估（pytorch)_地中海の养成记-CSDN博客
https://cn.bing.com/search?q=torch+gpu+available&cvid=540a61b39a274e15977e2ee2f324233a&aqs=edge.1.0l2j69i57j0l2j69i60l2.3677j0j1&pglt=43&FORM=ANNTA1&PC=HCTS | torch gpu available - 国内版 Bing
https://cn.bing.com/search?q=github+%E5%94%90%E5%AE%87%E8%BF%AA&qs=n&form=QBRE&sp=-1&pq=github+%E5%94%90%E5%AE%87di&sc=0-11&sk=&cvid=CF58C28B4E314CCAB9C144FD2E847084 | github 唐宇迪 - 国内版 Bing
https://github.com/tangyudi/Ai-learn | tangyudi/Ai-Learn: 人工智能学习路线图，整理近200个实战案例与项目，免费提供配套教材，零基础入门，就业实战！包括：Python，数学，机器学习，数据分析，深度学习，计算机视觉，自然语言处理，PyTorch tensorflow machine-learning,deep-learning data-analysis data-mining mathematics data-science artificial-intelligence python tensorflow tensorflow2 caffe keras pytorch algorithm numpy pandas matplotlib seaborn nlp cv等热门领域
https://pytorch.org/get-started/locally/ | Start Locally | PyTorch
https://github.com/murufeng | murufeng (Hu Zhang)
https://cn.bing.com/search?q=markdown+%E5%8F%82%E8%80%83%E6%96%87%E7%8C%AE&cvid=49d16ee7989d4518b7b1e925e12dc9db&aqs=edge..69i57.6983j0j4&FORM=ANAB01&PC=HCTS | markdown 参考文献 - 国内版 Bing
https://jingyan.baidu.com/article/c843ea0b4b16ab37921e4a3c.html | 如何在markdown中添加文献引用上标（页内跳转）-百度经验
https://pan.baidu.com/s/1CYJrfNV-YyQP_34yKKHqBg#list/path=%2F | 程序员代码面试指南：IT名企算法与数据结构题目最优解（第2版）.pdf_免费高速下载|百度网盘-分享无限制

edge://newtab/ | 新建标签页

https://www.php.cn/tool/vscode/472011.html | 使用VS Code插件打造完美的MarkDown编辑器-VSCode-PHP中文网
https://gitee.com/zane_young/vscode-markdown-editor | vscode-markdown-editor: 一个 vscode 插件，让你的vscode 秒变和 typora 一样的所见即所得 markdown 编辑器，支持表格可视化编辑，拖拽图片，各种图标等
https://www.zhihu.com/question/423855665 | VScode写Markdown能否做到像Typora一样的即时渲染？ - 知乎
https://segmentfault.com/a/1190000022631755 | 美貌与强大并存——VScode，我的markdown码字利器(内附插件推荐) - SegmentFault 思否
https://zhuanlan.zhihu.com/p/113222681#:~:text=%E6%8C%89%E5%BF%AB%E6%8D%B7%E9%94%AE%E2%80%9CCtrl%2BShift%2BP%E2%80%9D%E3%80%82.%20%E5%9C%A8%E2%80%9Cvscode%E2%80%9D%E9%A1%B6%E9%83%A8%E4%BC%9A%E5%87%BA%E7%8E%B0%E4%B8%80%E4%B8%AA%E6%90%9C%E7%B4%A2%E6%A1%86%E3%80%82.%20%E8%BE%93%E5%85%A5%E2%80%9Cconfigure%20language%E2%80%9D%EF%BC%8C%E7%84%B6%E5%90%8E%E5%9B%9E%E8%BD%A6%E3%80%82.,%E2%80%9Cvscode%E2%80%9D%E9%87%8C%E9%9D%A2%E5%B0%B1%E4%BC%9A%E6%89%93%E5%BC%80%E4%B8%80%E4%B8%AA%E8%AF%AD%E8%A8%80%E9%85%8D%E7%BD%AE%E6%96%87%E4%BB%B6%E3%80%82.%20%E5%B0%86%E2%80%9Cen-us%E2%80%9D%E4%BF%AE%E6%94%B9%E6%88%90%E2%80%9Czh-cn%E2%80%9D%E3%80%82.%20%E6%8C%89%E2%80%9CCtrl%2BS%E2%80%9D%E4%BF%9D%E5%AD%98%E8%AE%BE%E7%BD%AE%E3%80%82.%20%E5%85%B3%E9%97%AD%E2%80%9Cvscode%E2%80%9D%EF%BC%8C%E5%86%8D%E6%AC%A1%E6%89%93%E5%BC%80%E5%B0%B1%E5%8F%AF%E4%BB%A5%E7%9C%8B%E5%88%B0%E4%B8%AD%E6%96%87%E7%95%8C%E9%9D%A2%E4%BA%86%E3%80%82.%20%E5%BD%93%E7%84%B6%E5%A6%82%E6%9E%9C%E4%BD%A0%E4%B8%8D%E6%84%BF%E6%84%8F%E8%AE%BE%E7%BD%AE%EF%BC%8C%E4%B9%9F%E5%8F%AF%E4%BB%A5%E7%9B%B4%E6%8E%A5%E5%AE%89%E8%A3%85%E5%AE%83%E7%9A%84%E4%B8%AD%E6%96%87%E6%8F%92%E4%BB%B6%EF%BC%8C%E8%BF%98%E6%98%AF%E5%BE%88%E4%BA%BA%E6%80%A7%E5%8C%96%E7%9A%84%E3%80%82. | 史上最全vscode配置使用教程 - 知乎
https://zhuanlan.zhihu.com/p/139140492 | 搭建 Markdown 舒适写作环境（VSCode，含扩展盘点） - 知乎

edge://newtab/ | 新建标签页

https://cn.bing.com/search?q=dockerfile&cvid=f43ec91e0f4b47268343b94afdfdb1e6&aqs=edge.0.0j69i57j69i59j0j69i60j69i61j69i60.5457j0j1&pglt=2083&FORM=ANNTA1&PC=HCTS | dockerfile - 国内版 Bing
https://www.jianshu.com/p/233dd2b13ce2 | Docker 镜像构建之 Dockerfile - 简书
https://cn.bing.com/search?q=docker+%E6%95%99%E7%A8%8B&cvid=635bed95fe0842668ee0e8eef1268635&aqs=edge.0.0j69i59j0l5.6193j0j1&pglt=2083&FORM=ANNTA1&PC=HCTS | docker 教程 - 国内版 Bing
https://yeasy.gitbook.io/docker_practice/image/build | 使用 Dockerfile 定制镜像 - Docker —— 从入门到实践
https://cn.bing.com/search?q=github+%E6%96%87%E4%BB%B6%E4%B8%8B%E8%BD%BD%E6%96%87%E4%BB%B6+%E5%91%BD%E4%BB%A4%E8%A1%8C&qs=n&form=QBRE&sp=-1&pq=github+%E6%96%87%E4%BB%B6%E4%B8%8B%E8%BD%BD%E6%96%87%E4%BB%B6+%E5%91%BD%E4%BB%A4%E8%A1%8C&sc=0-17&sk=&cvid=50B44C5848C643C48E6827FE95A1EBE8 | github 文件下载文件命令行 - 国内版 Bing
https://cn.bing.com/search?q=pip%E5%BC%BA%E5%88%B6%E6%9B%B4%E6%96%B0&qs=n&form=QBRE&sp=-1&pq=pip%E5%BC%BA%E5%88%B6%E6%9B%B4%E6%96%B0&sc=0-7&sk=&cvid=75472C41FC4D4814894FF87E189FCBF7 | pip强制更新 - 国内版 Bing
https://www.coder.work/article/94633 | python - pip安装选项 “ignore-installed”和 “force-reinstall”之间的区别 - IT工具网
https://www.coder.work/article/94633 | python - pip安装选项 “ignore-installed”和 “force-reinstall”之间的区别 - IT工具网
https://github.com/ultralytics/yolov5/blob/master/requirements.txt | yolov5/requirements.txt at master · ultralytics/yolov5
https://github.com/theme-next/hexo-theme-next/blob/master/layout/_partials/head/head.swig | hexo-theme-next/head.swig at master · theme-next/hexo-theme-next
https://blog.csdn.net/loocao/article/details/5536932 | Ubuntu 清除缓存 apt-get命令参数_ProgrammingRoad-CSDN博客
https://cn.bing.com/search?q=hexo+%E9%98%85%E8%AF%BB%E5%85%A8%E6%96%87&cvid=1d46e5b81b2d499c81f3462cac06f297&aqs=edge..69i57.9271j0j4&FORM=ANAB01&PC=HCTS | hexo 阅读全文 - 国内版 Bing
https://blog.csdn.net/zhu_1997/article/details/87869617 | hexo博客设置阅读全文_Mumu’s Blogs-CSDN博客_hexo 阅读全文
https://blog.csdn.net/mqdxiaoxiao/article/details/93263923 | Hexo博客NexT主题下显示摘要和阅读原文按钮_野猿新一-CSDN博客
https://sherlockgy.github.io/2018/06/03/Hexo%E6%90%AD%E5%BB%BA%E5%8D%9A%E5%AE%A2%E2%80%94%E2%80%94%E5%AE%9E%E7%8E%B0%E6%96%87%E7%AB%A0%E9%98%85%E8%AF%BB%E5%85%A8%E6%96%87/ | Hexo搭建博客——实现文章阅读全文 | SherlockGy’s Blog
https://blog.csdn.net/yueyue200830/article/details/104470646 | 设置hexo首页只显示部分摘要（不显示全文）_yueyue200830的博客-CSDN博客_hexo首页只显示标题
https://cn.bing.com/search?q=markdown+%E5%88%97%E8%A1%A8&cvid=39c02f29889e49e9b83e77cd60287658&aqs=edge..69i57j69i60.10151j0j4&FORM=ANAB01&PC=HCTS | markdown 列表 - 国内版 Bing
https://markdown.com.cn/basic-syntax/lists.html | Markdown 列表语法 | Markdown 语法教程
https://www.zabbx.cn/archives/docker%E4%BD%BF%E7%94%A8dockerfile%E6%9E%84%E5%BB%BAmysql%E9%95%9C%E5%83%8F%E5%8D%81%E4%BA%94 | Docker 使用Dockerfile构建MySQL镜像（十五） - 勤奋的冬枣

