allensmile/awesome-chainer

Awesome Chainer

What is Chainer?

Chainer is a flexible framework for neural networks. One major goal is flexibility, so it must enable us to write complex architectures simply and intuitively.

More info here

ref Chainer.wiki - Chainer Wiki

Table of Contents

• Tutorials
• Models/Projects
• Examples
• Libraries
• Videos
• Papers
• Blog posts
• Community
• Books

Tutorials

• Chainer Tutorial Offical document - Chainer Tutorial Offical document

Hands-on

• hido/chainer-handson - Chainer Hands-on
• iwiwi/chainer-handson - Chainer Hands-on(JP CPU only)

Models/Projects

Preferred Networks official

• ChainerRL - ChainerRL is a deep reinforcement learning library built on top of Chainer.
• ChainerCV - Versatile set of tools for Deep Learning based Computer Vision
• Paint Chainer - Paints Chainer is line drawing colorizer using chainer.

Chainer External Examples

Preferred NetWorks Research

• LSGAN - Least Squares Generative Adversarial Network implemented in Chainer
• Chainer-gogh - Implementation of "A neural algorithm of Artistic style"
• Chainer Graph CNN - Chainer implementation of 'Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering' (https://arxiv.org/abs/1606.09375)
• chainer-segnet - SegNet implementation & experiments in Chainer
• chainer-pix2pix - chainer implementation of pix2pix

Examples

• hido/chainer-handson - Chainer Hands-on
• yusuketomoto/chainer-char-rnn - Recurrent neural network (RNN)
• mitmul/chainer-fast-rcnn - Fast R-CNN
• apple2373/chainer-simple-fast-rnn - Fast R-CNN
• mitmul/chainer-faster-rcnn - Faster R-CNN
• jekbradbury/qrnn.py - QRNN
• mitmul/chainer-siamese - Siamese Network
• mitmul/chainer-svm - Support Vector Machine (SVM)
• tscohen/GrouPy - Group Equivariant Convolutional Neural Networks

NLP

• odashi/chainer_examples - Machine Translation, word segmentation, and language model
• odashi/chainer_rnnlm.py - RNN Language Model
• odashi/chainer_encoder_decoder.py - Neural Encoder-Decoder Machine Translation
• prajdabre/chainer_examples - LSTM variants

Computer Vision

• mitmul/chainer-cifar10 - Cifar10
• mitmul/DeepPose - Deep pose
• mrkn/chainer-srcnn - Image super-resolution
• Hi-king/chainer_superresolution - Image super-resolution
• darashi/chainer-example-overfeat-classify - Overfeat
• RyotaKatoh/chainer-Variational-AutoEncoder - Variational autoencoder (VAE)
• mitmul/chainer-imagenet-vgg - VGG
• yasunorikudo/chainer-ResNet - ResNet
• yasunorikudo/chainer-DenseNet - DenseNet
• yasunorikudo/chainer-ResDrop - ResDrop
• mitmul/chainer-conv-vis - Convolution Filter Visualization
• yusuketomoto/chainer-fast-neuralstyle - Perceptual Losses for Real-Time Style Transfer and Super-Resolution
• rezoo/illustration2vec - illustration2vec
• apple2373/chainer_stylenet - StyleNet (A Neural Algorithm of Artistic Style)
• mattya/chainer-gogh - StyleNet (A Neural Algorithm of Artistic Style)
• apple2373/chainer_caption_generation - Show and Tell
• chainer-prednet - Deep Predictive Coding Networks
• hillbig/binary_net - BinaryNet
• stitchfix/fauxtograph - Variational Auto-Encoder (VAE), Generative Adversarial Nets (GAN), VAE-GAN
• rezoo/data.py - Generative Adversarial Nets (GAN)
• mattya/chainer-DCGAN - Deep Convolutional Generative Adversarial Network (DCGAN)
• mattya/chainer-fluid - Fluid simulation
• ktnyt/chainer_ca.py - Convolutional Autoencoder
• tochikuji/chainer-libDNN - Stacked Denoising Autoencoder
• masaki-y/ram - Recurrent Attention Model
• wkentaro/fcn - Fully Convolutional Networks
• hvy/chainer-gan-denoising-feature-matching - Generative Adversarial Networks with Denoising Feature Matching
• hvy/chainer-visualization - Visualizing and Understanding Convolutional Networks
• hvy/chainer-gan-trainer - Chainer GAN Trainer
• dsanno/chainer-dfi - Deep Feature Interpolation for Image Content Changes
• mitmul/chainer-segnet - SegNet
• musyoku/wasserstein-gan - WGAN
• weihua916/imsat - IMSAT
• Hakuyume/chainer-ssd - SSD
• leetenki/YOLOv2 - YOLOv2
• leetenki/YOLOtiny_v2 - YOLOtiny_v2
• deformable-conv - Deformable-conv
• chainercv - chainercv

Reinforcement Learning

• ugo-nama-kun/DQN-chainer - Deep Q-Network (DQN)

musyoku

• musyoku/LSGAN - LSGAN
• musyoku/began - BEGAN
• adversarial-autoencoder - adversarial-autoencoder
• dsanno/chainer-dfi - chainer-dfi
• chainer-VAE - chainer-VAE
• wavenet - wavenet
• chainer-sequential - chainer-sequential
• musyoku/IMSAT - IMSAT
• musyoku/unrolled-gan - unrolled-ga
• musyoku/improved-gan- improved-gan
• musyoku/mnist-oneshot - mnist-oneshot
• musyoku/adgm - adgm
• musyoku/vat - VAT
• musyoku/ddgm - DDGM
• musyoku/recurrent-batch-normalization - recurrent-batch-normalization
• musyoku/weight-normalization - weight-normalization
• musyoku/minibatch_discrimination - minibatch_discrimination
• musyoku/variational-autoencoder - VAE

