整理 | Jane
出品 | AI科技大本营（公众号id：rgznai100）
在过去一年里，我们每个月都会给大家推荐一些优质的、最新的机器学习研究成果或机器学习技术文章，很多文章是从近千篇文章中评选出来的。综合考虑这些文章的更新时间、文章质量、受欢迎程度等因素，这次我们将为大家其中的 Top50，有些文章我们以前也详细讲解过，大家可以进行再次详读。
新一年已经开始了，大家可以从这些文章涉及的领域、方向，告诉我们你们今年更想看到看到哪些内容，今年我们将继续为大家介绍、推荐更多优质的学习资源。我在留言区等你们~
一、深度视频（Deep Video）
1、Deepfakes
https://towardsdatascience.com/family-fun-with-deepfakes-or-how-i-got-my-wife-onto-the-tonight-show-a4454775c011
2、Deep Video Portraits
https://web.stanford.edu/~zollhoef/papers/SG2018_DeepVideo/page.html%0A
二、面部识别（Face Recognition）
3、用 Python 实现 iPhone X 面部识别的深度学习方法
https://towardsdatascience.com/how-i-implemented-iphone-xs-faceid-using-deep-learning-in-python-d5dbaa128e1d
https://github.com/normandipalo/faceID_beta
4、通过 OpenCV、Python 和深度学习方法进行人脸识别
https://www.pyimagesearch.com/2018/06/18/face-recognition-with-opencv-python-and-deep-learning/
https://towardsdatascience.com/cutting-edge-face-recognition-is-complicated-these-spreadsheets-make-it-easier-e7864dbf0e1a?utm_source=mybridge&utm_medium=blog&utm_campaign=read_more
三、物体检测（Object Detection）
6、Aibnb 平台上的照片分类任务，实现大规模深度学习模型
https://medium.com/airbnb-engineering/categorizing-listing-photos-at-airbnb-f9483f3ab7e3
7、基于 OpenCV 用 YOLO 实现目标检测
从YOLOv1到YOLOv3，目标检测的进化之路
https://www.pyimagesearch.com/2018/11/12/yolo-object-detection-with-opencv
https://towardsdatascience.com/object-detection-with-10-lines-of-code-d6cb4d86f606
四、游戏 AI
9、一份初学者指南
https://www.gamedev.net/articles/programming/artificial-intelligence/the-total-beginners-guide-to-game-ai-r4942
10、OpenAI 发布基于预测奖励的强化学习
https://blog.openai.com/reinforcement-learning-with-prediction-based-rewards/
11、Uber 发布的一种解决硬探索问题的新算法
https://eng.uber.com/go-explore/
12、DeepMind 在夺旗游戏中取得的成果
https://deepmind.com/blog/capture-the-flag/
13、OpenAI Five 在 DOTA2 中击败人类选手，取得的成果
https://blog.openai.com/openai-five
五、棋类游戏
14、AlphaZero，DeepMind 发布在围棋、国际象棋等棋类任务中取得的新成果
https://deepmind.com/blog/alphazero-shedding-new-light-grand-games-chess-shogi-and-go
15、基于 Python 和 Keras 实现的 AlphaZero AI
https://medium.com/applied-data-science/how-to-build-your-own-alphazero-ai-using-python-and-keras-7f664945c188
16、AI 学会围棋的一个简单解释
https://medium.freecodecamp.org/explained-simply-how-an-ai-program-mastered-the-ancient-game-of-go-62b8940a9080
六、医疗领域
17、深度学习在医疗图像数据集中存在哪些不合理的用法
https://lukeoakdenrayner.wordpress.com/2018/04/30/the-unreasonable-usefulness-of-deep-learning-in-medical-image-datasets
18、Scaling up molecular pattern recognition with DNA-based winner-take-all neural networks
https://www.nature.com/articles/s41586-018-0289-6.epdf
19、用深度学习研究大脑 MR 图像
https://medium.com/stanford-ai-for-healthcare/its-a-no-brainer-deep-learning-for-brain-mr-images-f60116397472
七、运动
https://carolineec.github.io/everybody_dance_now/
21、Berkeley 研究的虚拟特效人
https://bair.berkeley.edu/blog/2018/04/10/virtual-stuntman
22、OpenAI 研究的灵活机械人手
OpenAI发布最新「模拟机器人环境」,用「真实机器人」模型进行训练
https://blog.openai.com/learning-dexterity/
23、DeepMind 发布的Navigating with grid-like representations in artificial agents
https://deepmind.com/blog/grid-cells/
八、Web & APP
24、如何用 CoreML、PyTorch 和 React Native 在 ios 上完成一个神经网络