https://blog.csdn.net/weixin_43311978/article/details/106099052 | (22条消息) Hadoop3.2.1 HA 高可用集群的搭建（基于Zookeeper，NameNode高可用+Yarn高可用）_Captain.Y.的博客-CSDN博客
https://blog.csdn.net/weixin_43311978 | Captain.Y.的博客_CSDN博客-笔记领域博主
https://blog.csdn.net/weixin_43311978/article/details/105400694?spm=1001.2014.3001.5501 | (22条消息) CentOS7使用Docker安装hadoop集群_Captain.Y.的博客-CSDN博客
https://blog.csdn.net/weixin_43311978 | (22条消息) Captain.Y.的博客_CSDN博客-笔记领域博主
https://blog.csdn.net/piaoxi6587/article/details/103569376 | (22条消息) Hadoop3.2 +Spark3.0全分布式安装_piaoxi6587的博客-CSDN博客
https://blog.csdn.net/piaoxi6587 | (22条消息) piaoxi6587的博客_Jiaxing.Zhang_CSDN博客-领域博主
https://blog.csdn.net/piaoxi6587/article/details/103516600?spm=1001.2014.3001.5501 | (22条消息) 基于Hadoop3.2伪分布式安装Spark3.0_piaoxi6587的博客-CSDN博客
https://blog.csdn.net/zhangjunli/article/details/111308104 | (22条消息) hadoop-3.x、zookeeper-3.x、hbase-2.x、hive-3.x、sqoop1.x、spark3.x_zhangjunli的博客-CSDN博客
https://mirrors.tuna.tsinghua.edu.cn/apache/hive/stable-2/ | Index of /apache/hive/stable-2
https://blog.csdn.net/dsafds123/article/details/105932335 | (22条消息) Hadoop+zookeeper集群环境搭建步骤(百分百安装，亲测有效)_dsafds123的博客-CSDN博客
https://blog.csdn.net/xwyzsn/article/details/112338624 | (22条消息) hadoop +hbase+zookeeper 伪分布安装(超级无敌详细)_xwyzsn的博客-CSDN博客
https://blog.csdn.net/weixin_43487121/article/details/103589532 | (22条消息) Hadoop3.1.3+Zookeeper3.5.6+hbase2.2.2+hive3.1.2安装以及测试_井鱼的博客-CSDN博客
https://blog.csdn.net/weixin_43993764/article/details/113405025 | (22条消息) Docker配置Hadoop集群并使用WordCount测试_出大问题-CSDN博客
https://blog.csdn.net/Y_C_S_M/article/details/90731320 | Docker 从零搭建 Hadoop 3.2 集群 (镜像基于Ubuntu 14.04)_Y_C_S_M的博客-CSDN博客
https://cn.bing.com/search?q=hive+%E5%AE%8C%E5%85%A8%E5%88%86%E5%B8%83%E5%BC%8F&qs=n&form=QBRE&sp=-1&pq=hive+%E5%AE%8C%E5%85%A8%E5%88%86%E5%B8%83%E5%BC%8F&sc=0-10&sk=&cvid=F8D0F2158279442FB8264C47EA447C53 | hive 完全分布式 - 国内版 Bing
https://zhuanlan.zhihu.com/p/64858731 | Hbase完全分布式搭建及Hive整合 - 知乎
https://www.cnblogs.com/yinzhengjie2020/p/12239031.html | HBase完全分布式集群搭建 - JasonYin2020 - 博客园
https://www.cnblogs.com/rmxd/p/11316062.html#_label4_0 | HBase的完全分布式搭建 - coder、 - 博客园
https://www.jianshu.com/p/30d45fa044a2 | 3.Spark环境搭建-Spark完全分布式集群搭建 - 简书
https://blog.csdn.net/weixin_39030764/article/details/83032423 | (22条消息) Ubuntu18安装Scala（亲测）_迷彩侠的博客-CSDN博客_ubuntu安装scala
https://www.scala-lang.org/download/ | Download | The Scala Programming Language
https://scala-lang.org/download/scala2.html | Download Scala 2 | The Scala Programming Language
https://hadoop.apache.org/docs/r3.2.2/hadoop-project-dist/hadoop-common/ClusterSetup.html | Apache Hadoop 3.2.2 – Hadoop Cluster Setup
https://cn.bing.com/search?q=hadoop+%E9%9B%86%E7%BE%A4+hive+ubuntu&qs=n&form=QBRE&sp=-1&pq=hadoop+%E9%9B%86%E7%BE%A4+hive+ubuntu&sc=0-21&sk=&cvid=7F798F3ECA1745639068726F36F86E3E | hadoop 集群 hive ubuntu - 国内版 Bing
https://blog.csdn.net/chenyh_csdn/article/details/83509353 | (22条消息) ubuntu系统的mysql+hadoop+hive环境搭建_chenyh_csdn的博客-CSDN博客
https://blog.csdn.net/qq_40415753/article/details/106919223 | (22条消息) ubuntu搭建hadoop、mysql、hive环境_joker-G的博客-CSDN博客
https://zhuanlan.zhihu.com/p/89472385 | ubuntu下hadoop、spark、hive、azkaban 集群搭建 - 知乎
https://blog.csdn.net/weixin_43931044/article/details/104903746 | (22条消息) Ubuntu+Hadoop+Mysql+Hive+Sqoop_weixin_43931044的博客-CSDN博客
https://www.cnblogs.com/opsprobe/p/9126864.html | Ubuntu18.04下安装MySQL - OpsDrip - 博客园
https://blog.csdn.net/baidu_36414377/article/details/85554824 | (22条消息) Hadoop集群+Hive+MySQL搭建_baidu_36414377的博客-CSDN博客
https://blog.csdn.net/v3424/article/details/109784131 | (22条消息) Hive集群（3台虚拟机）搭建和交互方式基于Hadoop_v3424的博客-CSDN博客
https://www.cnblogs.com/kinginme/p/7249533.html | Hive集成mysql数据库 - Kinginme - 博客园
https://zhuanlan.zhihu.com/p/141233011 | hive及mysql的安装（最细） - 知乎
https://blog.csdn.net/agent_x/article/details/78660341 | (22条消息) hive安装及mysql配置_炼剑-CSDN博客_hive配置mysql
https://blog.csdn.net/szxiaohe/article/details/76650266 | (22条消息) Linux -bash: java: command not found 解决方法_格子的博客-CSDN博客
https://blog.edkso.cn/?p=469 | 十下博客 - Hadoop集群规划和配置以及启动
https://cn.bing.com/search?q=ubuntu+%E5%AE%89%E8%A3%85mysql5.7&qs=n&form=QBRE&sp=-1&pq=ubuntu+%E5%AE%89%E8%A3%85mysql5.7&sc=2-17&sk=&cvid=52A4F8E3DA964C00A5FC8EA7FE5C8821 | ubuntu 安装mysql5.7 - 国内版 Bing
https://www.jianshu.com/p/ba48f1e386f0 | Ubuntu20.04 安装 MySQL5.7 - 简书
https://cn.bing.com/search?q=%2Fetc%2Finit.d%2Fmysql%3A+line+70%3A+%2Flib%2Fapparmor%2Fprofile-load%3A+No+such+file+or+directory+..&qs=n&form=QBRE&sp=-1&pq=%2Fetc%2Finit.d%2Fmysql%3A+line+70%3A+%2Flib%2Fapparmor%2Fprofile-load%3A+no+such+file+or+directory+..&sc=0-84&sk=&cvid=A113130B95BD4F12BDCBE9C5D606AC18 | /etc/init.d/mysql: line 70: /lib/apparmor/profile-load: No such file or directory .. - 国内版 Bing
https://ubuntu.pkgs.org/18.04/ubuntu-universe-amd64/libmecab2_0.996-5_amd64.deb.html | libmecab2_0.996-5_amd64.deb 18.04 LTS Download
https://blog.csdn.net/qq_39504519/article/details/115486186 | (22条消息) Ubuntu18.04安装MySQL5.7_、Lu的博客-CSDN博客
https://www.jb51.net/article/202399.htm | Ubuntu 20.04 安装和配置MySql5.7的详细教程_Mysql_脚本之家
https://www.cnblogs.com/xuwujing/p/8322022.html | 大数据学习系列之九—- Hive整合Spark和HBase以及相关测试 - 虚无境 - 博客园
https://cloud.tencent.com/developer/article/1624245 | Spark on Hive & Hive on Spark，傻傻分不清楚 - 云+社区 - 腾讯云
https://www.jianshu.com/p/75b440c9b904 | Apache Spark整合Hive - 简书
https://www.jianshu.com/p/83b2f9cb3d33 | sparkSql和hive的整合（spark on hive） - 简书
https://www.cnblogs.com/xuwujing/p/8045821.html | 大数据学习系列之四 —– Hadoop+Hive环境搭建图文详解(单机) - 虚无境 - 博客园
