深度学习开源项目集合

1. Image Completion with Deep Learning in TensorFlow

bamos/dcgan-completion.tensorflow https://github.com/bamos/dcgan-completion.tensorflow

Image Completion with Deep Learning in TensorFlow 文章 http://bamos.github.io/2016/08/09/deep-completion/

2. NEURAL FACE

Neural Face is an Artificial Intelligence which generates face images and all images in this page are not REAL.

Neural Face | 프사 뉴럴 http://carpedm20.github.io/faces/

A torch implementation of http://arxiv.org/abs/1511.06434

3. Face recognition with deep neural networks.

OpenFace介绍网站 http://cmusatyalab.github.io/openface/

cmusatyalab/openface

4. 利用生成对抗网络(generative adversarial networks, GANs)生成二次元妹子

A simple PyTorch Implementation of Generative Adversarial Networks, focusing on anime face drawing.

jayleicn/animeGAN https://github.com/jayleicn/animeGAN

5. Tensorflow implementation of Human-Level Control through Deep Reinforcement Learning

devsisters/DQN-tensorflow https://github.com/devsisters/DQN-tensorflow

6. Flappy Bird hack using Deep Reinforcement Learning (Deep Q-learning).

This project follows the description of the Deep Q Learning algorithm described in Playing Atari with Deep Reinforcement Learning [2] and shows that this learning algorithm can be further generalized to the notorious Flappy Bird.

https://github.com/yenchenlin/DeepLearningFlappyBird

整个项目的根基就是这个游戏本身,在增强学习中可以看作是“Environment”,也就是能接收Action并反馈Next State和Reward的。作者使用了别人写好的现成的程序,不过还是在此基础上做了一定的修改,使游戏能够返回这两个信息。对于Next State,则是下一个时刻的图画信息,对于Reward,主要就是能分了,穿过一对管子得一分,死掉得负分。
预测部分:作者采用了E&E的策略,除了以模型进行判定之外,还以一定概率随机操作。
学习部分:当数据集收集到一定数量后开始学习,对于非完结的状态,要根据Bellman Equation进行Return的计算(Return和Reward有一定的差别)。然后将(state,action,return)扔给模型学习。
模型:CNN,具体的参数README讲得很清楚。总体来看这就是一个模仿Atari实现的项目,

7. 把一副图像变成铅笔水粉画

http://www.cse.cuhk.edu.hk/~leojia/projects/pencilsketch/pencil_drawing.htm

一个非官方的实现在这里:
https://github.com/fumin/pencil implementation of http://www.cse.cuhk.edu.hk/~leojia/projects/pencilsketch/npar12_pencil.pdf

8. Character-Aware Neural Language Models

Tensorflow implementation of Character-Aware Neural Language Models. The original code of author can be found here.
carpedm20/lstm-char-cnn-tensorflow https://github.com/carpedm20/lstm-char-cnn-tensorflow

9. “End-To-End Memory Networks” in Tensorflow

论文: http://arxiv.org/abs/1503.08895v4

代码: https://github.com/carpedm20/MemN2N-tensorflow

10. TensorFlow implementation of Pixel Recurrent Neural Networks

https://github.com/carpedm20/pixel-rnn-tensorflow

11. A recurrent neural network for generating little stories about images

你提供一张图,然后计算机根据这张图讲一个小故事。
https://github.com/ryankiros/neural-storyteller

12. Efficient Image Captioning code in Torch, runs on GPU

给出一张图片,然后生成一句描述,
Efficient Image Captioning code in Torch, runs on GPU
https://github.com/karpathy/neuraltalk2

13. Neural Artistic Style in Python

Implementation of A Neural Algorithm of Artistic Style https://arxiv.org/abs/1508.06576.
A method to transfer the style of one image to the subject of another image.
提供一张图片确定一种风格,就可以将输入图片风格化(艺术化)

https://github.com/andersbll/neural_artistic_style

14. Variational Auto Encoder

论文 http://arxiv.org/abs/1312.6114
A collection of generative methods implemented with TensorFlow (Deep Convolutional Generative Adversarial Networks (DCGAN), Variational Autoencoder (VAE) and DRAW: A Recurrent Neural Network For Image Generation).

ikostrikov/TensorFlow-VAE-GAN-DRAW

15. Generative Adversarial Networks 网络

论文https://arxiv.org/abs/1406.2661

这个模型有两个网络,一个“造假者”,一个“判别真假的警察”,互相博弈训练,最后使得“造假者”可以以假乱真。

https://github.com/carpedm20/DCGAN-tensorflow
A tensorflow implementation of “Deep Convolutional Generative Adversarial Networks” http://carpedm20.github.io/faces/

https://github.com/ericjang/draw
TensorFlow Implementation of “DRAW: A Recurrent Neural Network For Image Generation”

https://github.com/mattya/chainer-DCGAN
Chainer implementation of Deep Convolutional Generative Adversarial Network

参考
A curated list of deep learning resources for computer vision, kjw0612/awesome-deep-vision

Top Deep Learning Projects in Github

【Deep Learning】Github上关注最多的53个深度学习项目 - CSDN博客


技术交流学习,请加QQ微信:631531977
目录