Blog posts

• Introduction to Chainer: Neural Networks in Python
• The DIY Guide to Chainer
• CHAINER CHARACTER EMBEDDINGS
• A Fontastic Voyage: Generative Fonts with Adversarial Networks

Tools and extensions

• Deel; A High level deep neural network description language
• DEEPstation
• chainer_imagenet_tools
• scikit-chainer
• chainer-libDNN

Videos

Only Japanese

• Chainer の Trainer 解説とNStepLSTM について Published on Mar 15, 2017
• Chainer Meetup #04 Published on Feb 23, 2017
• 1014：深層学習フレームワークChainerの導入と化合物活性予測への応用 Published on Dec 2, 2015

Papers

• GP-GAN: Towards Realistic High-Resolution Image Blending
  • Conference: arXiv only
  • Codes: wuhuikai/GP-GAN
• Temporal Generative Adversarial Nets
  • Conference: arXiv only
• Reasoning with Memory Augmented Neural Networks for Language Comprehension
  • Conference: arXiv only
• PMI Matrix Approximations with Applications to Neural Language Modeling
  • Conference: arXiv only
• Neural Tree Indexers for Text Understanding
  • Conference: arXiv only
  • Codes: NTI
• Neural Semantic Encoders
  • Conference: arXiv only
• Networked Intelligence: Towards Autonomous Cyber Physical Systems
  • Conference: arXiv only
• Modeling the dynamics of human brain activity with recurrent neural networks
  • Conference: arXiv only
• A Deep-Learning Approach for Operation of an Automated Realtime Flare Forecast
  • Conference: arXiv only
• Convolutional Neural Networks using Logarithmic Data Representation
  • Conference: arXiv only
• context2vec: Learning Generic Context Embedding with Bidirectional LSTM
  • Conference: CoNLL 2016
• Mixing Dirichlet Topic Models and Word Embeddings to Make lda2vec
  • Conference: CoNLL 2016
• Deep Impression: Audiovisual Deep Residual Networks for Multimodal Apparent Personality Trait Recognition
  • Conference: ECCV 2016 Workshop
  • comments: "3rd place in Looking at People ECCV Challenge"
• Learning Joint Representations of Videos and Sentences with Web Image Search
  • Conference: ECCV 2016 Workshop
• Incorporating Discrete Translation Lexicons into Neural Machine Translation
  • Conference: EMNLP 2016
• Controlling Output Length in Neural Encoder-Decoders
  • Conference: EMNLP 2016
• Insertion Position Selection Model for Flexible Non-Terminals in Dependency Tree-to-Tree Machine Translation
  • Conference: EMNLP 2016
• Learning Representations Using Complex-Valued Nets
  • Conference: ICLR 2016
• Dynamic Coattention Networks For Qustion Answering
  • Conference: ICLR 2017 under review
• SqueezeNet: AlexNet-level Accuracy with 50x Fewer Parameters and < 0.5MB Model Size
  • Conference: ICLR 2017 under review
• Quasi-Recurrent Neural Networks
  • Conference: ICLR 2017 under review
• Steerable CNNs
  • Conference: ICLR 2017
  • comments: Chainer is not referred in the paper, but the authors kindly informed us.
• f-GAN: Training Generative Neural Samplers using Variational Divergence Minimization
  • Conference: NIPS 2016 Workshop
• QSGD: Randomized Quantization for Communication-Optimal Stochastic Gradient Descent
  • Conference: OPT 2016
• Evaluation of Deep Learning based Pose Estimation for Sign Language Recognition
  • Conference: PETRA 2016
• Machine-learning Selection of Optical Transients in Subaru/Hyper Suprime-Cam Survey
  • Conference: PASJ 2016
  • comments: PASJ: Publications of the Astronomical Society of Japan
• A Deep-Learning Approach for Operation of an Automated Realtime Flare Forecast
  • Conference: Space Weather 2016
• Dynamic Entity Representation with Max-pooling Improves Machine Reading
  • Conference: NAACL 2016
• Feature-based Model versus Convolutional Neural Network for Stance Detection
  • Conference: SemEval 2016
• Cross-Lingual Image Caption Generation
  • Conference: ACL 2016
• Composing Distributed Representations of Relational Patterns
  • Conference: ACL 2016
• Generating Natural Language Descriptions for Semantic Representations of Human Brain Activity
  • Conference: ACL 2016
• MetaMind Neural Machine Translation System for WMT 2016
  • Conference: WMT 2016
• Group Equivariant Convolutional Networks
  • Conference: ICML 2016
  • Codes: GitHub
  • comments: Chainer is not referred in the paper, but the authors kindly informed us.
• Robocodes: Towards Generative Street Addresses from Satellite Imagery
  • Conference: CVPR 2017 Workshop

Official announcements

• Chainer document - An introduction to Chainer
• Chainer blogs

Community

• Introduction to Chainer: Neural Networks in Python
• The DIY Guide to Chainer
• CHAINER CHARACTER EMBEDDINGS
• A Fontastic Voyage: Generative Fonts with Adversarial Networks

Community

Global

• @ChainerOfficial on Twitter
• Mailing List
• Slack

Japan

• @ChainerJP on Twitter
• Mailing List
• Slack
• connpass

Books

• Chainerによる実践深層学習 by 新納浩幸