https://cn.bing.com/search?q=hive+%E6%B2%A1%E6%9C%89%E9%9B%86%E7%BE%A4&qs=n&form=QBRE&sp=-1&pq=hive+%E6%B2%A1%E6%9C%89%E9%9B%86%E7%BE%A4&sc=0-9&sk=&cvid=E07D360724B34199BD355A86F1804154 | hive 没有集群 - 国内版 Bing
https://www.cnblogs.com/zlslch/p/6700695.html | Hive环境的安装部署（完美安装）（集群内或集群外都适用）（含卸载自带mysql安装指定版本） - 大数据和AI躺过的坑 - 博客园
https://blog.csdn.net/Allenzyg/article/details/105762896 | (22条消息) Hive集群安装（3个节点）-基于Hadoop_Allenzyg的博客-CSDN博客
https://cn.bing.com/search?q=FAILED%3A+SemanticException+org.apache.hadoop.hive.ql.metadata.HiveException%3A+java.lang.RuntimeException%3A+Unable+to+instantiate+org.apache.hadoop.hive.ql.metadata.SessionHiveMetaStoreClient&cvid=4e7e75c762f04707b2811e7bb59ee39b&aqs=edge..69i57j69i58.436j0j4&FORM=ANAB01&PC=HCTS | FAILED: SemanticException org.apache.hadoop.hive.ql.metadata.HiveException: java.lang.RuntimeException: Unable to instantiate org.apache.hadoop.hive.ql.metadata.SessionHiveMetaStoreClient - 国内版 Bing
https://blog.csdn.net/qq_35078688/article/details/86137440 | (22条消息) FAILED: SemanticException org.apache.hadoop.hive.ql.metadata.HiveException: java.lang.RuntimeExcepti_大数据-CSDN博客
https://www.jianshu.com/p/776802db315e | Hive环境搭建 - 简书
https://blog.csdn.net/GentleCP/article/details/87936263 | (22条消息) Mysql 8.0.13 开启远程访问权限（ERROR 1064 (42000): You have an error in your SQL syntax; check the manual th）_GentleCP的博客-CSDN博客
https://blog.csdn.net/qq_41943867/article/details/90574135 | (22条消息) JDBC连接数据库出现Loading class `com.mysql.jdbc.Driver’. This is deprecated.问题的相关解决方法_一只小菜鸡的blog-CSDN博客
https://www.cnblogs.com/EnzoDin/p/7376707.html | hive启动时报错 java.lang.IllegalArgumentException: java.net.URISyntaxException: Relative path in absolute URI: ${system:java.io.tmpdir%7D/$%7Bsystem:user.name%7D at org.apache.hadoop.fs.Path.initialize - 嘣嘣嚓 - 博客园
https://www.cnblogs.com/syq816/p/12632028.html | Hive启动报错：java.lang.NoSuchMethodError: com.google.common.base.Preconditions.checkArgument - 狗子的进阶史 - 博客园
https://blog.csdn.net/Milan__Kundera/article/details/81182757 | (22条消息) mysql-connector-java各种版本下载地址_Bruce_Json的博客-CSDN博客_mysql-connector-java-
https://mvnrepository.com/artifact/mysql/mysql-connector-java/8.0.25 | Maven Repository: mysql » mysql-connector-java » 8.0.25
https://blog.csdn.net/zhangbinlong/article/details/109777234 | (22条消息) mysql8 提示 ERROR 1410 (42000): You are not allowed to create a user with GRANT__binlong的博客-CSDN博客
https://cn.bing.com/search?q=mysql8&cvid=4bff9c3d3b2c4b5e87fa6dfe39c25890&aqs=edge..69i57j0l6.2070j0j4&FORM=ANAB01&PC=HCTS | mysql8 - 国内版 Bing
https://dev.mysql.com/doc/refman/8.0/en/grant.html | MySQL :: MySQL 8.0 Reference Manual :: 13.7.1.6 GRANT Statement
https://www.cnblogs.com/lucas-zhao/p/12012401.html | 报错“bash: jps: command not found” - Lucas_zhao - 博客园
https://blog.csdn.net/qq_34864753/article/details/95063774 | (22条消息) linux jps命令报错-bash: jps: command not found 解决方案_me凡的博客-CSDN博客
https://blog.csdn.net/shilu963/article/details/87966912 | (22条消息) 搭建hadoop全分布遇到的问题_shilu963的博客-CSDN博客
https://blog.csdn.net/appleyuchi/article/details/114965509 | (22条消息) Desktop: ERROR: Cannot set priority of namenode process 35122_微电子学与固体电子学-CSDN博客
https://blog.csdn.net/agent_x/article/details/78660341 | (22条消息) hive安装及mysql配置_炼剑-CSDN博客_hive配置mysql
https://cn.bing.com/search?q=torch+%E5%A4%9A%E7%B2%BE%E5%BA%A6&qs=n&form=QBRE&sp=-1&pq=torch+duo%27jing%27du&sc=0-17&sk=&cvid=89FFE4E43B58420A8CC43B206334D066 | torch 多精度 - 国内版 Bing
https://zhuanlan.zhihu.com/p/158375055 | PyTorch 21.单机多卡操作(分布式DataParallel，混合精度，Horovod) - 知乎
https://zhuanlan.zhihu.com/p/113694038 | PyTorch分布式训练简明教程 - 知乎
https://www.jianshu.com/p/ea8b705f0915 | Ubuntu如何卸载mysql - 简书
https://blog.csdn.net/qq_41238579/article/details/102778243 | (22条消息) Ubuntu修改文件（文件夹）的拥有者和用户组_哈哈拉拉hxl的博客-CSDN博客
https://cn.bing.com/search?q=WARNING%3A+HADOOP_SECURE_DN_USER+has+been+replaced+by+HDFS_DATANODE_SECURE_USER.+Using+value+of+HADOOP_SECURE_DN_USER&cvid=3b2923cbb83d413e997f0087d9a2f8de&aqs=edge..69i57j69i58.953j0j4&FORM=ANAB01&PC=HCTS | WARNING: HADOOP_SECURE_DN_USER has been replaced by HDFS_DATANODE_SECURE_USER. Using value of HADOOP_SECURE_DN_USER - 国内版 Bing
https://cn.bing.com/search?q=ERROR%3a+Cannot+set+priority+of+datanode+process&cvid=b5e2c41f61dc4c27b1b593918bda97c9&aqs=edge..69i57j69i59l2j69i58j69i60l3.1026j0j4&PC=HCTS&first=11&FORM=PORE | ERROR: Cannot set priority of datanode process - 国内版 Bing
https://cn.bing.com/search?q=ERROR%3A+Cannot+set+priority+of+journalnode+process+5355&qs=n&form=QBRE&sp=-1&pq=error%3A+cannot+set+priority+of+journalnode+process+5355&sc=0-54&sk=&cvid=53CF6BF63F4B49FC9D880353C276CFDB | ERROR: Cannot set priority of journalnode process 5355 - 国内版 Bing
https://blog.csdn.net/yanghuadong_1992/article/details/113622249 | (22条消息) hadoop入门启动 journalnode报错_yanghuadong的博客-CSDN博客
https://www.cnblogs.com/FireLL/p/13755169.html | (oﾟvﾟ)ノ Hi - WARNING: HADOOP_SECURE_DN_USER has been replaced by HDFS_DATANODE_SECURE_USER. Using value of HADOOP_SECURE_DN_USER.
https://www.cnblogs.com/19930521zhang/p/14617926.html | Hadoop 3.3.0问题解决汇总 - 随风去远方 - 博客园
https://www.cnblogs.com/jackchen001/p/6398423.html | hadoop集群异常问题总结 - jackchen007 - 博客园
https://blog.csdn.net/zhaogang1993/article/details/92727952 | (22条消息) hadoop入门1：ERROR Cannot set priority of datanode process_小钢炮-CSDN博客
https://www.codeleading.com/article/72031400447/ | 记Hadoop3.1.2安装排错之ERROR: Cannot set priority of datanode process - 代码先锋网
https://hu1104.github.io/2021/06/28/Hadoop_WSL2/ | Hadoop_WSL2 | Hexo