https://attardi.org/pytorch-and-coreml
25、如何训练一个 AI 将模型设计转换成 HTML 和 CSS
https://medium.freecodecamp.org/how-you-can-train-an-ai-to-convert-your-design-mockups-into-html-and-css-cc7afd82fed4
九、翻译任务
26、由 Facebook Code 发布将神经机器翻译推广至更大的数据集上
https://code.fb.com/ai-research/scaling-neural-machine-translation-to-bigger-data-sets-with-faster-training-and-inference/
27、Building a language translator from scratch with deep learning
https://blog.floydhub.com/language-translator/
28、Facebook Research 研究的无监督机器翻译
https://code.fb.com/ai-research/unsupervised-machine-translation-a-novel-approach-to-provide-fast-accurate-translations-for-more-languages
十、NLP
29、图解BERT, ELMo等模型（NLP如何破解迁移学习）
http://jalammar.github.io/illustrated-bert/
30、The Annotated Transformer — Harvard NLP
http://nlp.seas.harvard.edu/2018/04/03/attention.html
https://medium.com/@ageitgey/natural-language-processing-is-fun-9a0bff37854e?utm_source=mybridge&utm_medium=blog&utm_campaign=read_more
十一、神经网络
32、如何用 Python 从头构建一个神经网络
https://towardsdatascience.com/how-to-build-your-own-neural-network-from-scratch-in-python-68998a08e4f6
33、用 Numpy 实现一个神经网络
https://towardsdatascience.com/lets-code-a-neural-network-in-plain-numpy-ae7e74410795
十二、CNN
34、Differentiable Image Parameterizations
https://distill.pub/2018/differentiable-parameterizations
35、Feature-wise 转换
https://distill.pub/2018/feature-wise-transformations/
36、Keras 与 CNN
https://www.pyimagesearch.com/2018/04/16/keras-and-convolutional-neural-networks-cnns
37、The Building Blocks of Interpretability
https://distill.pub/2018/building-blocks/
38、Facebook 公开的 Rosetta 系统，识别图像或视频中的文字
https://code.fb.com/ai-research/rosetta-understanding-text-in-images-and-videos-with-machine-learning/
39、Uber 发表的一篇文章，关于 CNN 和 CoordConv Solution 一个有意思的缺陷
https://eng.uber.com/coordconv
十三、RNN
https://ai.googleblog.com/2018/05/duplex-ai-system-for-natural-conversation.html
41、World Models
https://worldmodels.github.io/
十四、强化学习
42、Lessons Learned Reproducing a Deep Reinforcement Learning Paper
http://amid.fish/reproducing-deep-rl
43、Berkeley发布的 Dexterous Manipulation with Reinforcement Learning
https://bair.berkeley.edu/blog/2018/08/31/dexterous-manip/
44、强化学习还不能发挥作用
https://www.alexirpan.com/2018/02/14/rl-hard.html
十五、TensorFlow
45、Triplet Loss and Online Triplet Mining in TensorFlow
https://omoindrot.github.io/triplet-loss
46、TensorFlow 答疑解惑
http://jacobbuckman.com/post/tensorflow-the-confusing-parts-1/
47、Tensorflow-Project-Template
https://github.com/Mrgemy95/Tensorflow-Project-Template
48、用 TF.js 在浏览器中实现实时人体姿态估计
https://medium.com/tensorflow/real-time-human-pose-estimation-in-the-browser-with-tensorflow-js-7dd0bc881cd5
十六、学习指南
49、Rules of Machine Learning: | ML Universal Guides | Google Developers
https://developers.google.com/machine-learning/guides/rules-of-ml
50、Model-based machine learning
http://mbmlbook.com/toc.html
想看、学习哪些 AI内容，欢迎给我们留言哦~
（本文为AI科技大本营整理文章，转载请微信联系 1092722531）
群招募
扫码添加小助手微信，回复：公司+研究方向（学校+研究方向），邀你加入技术交流群。技术群审核较严，敬请谅解。
推荐阅读：
❤点击“阅读原文”，查看历史精彩文章。
212
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
