Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
:notebook: DeepLearning and CV notes.
| Date | Stars |
|---|---|
| 2026-07-24 | 353 |
| 2026-07-25 | 353 |
| 2026-07-28 | 353 |
| 2026-07-30 | 353 |
| 2026-08-06 | 353 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
<a name="top"></a> [toc] 深度学习、计算机视觉学习过程。 # 写在前面 人工智能最新学术研究和技术实现追寻可关注: - [量子位 - 知乎 - 专栏](https://zhuanlan.zhihu.com/qbitai) - [机器之心 - 知乎 - 专栏](https://zhuanlan.zhihu.com/jiqizhixin) - [新智元 - 知乎](<https://zhuanlan.zhihu.com/aiera>) - [计算机视觉论文速递 - 知乎 - 专栏](https://zhuanlan.zhihu.com/c_172507674) - [PaperWeekly - 知乎 - 专栏](https://zhuanlan.zhihu.com/paperweekly) - [计算机视觉life - 知乎 - 专栏](https://zhuanlan.zhihu.com/c_150246914) - 相关网站: - [专知 - 专知](http://www.zhuanzhi.ai) - [AI研习社](<https://ai.yanxishe.com/>) - [极市开发者社区](<http://bbs.cvmart.net/>) - [PaperWeekly](https://www.paperweekly.site/home) - [论文投稿 - 学术交流区 - 小木虫论坛-学术科研互动平台](http://muchong.com/f-125-1) 学习经验参考: - [关于学习机器学习算法的一些建议(忠告) - 知乎](<https://zhuanlan.zhihu.com/p/69104881>) - [深度学习的一些经验总结和建议| To do v.s Not To Do](https://mp.weixin.qq.com/s/5WpdUrv8caCMYSEEt5hS9w) AI Conference Deadlines:<https://aideadlin.es/?sub=ML,CV,NLP,RO,SP,DM> 比赛: - Kaggle官网:https://www.kaggle.com/ - 天池AI开发者社区:https://tianchi.aliyun.com/ # 一、基础学习 ## 1.1 数学基础 - [标准差和方差](<https://www.shuxuele.com/data/standard-deviation.html>) [荐]★★★ - [理解梯度下降](<http://liuchengxu.org/blog-cn/posts/dive-into-gradient-decent/>) [荐]★★★ - [one-hot 编码](<https://www.jianshu.com/p/a47a1c1fa3f1>) - [数据标准化/归一化normalization](<https://blog.csdn.net/pipisorry/article/details/52247379>) [荐] ## 1.2 深度学习 ### 1.2.1 基础入门 科普文章: - [推荐 | 机器学习经典总结,入门必读【17000字,可下载PDF】](https://mp.weixin.qq.com/s?__biz=MzIxODM4MjA5MA==&mid=2247485716&idx=1&sn=5b182c1c0b6578b1b1f3b75878ec1364&chksm=97ea2371a09daa6713afbe506d2bc40a7b2062be151c58425cefabf6cb3c41de37e9f51fdb0b&mpshare=1&scene=1&srcid=1209sLVUbZIHaCnoa9sLlfZ4#rd) - [一图看懂| 人工智能知识体系大全](https://mp.weixin.qq.com/s?__biz=MzU2MDc1MjEyMQ==&mid=2247486182&idx=1&sn=6174593f5862193e98cda311022aeb94&source=41#wechat_redirect) - [云计算、大数据和人工智能这么火,究竟是什么关系?](https://mp.weixin.qq.com/s?__biz=MzU2MDc1MjEyMQ==&mid=2247486185&idx=1&sn=0690fac9da75b1ea9f44c4f79df461a9&source=41#wechat_redirect) - [漫画:人工智能简史](https://mp.weixin.qq.com/s/0ps-tSx2xV1090NeZNQAFA) 入门: - [神经网络入门 - 阮一峰的网络日志](<http://www.ruanyifeng.com/blog/2017/07/neural-network.html>) [荐] - [如何识别图像边缘? - 阮一峰的网络日志](http://www.ruanyifeng.com/blog/2016/07/edge-recognition.html) - [从入门到精通:卷积神经网络初学者指南 | 机器之心](<https://www.jiqizhixin.com/articles/2016-08-01-3>) - [能否对卷积神经网络工作原理做一个直观的解释? - YJango的回答 - 知乎](<https://www.zhihu.com/question/39022858/answer/194996805>) - [深度学习入门指南:25个初学者要知道的概念 - 知乎](<https://zhuanlan.zhihu.com/p/27186569>) [荐] 基础: - [李理:详解卷积神经网络 - qunnie_yi的博客 - CSDN博客](<https://blog.csdn.net/qunnie_yi/article/details/80127218>) - [神经网络激励函数的作用是什么?有没有形象的解释? - 忆臻的回答 - 知乎](https://www.zhihu.com/question/22334626/answer/147543436) [荐] - [为什么都说神经网络是个黑箱? - mileistone的回答 - 知乎](https://www.zhihu.com/question/263672028/answer/430179912) - [【深度学习系列】卷积神经网络CNN原理详解(一)——基本原理 - Charlotte77](https://www.cnblogs.com/charlotte77/p/7759802.html) [荐] - [一文弄懂神经网络中的反向传播法——BackPropagation - Charlotte77](https://www.cnblogs.com/charlotte77/p/5629865.html) [荐] - [如何直观地解释 backpropagation 算法? - Anonymous的回答 - 知乎](https://www.zhihu.com/question/27239198/answer/89853077) 深度学习系列文章: - MachineLP:[MachineLP博客目录](https://blog.csdn.net/u014365862/article/details/78422372) - hanbingtao:[《零基础入门深度学习》系列文章](https://www.zybuluo.com/hanbingtao/note/433855) ### 1.2.2 迁移学习 - [一文看懂迁移学习:怎样用预训练模型搞定深度神经网络? - 知乎](<https://zhuanlan.zhihu.com/p/27657264>) [荐] - [pytorch-handbook/chapter4/04_1-fine-tuning.ipynb](<https://github.com/zergtant/pytorch-handbook/blob/a95c3a28564050640676818adbfd75eb104e95ea/chapter4/04_1-fine-tuning.ipynb>) [荐] > 把别人现成的训练好了的模型拿过来,换成自己的数据,调整一下参数,再训练一遍,这就是微调(fine-tune) > > 。。。 > > 其实 "Transfer Learning" 和 "Fine-tune" 并没有严格的区分,含义可以相互交换,只不过后者似乎更常用于形容迁移学习的后期微调中。 **我个人的理解,微调应该是迁移学习中的一部分。微调只能说是一个trick。** > > 下面只介绍下计算机视觉方向的微调,摘自 [cs231](http://cs231n.github.io/transfer-learning/) > > - ConvNet as fixed feature extractor > - Fine-tuning the ConvNet:固定前几层的参数,只对最后几层进行 fine-tuning ... > - Pretrained mod
Excerpt of 19,462 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:08c174006910e114, topic:tensorflow
matched fp:08c174006910e114, topic:opencv