https://www.cnblogs.com/brooksj/p/10765377.html | 多机同步管理hexo博客 - 编程匠心者 - 博客园
https://www.cnblogs.com/liziczh/p/9318670.html | Hexo | （四）多机同步更新博客 - 陈栗 - 博客园
https://www.longger.net/article/15012.html | Hexo多台电脑同步 | 网络运维
https://yeasy.gitbook.io/docker_practice/image/dockerfile/entrypoint | ENTRYPOINT 入口点 - Docker —— 从入门到实践
https://www.jianshu.com/p/592ff5ee3b62 | 深入理解 Docker 构建上下文 - 简书
https://zhuanlan.zhihu.com/p/30555962 | Dockerfile: ENTRYPOINT和CMD的区别 - 知乎
https://www.cnblogs.com/sddai/p/11032879.html#:~:text=%E2%80%9Cdocker%20run%E2%80%9D%E5%92%8C%E2%80%9Cdocker%20exec%E2%80%9D%E9%83%BD%E6%98%AF%20Docker,%E5%AE%B9%E5%99%A8%E4%B8%AD%E7%94%A8%E4%BA%8E%E6%89%A7%E8%A1%8C%E7%9A%84%E5%91%BD%E4%BB%A4%E3%80%82%20%E7%84%B6%E8%80%8C%EF%BC%8C%E5%9C%A8%E4%B8%8D%E5%90%8C%E7%9A%84%E6%83%85%E5%86%B5%E4%B8%8B%EF%BC%8C%E5%AE%83%E4%BB%AC%E7%9A%84%E4%BD%BF%E7%94%A8%E6%9C%89%E7%9D%80%E6%9C%AC%E8%B4%A8%E4%B8%8A%E7%9A%84%E5%8C%BA%E5%88%AB%E3%80%82%20%E2%80%9Cdocker%20run%E2%80%9D%E9%80%9A%E5%B8%B8%E6%98%AF%E5%9C%A8%E6%96%B0%E5%88%9B%E5%BB%BA%E7%9A%84%E5%AE%B9%E5%99%A8%E4%B8%AD%E6%89%80%E4%BD%BF%E7%94%A8%E7%9A%84%E5%91%BD%E4%BB%A4%E3%80%82%20%E5%AE%83%E9%80%82%E7%94%A8%E4%BA%8E%E5%9C%A8%E6%B2%A1%E6%9C%89%E5%85%B6%E4%BB%96%E5%AE%B9%E5%99%A8%E8%BF%90%E8%A1%8C%E7%9A%84%E6%83%85%E5%86%B5%E4%B8%8B%EF%BC%8C%E6%82%A8%E6%83%B3%E8%A6%81%E5%88%9B%E5%BB%BA%E4%B8%80%E4%B8%AA%E5%AE%B9%E5%99%A8%EF%BC%8C%E5%B9%B6%E4%B8%94%E8%A6%81%E5%90%AF%E5%8A%A8%E5%AE%83%EF%BC%8C%E7%84%B6%E5%90%8E%E5%9C%A8%E5%85%B6%E4%B8%8A%E8%BF%90%E8%A1%8C%E4%B8%80%E4%B8%AA%E8%BF%9B%E7%A8%8B%E3%80%82 | docker run VS docker exec 的区别 - stardsd - 博客园
https://cn.bing.com/search?q=yolov5&cvid=abbb10f9d89b4c7f9029dae0e7ce7084&aqs=edge.0.69i59l3j69i57j69i60l3.2412j0j4&FORM=ANAB01&PC=HCTS | yolov5 - 国内版 Bing
https://github.com/ultralytics/yolov5 | ultralytics/yolov5: YOLOv5 in PyTorch > ONNX > CoreML > TFLite
https://github.com/ultralytics/yolov5/wiki/Tips-for-Best-Training-Results | Tips for Best Training Results · ultralytics/yolov5 Wiki
https://cn.bing.com/search?q=%e5%8f%ac%e5%9b%9e%e7%8e%87%e5%81%8f%e4%bd%8e&qs=n&sp=-1&pq=%e5%8f%ac%e5%9b%9e%e7%8e%87%e5%81%8f%e4%bd%8e&sc=0-5&sk=&cvid=BB32A9A72CD74E8594307B6DAEB8F525&first=11&FORM=PORE | 召回率偏低 - 国内版 Bing
https://blog.csdn.net/weixin_39122088/article/details/109340318 | yolov5如何找出召回率以及map较低的原因（或者说找出难以预测的图片）一只很菜很菜的tfer的博客-CSDN博客
https://cn.bing.com/search?q=%E6%9C%BA%E5%99%A8%E5%AD%A6%E4%B9%A0+%E6%8C%87%E6%A0%87+micro&qs=n&form=QBRE&sp=-1&pq=%E6%9C%BA%E5%99%A8%E5%AD%A6%E4%B9%A0+%E6%8C%87%E6%A0%87+micro&sc=0-13&sk=&cvid=396867F0B24B4DF49813029FC3FA317A | 机器学习指标 micro - 国内版 Bing
https://www.cnblogs.com/laozhanghahaha/p/12499979.html | 机器学习–Micro Average,Macro Average, Weighted Average - 老张哈哈哈 - 博客园
https://blog.csdn.net/qq_29598161/article/details/105081369 | 推荐的准确度评价指标AP，MAP，AUC，NDCG_linkstack的博客-CSDN博客
https://blog.csdn.net/xys430381_1/article/details/90770520 | 机器学习基础—分类与检测的评价指标—AP，mAP，PRC曲线_xys430381_1的专栏-CSDN博客_机器学习ap
https://zhuanlan.zhihu.com/p/93107394#:~:text=%E6%9C%BA%E5%99%A8%E5%AD%A6%E4%B9%A0%E6%9C%89%E5%BE%88%E5%A4%9A%E8%AF%84%E4%BC%B0%E7%9A%84%E6%8C%87%E6%A0%87%E3%80%82.%20%E6%9C%89%E4%BA%86%E8%BF%99%E4%BA%9B%E6%8C%87%E6%A0%87%E6%88%91%E4%BB%AC%E5%B0%B1%E6%A8%AA%E5%90%91%E7%9A%84%E6%AF%94%E8%BE%83%E5%93%AA%E4%BA%9B%E6%A8%A1%E5%9E%8B%E7%9A%84%E8%A1%A8%E7%8E%B0%E6%9B%B4%E5%A5%BD%E3%80%82.%20%E6%88%91%E4%BB%AC%E5%85%88%E4%BB%8E%E6%95%B4%E4%BD%93%E4%B8%8A%E6%9D%A5%E7%9C%8B%E7%9C%8B%E4%B8%BB%E6%B5%81%E7%9A%84%E8%AF%84%E4%BC%B0%E6%8C%87%E6%A0%87%E9%83%BD%E6%9C%89%E5%93%AA%E4%BA%9B%EF%BC%9A.%20%E5%88%86%E7%B1%BB%E9%97%AE%E9%A2%98%E8%AF%84%E4%BC%B0%E6%8C%87%E6%A0%87%EF%BC%9A.%20%E5%87%86%E7%A1%AE%E7%8E%87%20%E2%80%93,Accuracy.%20%E7%B2%BE%E7%A1%AE%E7%8E%87%EF%BC%88%E5%B7%AE%E5%87%86%E7%8E%87%EF%BC%89-%20Precision.%20%E5%8F%AC%E5%9B%9E%E7%8E%87%EF%BC%88%E6%9F%A5%E5%85%A8%E7%8E%87%EF%BC%89-%20Recall.%20F1%E5%88%86%E6%95%B0.%20ROC%E6%9B%B2%E7%BA%BF. | 一文看懂机器学习指标：准确率、精准率、召回率、F1、ROC曲线、AUC曲线 - 知乎
https://blog.csdn.net/ruyingcai666666/article/details/109670567 | mAP@0.5与mAP@0.5:0.95的含义，YOLO_ruyingcai666666的博客-CSDN博客_map0.5:0.95
https://blog.csdn.net/qq_36756866/article/details/109111065 | YOLOv5训练自己的数据集（超详细完整版）_深度学习菜鸟的博客-CSDN博客_yolov5训练自己的数据集
https://blog.csdn.net/a_cheng/article/details/111401500 | YOLOV5训练自己的数据集（踩坑经验之谈）_a_cheng_的博客-CSDN博客_yolov5训练很慢
https://blog.csdn.net/weixin_48994268/article/details/115282688 | yolov5训练自己的数据集（一文搞定训练）_小屋-CSDN博客
https://blog.csdn.net/qq_44703886/article/details/112801975 | (22条消息) 【YOLO】YOLOv5训练自己的数据集_摇曳的树的博客-CSDN博客
https://pan.baidu.com/s/1vlTmjNofB5kD3BOaSy4SnA#list/path=%2Fsharelink3882671082-403192674953674%2Fyolov5%E6%9D%83%E9%87%8D&parentPath=%2Fsharelink3882671082-403192674953674 | yolov5权重_免费高速下载|百度网盘-分享无限制
https://github.com/wudashuo/yolov5 | wudashuo/yolov5: YOLOv5 汉化版，保持官方同步更新
https://blog.51cto.com/u_15194128/2761816 | YoloV5实战：手把手教物体检测——YoloV5_AI小浩的技术博客_51CTO博客
https://zhuanlan.zhihu.com/p/29393415 | COCO数据集的标注格式 - 知乎
https://www.jianshu.com/p/34b3312b71db | [图像算法]-yolov5全解-一文通关 - 简书
https://blog.csdn.net/mary_0830/article/details/107081030 | (22条消息) YOLOv5代码详解（test.py部分）_Liaojiajia-Blog-CSDN博客
https://blog.csdn.net/mary_0830/article/details/107081030 | YOLOv5代码详解（test.py部分）_Liaojiajia-Blog-CSDN博客
https://blog.csdn.net/mary_0830/article/details/107081030 | (22条消息) YOLOv5代码详解（test.py部分）_Liaojiajia-Blog-CSDN博客
https://zhuanlan.zhihu.com/p/56961620 | 详解object detection中的mAP - 知乎
https://blog.csdn.net/tpz789/article/details/110675268 | (22条消息) yolov5训练自己的数据集并计算mAP_tpz789的博客-CSDN博客
https://cn.bing.com/search?q=yolov5+%E6%A8%A1%E5%9E%8B%E4%B8%8B%E8%BD%BD&qs=n&form=QBRE&sp=-1&pq=yolov5+%E6%A8%A1%E5%9E%8Bx%27z&sc=0-12&sk=&cvid=FE3495801DE745F0BEFF5E3E82144CDF | yolov5 模型下载 - 国内版 Bing
https://cn.bing.com/search?q=json+dump+dumps+loads+load%E5%8C%BA%E5%88%AB&qs=n&form=QBRE&sp=-1&pq=json+dump+dumps+loads+loadqu%27b&sc=0-30&sk=&cvid=039CD071253049048757BCD80C02CBB5 | json dump dumps loads load区别 - 国内版 Bing
https://blog.csdn.net/ybw_2569/article/details/99839529 | (22条消息) Json中dump 和dumps，load和loads的区别_BeefpasteC的博客-CSDN博客_json load和dump
https://cloud.tencent.com/developer/article/1503898 | python中json.load()、json.loads()、json.dump()、json.dumps()的区别 - 云+社区 - 腾讯云
https://blog.csdn.net/yuyexiaohan/article/details/84970653 | (22条消息) json.dumps()、json.loads()、json.dump()、json.load()方法讲解_yuyexiaohan的博客-CSDN博客
https://blog.csdn.net/mr_evanchen/article/details/77879967 | (22条消息) 【Python】Json模块dumps、loads、dump、load函数介绍_Mr_EvanChen的Blog-CSDN博客_json.dumps
https://cn.bing.com/search?q=dataloader.dataset&qs=n&form=QBRE&sp=-1&pq=dataloader.datase&sc=0-17&sk=&cvid=8942E2F3A2854FE99F93CC9F78FA1025 | dataloader.dataset - 国内版 Bing
https://www.cnblogs.com/sigai/p/8074329.html | [pathlib]内置pathlib库的常用属性和方法 - [sigai] - 博客园
https://cn.bing.com/search?q=yolov3&qs=n&form=QBRE&sp=-1&pq=yolov3&sc=8-6&sk=&cvid=FBD07757690B4D6687BA2912FE76D980 | yolov3 - 国内版 Bing
https://www.jianshu.com/p/043966013dde | YOLOv3详解 - 简书
https://zhuanlan.zhihu.com/p/172121380 | 深入浅出Yolo系列之Yolov5核心基础知识完整讲解 - 知乎
https://blog.csdn.net/csdn1e/article/details/114637487 | (22条消息) yolov5使用笔记_greatwind的博客-CSDN博客
https://www.bing.com/search?q=.+%3A+%E6%97%A0%E6%B3%95%E5%8A%A0%E8%BD%BD%E6%96%87%E4%BB%B6+C%3A%5CUsers%5Czongxing%5CDocuments%5CWindowsPowerShell%5Cprofile.ps1%EF%BC%8C%E5%9B%A0%E4%B8%BA%E5%9C%A8%E6%AD%A4%E7%B3%BB%E7%BB%9F%E4%B8%8A%E7%A6%81%E6%AD%A2%E8%BF%90%E8%A1%8C%E8%84%9A%E6%9C%AC%E3%80%82%E6%9C%89%E5%85%B3%E8%AF%A6%E7%BB%86%E4%BF%A1%E6%81%AF%EF%BC%8C+%E8%AF%B7%E5%8F%82%E9%98%85+https%3A%2Fgo.microsoft.com%2Ffwlink%2F%3FLinkID%3D135170+%E4%B8%AD%E7%9A%84+about_Execution_Policies%E3%80%82+%E6%89%80%E5%9C%A8%E4%BD%8D%E7%BD%AE+%E8%A1%8C%3A1+%E5%AD%97%E7%AC%A6%3A+3+%2B+.+%27C%3A%5CUsers%5Czongxing%5CDocuments%5CWindowsPowerShell%5Cprofile.ps1%27+%2B+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~+%2B+CategoryInfo+%3A+SecurityError%3A+(%3A)+%5B%5D%EF%BC%8CPSSecurityException+%2B+FullyQualifiedErrorId+%3A+UnauthorizedAccess&cvid=b21084e284b3482380228480e902270b&aqs=edge..69i57.959j0j4&pglt=43&FORM=ANAB01&PC=HCTS | . : 无法加载文件 C:\Users\zongxing\Documents\WindowsPowerShell\profile.ps1，因为在此系统上禁止运行脚本。有关详细信息，请参阅 https:/go.microsoft.com/fwlink/?LinkID=135170 中的 about_Execution_Policies。所在位置行:1 字符: 3 + . ‘C:\Users\zongxing\Documents\WindowsPowerShell\profile.ps1’ + ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + CategoryInfo : SecurityError: (:) []，PSSecurityException + FullyQualifiedErrorId : UnauthorizedAccess - 国内版 Bing
https://cn.bing.com/search?q=%3A+Sub-process+%2Fusr%2Fbin%2Fdpkg+returned+an+error+code+(1&cvid=8041d187fa5d46e084066dd5a39e755f&aqs=edge..69i57.1300j0j4&FORM=ANAB01&PC=HCTS | : Sub-process /usr/bin/dpkg returned an error code (1 - 国内版 Bing
https://blog.csdn.net/u013832707/article/details/113104006 | (22条消息) 解决 dpkg 安装出错后的 Sub-process /usr/bin/dpkg returned an error code (1) 错误_持久决心的博客-CSDN博客

https://yeasy.gitbook.io/docker_practice/container | 操作容器 - Docker —— 从入门到实践
https://blog.csdn.net/weixin_43487121/article/details/103589532 | (22条消息) Hadoop3.1.3+Zookeeper3.5.6+hbase2.2.2+hive3.1.2安装以及测试_井鱼的博客-CSDN博客
https://blog.csdn.net/shanhai3000/article/details/107682499 | (22条消息) 安装并配置HBase集群（5个节点）_shanhai3000的博客-CSDN博客
https://blog.csdn.net/qq_35516745/article/details/103822597 | (22条消息) ImportError: libGL.so.1: cannot open shared object file: No such file or directory——docker容器内问题报错_Atsea0107的博客-CSDN博客
https://blog.csdn.net/weixin_43311978/article/details/106099052 | (22条消息) Hadoop3.2.1 HA 高可用集群的搭建（基于Zookeeper，NameNode高可用+Yarn高可用）_Captain.Y.的博客-CSDN博客
https://www.zabbx.cn/archives/docker%E4%BD%BF%E7%94%A8dockerfile%E6%9E%84%E5%BB%BAmysql%E9%95%9C%E5%83%8F%E5%8D%81%E4%BA%94 | Docker 使用Dockerfile构建MySQL镜像（十五） - 勤奋的冬枣

edge://newtab/ | 新建标签页

edge://newtab/ | 新建标签页

https://cn.bing.com/search?q=cv2.adaptivethreshold&cvid=e73e33a617174b55b732ac0068658d88&aqs=edge.1.69i59i450l7.481951024j0j1&FORM=ANAB01&PC=HCTS | cv2.adaptivethreshold - 国内版 Bing
https://blog.csdn.net/weixin_35732969/article/details/83756628 | opencv-python自适应阈值二值化函数cv2.adaptiveThreshold使用及效果_do it-CSDN博客
https://cn.bing.com/search?q=opencv%E4%BA%8C%E5%80%BC%E5%8C%96%E5%8F%96%E5%8F%8D&qs=n&form=QBRE&sp=-1&pq=opencv%E4%BA%8C%E5%80%BC%E5%8C%96%E5%8F%96%E5%8F%8D&sc=0-11&sk=&cvid=12F6AFADB4B8407E9F342A8BFEA89658 | opencv二值化取反 - 国内版 Bing
https://www.zhangshengrong.com/p/zD1yD9Y2Xr/ | Python-OpenCV教程之图像的位运算详解 / 张生荣
https://www.jb51.net/article/215462.htm | Python-OpenCV教程之图像的位运算详解_python_脚本之家
https://blog.csdn.net/qq_30241709/article/details/78759726 | OpenCV_提取直线、轮廓和区域_birdguan-CSDN博客_opencv 直线提取
https://cn.bing.com/search?q=cv2.inRange&qs=n&form=QBRE&sp=-1&pq=cv2.inrange&sc=8-11&sk=&cvid=278A82A05B24493CBE20D803B22D63ED | cv2.inRange - 国内版 Bing
https://cn.bing.com/search?q=hsv&qs=n&form=QBRE&sp=-1&pq=hsv&sc=9-3&sk=&cvid=4BF48C2B94D9434D8CE0256B57BC1AF7 | hsv - 国内版 Bing
https://cn.bing.com/search?q=%E7%BA%A2%E8%89%B2%2B%E8%93%9D%E8%89%B2+%E6%98%AF%E4%BB%80%E4%B9%88%E9%A2%9C%E8%89%B2&qs=n&form=QBRE&sp=-1&pq=%E7%BA%A2%E8%89%B2%2B%E8%93%9D%E8%89%B2+%E6%98%AF%E4%BB%80%E4%B9%88%E9%A2%9C%E8%89%B2&sc=0-11&sk=&cvid=F1B127178663446894C6CCD13530F931 | 红色+蓝色是什么颜色 - 国内版 Bing
https://cbgc.scol.com.cn/news/1282515 | 三基色（红绿蓝）混合出来的颜色是合成了单色光还是一种错觉？ - 川观新闻
https://zhuanlan.zhihu.com/p/60578464#:~:text=%E4%BD%BF%E7%94%A8%20Hexo%2BGitHub%20%E6%90%AD%E5%BB%BA%E4%B8%AA%E4%BA%BA%E5%85%8D%E8%B4%B9%E5%8D%9A%E5%AE%A2%E6%95%99%E7%A8%8B%EF%BC%88%E5%B0%8F%E7%99%BD%E5%90%91%EF%BC%89%201%20%E5%87%86%E5%A4%87%202%20%E8%BF%9E%E6%8E%A5%20Github....,Hexo%20%E5%88%B0%20GitHub%20Pages%206%20%E7%BB%91%E5%AE%9A%E5%9F%9F%E5%90%8D%EF%BC%88%E5%8F%AF%E9%80%89%EF%BC%89....%207%20%E5%BC%80%E5%A7%8B%E4%BD%BF%E7%94%A8 | 使用 Hexo+GitHub 搭建个人免费博客教程（小白向） - 知乎
https://cn.bing.com/search?q=%e8%bf%81%e7%a7%bb%e5%ad%a6%e4%b9%a0&qs=AS&pq=%e8%bf%81%e7%a7%bb&sc=7-2&cvid=2C420C8CDE80497498E53BA177256815&FORM=QBRE&sp=1 | 迁移学习 - 国内版 Bing
https://blog.csdn.net/huang_nansen/article/details/102793744 | 【opencv】HSV颜色值对照表_huang_nansen的博客-CSDN博客_hsv颜色对照表
https://blog.csdn.net/u013921430/article/details/84532068 | 【图像处理】——双边滤波_不用先生的博客-CSDN博客_双边滤波
https://blog.csdn.net/weixin_41115751/article/details/84568029 | python_opencv图像算术运算（cv2.bitwise_and(),cv2.bitwise_not()）按位运算_C++是世界上最好的语言，手动狗头-CSDN博客
https://blog.csdn.net/Hren0412/article/details/97623740 | cv2.bitwise_and() 函数_Hren0412的博客-CSDN博客_cv2.bitwise_and
https://www.jianshu.com/p/200526c9a109 | openCV-python（四）图像操作——调节对比度和亮度 - 简书
https://www.cnblogs.com/lfri/p/10753019.html#:~:text=Python-OpenCV%E2%80%94%E2%80%94%E4%BA%AE%E5%BA%A6%E5%92%8C%E5%AF%B9%E6%AF%94%E5%BA%A6.%20%E4%BA%AE%E5%BA%A6%E4%B8%8E%E5%AF%B9%E6%AF%94%E5%BA%A6.%20%E4%BA%AE%E5%BA%A6%E8%B0%83%E6%95%B4%E6%98%AF%E5%B0%86%E5%9B%BE%E5%83%8F%E5%83%8F%E7%B4%A0%E7%9A%84%E5%BC%BA%E5%BA%A6%E6%95%B4%E4%BD%93%E5%8F%98%E5%A4%A7%2F%E5%8F%98%E5%B0%8F%EF%BC%8C%E5%AF%B9%E6%AF%94%E5%BA%A6%E8%B0%83%E6%95%B4%E6%8C%87%E7%9A%84%E6%98%AF%E5%9B%BE%E5%83%8F%E6%9A%97%E5%A4%84%E5%8F%98%E5%BE%97%E6%9B%B4%E6%9A%97%EF%BC%8C%E4%BA%AE%E5%87%BA%E5%8F%98%E5%BE%97%E6%9B%B4%E4%BA%AE%EF%BC%8C%E4%BB%8E%E8%80%8C%E6%8B%93%E5%AE%BD%E6%9F%90%E4%B8%AA%E5%8C%BA%E5%9F%9F%E5%86%85%E7%9A%84%E6%98%BE%E7%A4%BA%E7%B2%BE%E5%BA%A6%E3%80%82.%20OpenCV%E4%B8%AD%E4%BA%AE%E5%BA%A6%E5%92%8C%E5%AF%B9%E6%AF%94%E5%BA%A6%E5%BA%94%E7%94%A8%E8%BF%99%E4%B8%AA%E5%85%AC%E5%BC%8F%E6%9D%A5%E8%AE%A1%E7%AE%97%EF%BC%9Ag%28x%29,%3D%20%CE%B1f%28x%29%20%2B%20%CE%B2%EF%BC%8C%E5%85%B6%E4%B8%AD%EF%BC%9A%CE%B1%28%3E0%29%E3%80%81%CE%B2%E5%B8%B8%E7%A7%B0%E4%B8%BA%E5%A2%9E%E7%9B%8A%E4%B8%8E%E5%81%8F%E7%BD%AE%E5%80%BC%EF%BC%8C%E5%88%86%E5%88%AB%E6%8E%A7%E5%88%B6%E5%9B%BE%E7%89%87%E7%9A%84%E5%AF%B9%E6%AF%94%E5%BA%A6%E5%92%8C%E4%BA%AE%E5%BA%A6%E3%80%82.%20%E6%B3%A8%EF%BC%9A%E6%AD%A4%E5%A4%84%E5%AF%B9%CE%B1%2F%CE%B2%E6%8E%A7%E5%88%B6%E5%AF%B9%E6%AF%94%E5%BA%A6%E5%92%8C%E4%BA%AE%E5%BA%A6%E6%9C%89%E4%BA%89%E8%AE%AE%EF%BC%8C%E5%85%B7%E4%BD%93%E8%AF%B7%E5%8F%82%E8%80%83%EF%BC%9AOpenCV%E5%85%B3%E4%BA%8E%E5%AF%B9%E6%AF%94%E5%BA%A6%E5%92%8C%E4%BA%AE%E5%BA%A6%E7%9A%84%E8%AF%AF%E8%A7%A3. | Python-OpenCV——亮度和对比度 - Rogn - 博客园
https://lil-q.github.io/blog/opencv-python%E5%AD%A6%E4%B9%A0%E7%AC%94%E8%AE%B0%E5%9B%9B/ | OpenCV-python：亮度与对比度 | Homeward
https://blog.csdn.net/ivan_9/article/details/113572880 | Opencv图像的亮度和对比度调整_ivan_9的博客-CSDN博客
https://cn.bing.com/search?q=np.histogram&go=%E6%90%9C%E7%B4%A2&qs=ds&form=QBRE | np.histogram - 国内版 Bing
https://blog.csdn.net/hankobe1/article/details/104931766 | np.histogram用法_hankobe1的博客-CSDN博客
http://localhost:8888/tree | Home Page - Select or create a notebook
http://localhost:8888/notebooks/Untitled.ipynb | Untitled - Jupyter Notebook
https://www.baidu.com/s?wd=%E6%8E%A8%E8%8D%90%E7%B3%BB%E7%BB%9F%20github&rsv_spt=1&rsv_iqid=0xe3a775be0052bceb&issp=1&f=8&rsv_bp=1&rsv_idx=2&ie=utf-8&rqlang=cn&tn=baiduhome_pg&rsv_enter=1&rsv_dl=tb&rsv_t=c9a8vT79Qv0nhh303Cgmpd5%2FwXxxOLTxTtMalxZW%2F055poSAN01PMjTZYQOZNImCz2vM&oq=youtube%2520%25E6%258E%25A8%25E8%258D%2590%25E7%25B3%25BB%25E7%25BB%259F%2520github&rsv_btype=t&inputT=274&rsv_pq=a07409ea001019f7&rsv_sug3=48&rsv_sug1=63&rsv_sug7=100&rsv_sug2=0&rsv_sug4=3218 | 推荐系统 github_百度搜索
https://blog.csdn.net/algorithmPro/article/details/110944165 | Github星标超3k的推荐系统入门资料合集_机器学习算法那些事-CSDN博客
https://github.com/hu1104/RSPapers | hu1104/RSPapers
https://github.com/qcymkxyc/RecSys | qcymkxyc/RecSys: 项亮的《推荐系统实践》的代码实现
https://nbviewer.jupyter.org/github/qcymkxyc/RecSys/blob/master/Recommend%20System.ipynb | Jupyter Notebook Viewer
https://blog.csdn.net/DASlab/article/details/48598095?locationNum=12&fps=1 | YouTube推荐系统_DASlab的博客-CSDN博客
https://zhuanlan.zhihu.com/p/114703091 | 深入理解YouTube推荐系统算法 - 知乎
https://zhuanlan.zhihu.com/p/99953120 | YouTube推荐系统算法梳理 - 知乎
https://zhuanlan.zhihu.com/p/138777815 | 经典！YouTube推荐系统-年度最佳Paper(附实践代码) - 知乎
https://github.com/datawhalechina/fun-rec/blob/master/codes/FM.py | fun-rec/FM.py at master · datawhalechina/fun-rec
https://www.bing.com/search?q=%E6%8E%A8%E8%8D%90%E7%B3%BB%E7%BB%9F+%E7%89%B9%E5%BE%81%E5%B7%A5%E7%A8%8B&go=%E6%90%9C%E7%B4%A2&qs=ds&form=QBRE | 推荐系统特征工程 - Bing
https://zhuanlan.zhihu.com/p/189960926 | 推荐系统（6）–特征工程 - 知乎
http://www.woshipm.com/pd/480396.html | 推荐系统老司机的十条经验 | 人人都是产品经理
https://www.sohu.com/a/205279542_610300 | 干货分享 | 详解特征工程与推荐系统及其实践（附PPT）
https://www.jianshu.com/p/73adce3fd975 | GBDT+LR CTR预估-Kaggle实例[有数据集] - 简书
https://zhuanlan.zhihu.com/p/143763320 | 推荐系统 embedding 技术实践总结 - 知乎
https://www.cnblogs.com/asxinyu/articles/4372699.html | 关于推荐系统中的特征工程 - 数据之巅 - 博客园
https://www.zhihu.com/question/19711823 | 有哪些好用的开源推荐系统？ - 知乎
https://zhuanlan.zhihu.com/p/52504407 | YouTube深度学习推荐系统的十大工程问题 - 知乎
https://www.bing.com/search?q=%E5%9F%BA%E4%BA%8E%E7%94%A8%E6%88%B7%E7%9A%84%E6%8E%A8%E8%8D%90%E7%B3%BB%E7%BB%9F&cvid=f2bcaf9f8e564cccb07d4a36a178233d&aqs=edge..69i57.6295j0j9&FORM=ANAB01&PC=HCTS | 基于用户的推荐系统 - Bing
https://blog.csdn.net/u012050154/article/details/52268057 | 基于用户的协同过滤算法的电影推荐系统_行者小朱的博客-CSDN博客
https://github.com/Shicoder/Deep_Rec | Shicoder/Deep_Rec: 存放推荐算法相关代码、文档、资料
https://www.bing.com/search?q=%E6%8E%A8%E8%8D%90%E7%B3%BB%E7%BB%9F+%E7%89%B9%E5%BE%81%E5%B7%A5%E7%A8%8B&cvid=aec7911b8fb347b38367345dde81dcb8&aqs=edge..69i57.10872j0j1&pglt=2083&FORM=ANNTA1&PC=HCTS | 推荐系统特征工程 - Bing
https://zhuanlan.zhihu.com/p/99259582 | 推荐系统之数据与特征工程 - 知乎
https://www.bing.com/search?q=github&cvid=807461369b2d4c7880583a9daed10c79&aqs=edge.0.69i59j0j69i60l5.2920j0j1&pglt=2083&FORM=ANNTA1&PC=HCTS | github - Bing
https://github.com/ogerhsou/Youtube-Recommendation-Tensorflow/blob/master/youtube_recommendation.py | Youtube-Recommendation-Tensorflow/youtube_recommendation.py at master · ogerhsou/Youtube-Recommendation-Tensorflow
https://www.baidu.com/s?ie=utf-8&f=8&rsv_bp=1&rsv_idx=2&tn=baiduhome_pg&wd=%E6%8E%A8%E8%8D%90%E7%B3%BB%E7%BB%9F%20%E7%89%B9%E5%BE%81%E5%B7%A5%E7%A8%8B&rsv_spt=1&oq=%25E7%259F%25A5%25E8%25AF%2586%25E5%259B%25BE%25E8%25B0%25B1&rsv_pq=d130c4c700012671&rsv_t=9e60NUt1gs7pLu0C93QAE%2B6cEA5KvlTjVnBQaxzoORZL1uGkHSyqCekZ223h72BcUcSA&rqlang=cn&rsv_enter=1&rsv_dl=tb&rsv_sug3=20&rsv_sug1=11&rsv_sug7=100&rsv_sug2=0&rsv_btype=t&inputT=7907&rsv_sug4=7907 | 推荐系统特征工程_百度搜索
https://www.cnblogs.com/x739400043/p/12394135.html | 推荐系统的特征工程 - 小小小的程序猿 - 博客园
https://zhuanlan.zhihu.com/p/99259582 | 推荐系统之数据与特征工程 - 知乎

动态规划

发表于 2022-06-09 更新于 2022-07-11
本文字数： 0 阅读时长 ≈ 1 分钟

代码规范

发表于 2022-03-14 更新于 2022-07-11
本文字数： 1.2k 阅读时长 ≈ 1 分钟

代码应当易于理解

关键思想：代码应当易于理解

关键思想：代码的写法应当使别人理解它所需的时间最小化

表面层次的改进

把信息装到名字里

选择专业的词

get 与 fetch

size() 与 NumNodes()、MemoryBytes()

stop() 与 Kill()、Pause()、Resume()

选用更有表现力的单词：

避免泛泛的名字

因为懒惰而使用：tmp retval result

result 与 userList orderList cateList

循环迭代时的 i j k 变成更有意思的 cateLength xxxLen

用具体的名字代替抽象的名字

在给变量、函数或其它元素命名时，要把它描述得更具体而不是更抽象。

DISALLOW_EVIL_CONSTRUCTORS

DISALLOW_COPY_AND_ASSIGN

使用前缀或后缀来给名字附带更多信息

如果变量是一个度量，最好把名字带上单位

附带其它重要信息：

决定名字的长度

BEManager与BackEndManager

小的作用域可心使用短名字

经验原则：团队新成员是否能理解这个名字的含义？

doc document eval evaluation

丢掉没用的词：

convertToString() 与 ToString()

long2Str

利用名字的格式来表达含义

这个比较成熟，遵照规范和团队约定即可

不会误解的名字

关键思想：名字不要有歧义

objList.filter(“age > 18”);

objList.leaveAdults(“age>18”);

给布尔值命名

userIsAuthentiated

与使用者的期望相匹配

审美

来比较一下下面的代码：

用换行来保持一致和紧凑

用方法来整理不规则的东西

在需要时使用列对齐

把代码分段落

使用一致的风格

该写什么样的注释

不要为了注释而注释

不要给不好的名字加注释——应该把名字改好

记录你的思想

加入“导演评论”

为代码中的瑕疵写注释

给常量加注释

意料中的提问

公布可能的陷阱

“全局观”注释

最后的思考-克服“作者心理阻滞”

写出言简意赅的注释

简化循环和逻辑

把控制流变得易读

条件语句中的参数顺序

if/else语句块的顺序

?:条件表达式

避免do/while循环

从函数中提前返回

最小化嵌套

通过提前返回来减少嵌套

减少循环内的嵌套

总结：

拆分超长表达式

用作解释的变量

总结变量

使用德摩根定理

找到更优雅的方式

拆分巨大的语句

优雅的例子：

变量与可读性

减少变量

没有价值的临时变量

减少中间结果

减少控制流变量

缩小变量的作用域

只写一次的变量更好

重新组织代码

抽取不相关的子问题

纯工具代码

其它用途代码

意料之外的好处

创建大量通用代码

项目专有的功能

简化已有接口

一次只做一件事

任务可以很小

应用“一次只做一件事情”原则

把想法变成代码

清楚的描述逻辑

用自然语言描述解决文案

少写代码

质疑和拆分你的需求

删除没用的代码

熟悉你周边的库

优秀博文

发表于 2022-01-19 更新于 2022-07-11
本文字数： 82 阅读时长 ≈ 1 分钟

yolo#

博文链接	重点
你一定从未看过如此通俗易懂的YOLO系列(从v1到v5)模型解读 (上) - 知乎 (zhihu.com)	生动形象的将yolo系列从0到1的思考过程展现

激活函数

发表于 2022-01-13 更新于 2022-07-11
本文字数： 1.4k 阅读时长 ≈ 1 分钟

1、恒等函数#

$$
f(x)=x \ \ \ \ \ \ \ f^{‘}(x) =1
$$

2. 单位阶跃函数#

$$
f(x)=\begin{cases}
0, & \text x<0\
1, & \text x \geq 0
\end{cases}
\ \ \ \ \ \ \ \
f(x)=\begin{cases}
0, & \text x \neq 0\
?, & \text x = 0
\end{cases}
$$

逻辑函数#

$$
f(x)=\sigma (x)=\frac{1}{1+e^{-x}}\ \ \ \ \ \ \ \ \
f^{‘}(x) = f(x)(1-f(x))
$$

range:(0,1),存在梯度消失的风险，原函数满足单调性，一阶导数不满足；

优点：平滑，易于求导

缺点：梯度消失；不是关于原点对称；exp耗时

双曲正切函数#

$$
f(x)=tanh(x)=\frac{e^x - e^{-x}}{e^x + e^{-x}}\ \ \ \ \ \ \ \
f^{‘}(x)=1-f(x)^2
$$

range:(-1,1);——解决sigmoid不是关于原点对称问题，但是梯度消失（缓解依然存在）；

RELU#

修正线性单元函数
$$
f(x)=max(0,x)\ \ \ \ \ \ \
f(x)=\begin{cases}
0 & \text x<0\
1 &\text x\geq 0
\end{cases}
$$

解决梯度消失，但是存在梯度爆炸风险；因为没有exp运算，运算快于sigmoid和tanh；收敛因梯度值将快于sigmoid和tanh；但是输入如果为负数时，则将输出总是0，导致神经元不被激活。

Swish#

$$
swish(x)=x \cdot \sigma(\beta x)
$$

其中$\beta$ 是个常数或者可训练的参数。

Swish激活函数具备无上界有下届、平滑、非单调的特性，Swish在深层模型上效果优于ReLU。

hard-swish#

$$
h-swish(x)=x\frac{ReLU6(x+3)}{6}
$$

该激活函数在MobileNetV3论文中提出，相较于swish函数，具有数值稳定性好，计算速度快等优点。

嵌入式时，计算sigmoid函数代价太大

在论文中，作者提到，虽然这种Swish非线性提高了精度，但是在嵌入式环境中，他的成本是非零的，因为在移动设备上计算sigmoid函数代价要大得多。

因此作者使用hard-Swish和hard-Sigmoid替换了ReLU6和SE-block中的Sigmoid层，但是只是在网络的后半段才将ReLU6替换为h-Swish，因为作者发现Swish函数只有在更深的网络层使用才能体现其优势。

首先是肯定了Swish的重要性，然后指出在量化模式下，Sigmoid函数比ReLU6的计算代价大的多，所以才有了这个ReLU6版本的h-Swish。

ReLU6#

抑制最大值

$$
ReLu6(x) = min(6, max(0, x))
$$

主要是为了在移动端float16的低精度的时候，也能有很好的数值分辨率，如果对ReLu的输出值不加限制，那么输出范围就是0到正无穷，而低精度的float16无法精确描述其数值，带来精度损失。

softplus#

$$
f(x)=ln(1+e^x)\ \ \ \ \ \ \ \
f^{‘}(x)=\frac {e^x}{1+e^x}
$$

Mish#

$$
Mish(x) = x \cdot tanh(ln(1+e^x))
$$

DIEN

发表于 2021-11-08 更新于 2022-07-11
本文字数： 3.6k 阅读时长 ≈ 3 分钟

1. DIEN提出的动机
2. DIEN模型原理
- 2.1 Interest Exterator Layer
- 2.2 Interest Evolving Layer

1. DIEN提出的动机

在推荐场景，用户无需输入搜索关键词来表达意图，这种情况下捕捉用户兴趣并考虑兴趣的动态变化将是提升模型效果的关键。以Wide&Deep为代表的深度模型更多的是考虑不同field特征之间的相互作用，未关注用户兴趣。

DIN模型考虑了用户兴趣，并且强调用户兴趣是多样的，该模型使用注意力机制来捕捉和target item的相关的兴趣，这样以来用户的兴趣就会随着目标商品自适应的改变。但是大多该类模型包括DIN在内，直接将用户的行为当做用户的兴趣(因为DIN模型只是在行为序列上做了简单的特征处理)，但是用户潜在兴趣一般很难直接通过用户的行为直接表示，大多模型都没有挖掘用户行为背后真实的兴趣，捕捉用户兴趣的动态变化对用户兴趣的表示非常重要。DIEN相比于之前的模型，即对用户的兴趣进行建模，又对建模出来的用户兴趣继续建模得到用户的兴趣变化过程。

阅读全文 »

DIN

发表于 2021-10-29 更新于 2022-07-11
本文字数： 9.2k 阅读时长 ≈ 8 分钟

1. 动机
2. DIN模型结构及原理
3.实现

1. 动机

Deep Interest Network(DIN)是2018年阿里巴巴提出来的模型，该模型基于业务的观察，从实际应用的角度进行改进，相比于之前很多“学术风”的深度模型，该模型更加具有业务气息。该模型的应用场景是阿里巴巴的电商广告推荐业务，这样的场景下一般会有大量的用户历史行为信息，这个其实是很关键的，因为DIN模型的创新点或者解决的问题就是使用了注意力机制来对用户的兴趣动态模拟，而这个模拟过程存在的前提就是用户之前有大量的历史行为了，这样我们在预测某个商品广告用户是否点击的时候，就可以参考他之前购买过或者查看过的商品，这样就能猜测出用户的大致兴趣来，这样我们的推荐才能做的更加到位，所以这个模型的使用场景是非常注重用户的历史行为特征（历史购买过的商品或者类别信息），也希望通过这一点，能够和前面的一些深度学习模型对比一下。

在个性化的电商广告推荐业务场景中，也正式由于用户留下了大量的历史交互行为，才更加看出了之前的深度学习模型(作者统称Embeding&MLP模型)的不足之处。如果学习了前面的各种深度学习模型，就会发现Embeding&MLP模型对于这种推荐任务一般有着差不多的固定处理套路，就是大量稀疏特征先经过embedding层，转成低维稠密的，然后进行拼接，最后喂入到多层神经网络中去。

这些模型在这种个性化广告点击预测任务中存在的问题就是无法表达用户广泛的兴趣，因为这些模型在得到各个特征的embedding之后，就蛮力拼接了，然后就各种交叉等。这时候根本没有考虑之前用户历史行为商品具体是什么，究竟用户历史行为中的哪个会对当前的点击预测带来积极的作用。而实际上，对于用户点不点击当前的商品广告，很大程度上是依赖于他的历史行为的，王喆老师举了个例子

假设广告中的商品是键盘，如果用户历史点击的商品中有化妆品，包包，衣服，洗面奶等商品，那么大概率上该用户可能是对键盘不感兴趣的，而如果用户历史行为中的商品有鼠标，电脑，iPad，手机等，那么大概率该用户对键盘是感兴趣的，而如果用户历史商品中有鼠标，化妆品， T-shirt和洗面奶，鼠标这个商品embedding对预测“键盘”广告的点击率的重要程度应该大于后面的那三个。

这里也就是说如果是之前的那些深度学习模型，是没法很好的去表达出用户这广泛多样的兴趣的，如果想表达的准确些，那么就得加大隐向量的维度，让每个特征的信息更加丰富，那这样带来的问题就是计算量上去了，毕竟真实情景尤其是电商广告推荐的场景，特征维度的规模是非常大的。并且根据上面的例子，也并不是用户所有的历史行为特征都会对某个商品广告点击预测起到作用。所以对于当前某个商品广告的点击预测任务，没必要考虑之前所有的用户历史行为。

这样， DIN的动机就出来了，在业务的角度，我们应该自适应的去捕捉用户的兴趣变化，这样才能较为准确的实施广告推荐；而放到模型的角度，我们应该考虑到用户的历史行为商品与当前商品广告的一个关联性，如果用户历史商品中很多与当前商品关联，那么说明该商品可能符合用户的品味，就把该广告推荐给他。而一谈到关联性的话，我们就容易想到“注意力”的思想了，所以为了更好的从用户的历史行为中学习到与当前商品广告的关联性，学习到用户的兴趣变化，作者把注意力引入到了模型，设计了一个”local activation unit”结构，利用候选商品和历史问题商品之间的相关性计算出权重，这个就代表了对于当前商品广告的预测，用户历史行为的各个商品的重要程度大小，而加入了注意力权重的深度学习网络，就是这次的主角DIN，下面具体来看下该模型。

阅读全文 »

AFM

发表于 2021-10-29 更新于 2022-07-11
本文字数： 2.9k 阅读时长 ≈ 3 分钟

1. 动机
AFM模型原理

1. 动机

AFM的全称是Attentional Factorization Machines, 从模型的名称上来看是在FM的基础上加上了注意力机制，FM是通过特征隐向量的内积来对交叉特征进行建模，从公式中可以看出所有的交叉特征都具有相同的权重也就是1，没有考虑到不同的交叉特征的重要性程度： $$ y_{fm} = w_0+\sum_{i=1}^nw_ix_i+\sum_{i=1}^{n}\sum_{i+1}^n\lt v_i,v_j\gt x_ix_j $$ 如何让不同的交叉特征具有不同的重要性就是AFM核心的贡献，在谈论AFM交叉特征注意力之前，对于FM交叉特征部分的改进还有FFM，其是考虑到了对于不同的其他特征，某个指定特征的隐向量应该是不同的（相比于FM对于所有的特征只有一个隐向量，FFM对于一个特征有多个不同的隐向量）。

阅读全文 »

目标检测损失函数

发表于 2021-10-29 更新于 2022-07-11
本文字数： 3k 阅读时长 ≈ 3 分钟

Smooth L1 Loss
IoU Loss（2016）
GIOU loss(2019)
DIoU Loss（2019）
CIoU Loss（2019）
EIoU Loss（2021）

Smooth L1 Loss

【动机】 Smooth L1 Loss完美的避开了L1和L2 Loss的缺点

L1 Loss的问题：损失函数对x的导数为常数，在训练后期，x很小时，如果learning rate 不变，损失函数会在稳定值附近波动，很难收敛到更高的精度。
L2 Loss的问题：损失函数对x的导数在x值很大时，其导数也非常大，在训练初期不稳定。
阅读全文 »

DCN

发表于 2021-10-29 更新于 2022-07-11
本文字数： 3.4k 阅读时长 ≈ 3 分钟

1. 动机
2. 模型结构与原理
- 2.1 Embedding和Stacking 层
- 2.2 Cross Network

1. 动机

Wide&Deep模型的提出不仅综合了“记忆能力”和“泛化能力”，而且开启了不同网络结构融合的新思路。所以后面就有各式各样的模型改进Wide部分或者Deep部分，而Deep&Cross模型(DCN)就是其中比较典型的一个，这是2017年斯坦福大学和谷歌的研究人员在ADKDD会议上提出的，该模型针对W&D的wide部分进行了改进，因为Wide部分有一个不足就是需要人工进行特征的组合筛选，过程繁琐且需要经验，而2阶的FM模型在线性的时间复杂度中自动进行特征交互，但是这些特征交互的表现能力并不够，并且随着阶数的上升，模型复杂度会大幅度提高。于是乎，作者用一个Cross Network替换掉了Wide部分，来自动进行特征之间的交叉，并且网络的时间和空间复杂度都是线性的。通过与Deep部分相结合，构成了深度交叉网络（Deep & Cross Network），简称DCN。

阅读全文 »