今天給大家介紹自 2014 年以來，計(jì)算機(jī)視覺 CV 領(lǐng)域圖像分類方向文獻(xiàn)和代碼的超全總結(jié)和列表！總共涉及 36 種 ConvNet 模型。該 GitHub 項(xiàng)目作者是 weiaicunzai，項(xiàng)目地址是：

https://github.com/weiaicunzai/awesome-image-classification

背景

我相信圖像識(shí)別是深入到其它機(jī)器視覺領(lǐng)域一個(gè)很好的起點(diǎn)，特別是對(duì)于剛剛?cè)腴T深度學(xué)習(xí)的人來說。當(dāng)我初學(xué) CV 時(shí)，犯了很多錯(cuò)。我當(dāng)時(shí)非常希望有人能告訴我應(yīng)該從哪一篇論文開始讀起。到目前為止，似乎還沒有一個(gè)像 deep-learning-object-detection 這樣的 GitHub 項(xiàng)目。因此，我決定建立一個(gè) GitHub 項(xiàng)目，列出深入學(xué)習(xí)中關(guān)于圖像分類的論文和代碼，以幫助其他人。

對(duì)于學(xué)習(xí)路線，我的個(gè)人建議是，對(duì)于那些剛?cè)腴T深度學(xué)習(xí)的人，可以試著從 vgg 開始，然后是 googlenet、resnet，之后可以自由地繼續(xù)閱讀列出的其它論文或切換到其它領(lǐng)域。

性能表

基于簡化的目的，我只從論文中列舉出在 ImageNet 上準(zhǔn)確率最高的 top1 和 top5。注意，這并不一定意味著準(zhǔn)確率越高，一個(gè)網(wǎng)絡(luò)就比另一個(gè)網(wǎng)絡(luò)更好。因?yàn)橛行┚W(wǎng)絡(luò)專注于降低模型復(fù)雜性而不是提高準(zhǔn)確性，或者有些論文只給出 ImageNet 上的 single crop results，而另一些則給出模型融合或 multicrop results。

關(guān)于性能表的標(biāo)注：

ConvNet：卷積神經(jīng)網(wǎng)絡(luò)的名稱

ImageNet top1 acc：論文中基于 ImageNet 數(shù)據(jù)集最好的 top1 準(zhǔn)確率

ImageNet top5 acc：論文中基于 ImageNet 數(shù)據(jù)集最好的 top5 準(zhǔn)確率

Published In：論文發(fā)表在哪個(gè)會(huì)議或期刊

論文&代碼

1. VGG

Very Deep Convolutional Networks for Large-Scale Image Recognition.

Karen Simonyan, Andrew Zisserman

pdf: https://arxiv.org/abs/1409.1556

code: torchvision :

https://github.com/pytorch/vision/blob/master/torchvision/models/vgg.py

code: keras-applications :

https://github.com/keras-team/keras-applications/blob/master/keras_applications/vgg16.py

code: keras-applications :

https://github.com/keras-team/keras-applications/blob/master/keras_applications/vgg19.py

2. GoogleNet

Going Deeper with Convolutions

Christian Szegedy, Wei Liu, Yangqing Jia, Pierre Sermanet, Scott Reed, Dragomir Anguelov, Dumitru Erhan, Vincent Vanhoucke, Andrew Rabinovich

pdf:https://arxiv.org/abs/1409.4842

code: unofficial-tensorflow :

https://github.com/conan7882/GoogLeNet-Inception

code: unofficial-caffe :

https://github.com/lim0606/caffe-googlenet-bn

3.PReLU-nets

Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification

Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun

pdf:https://arxiv.org/abs/1502.01852

code: unofficial-chainer :

https://github.com/nutszebra/prelu_net

4.ResNet

Deep Residual Learning for Image Recognition

Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun

pdf:https://arxiv.org/abs/1512.03385

code: facebook-torch :

https://github.com/facebook/fb.resnet.torch

code: torchvision :

https://github.com/pytorch/vision/blob/master/torchvision/models/resnet.py

code: keras-applications :

https://github.com/keras-team/keras-applications/blob/master/keras_applications/resnet.py

code: unofficial-keras :

https://github.com/raghakot/keras-resnet

code: unofficial-tensorflow :

https://github.com/ry/tensorflow-resnet

5.PreActResNet

Identity Mappings in Deep Residual Networks

Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun

pdf:https://arxiv.org/abs/1603.05027

code: facebook-torch :

https://github.com/facebook/fb.resnet.torch/blob/master/models/preresnet.lua

code: official :

https://github.com/KaimingHe/resnet-1k-layers

code: unoffical-pytorch :

https://github.com/kuangliu/pytorch-cifar/blob/master/models/preact_resnet.py

code: unoffical-mxnet :

https://github.com/tornadomeet/ResNet

6.Inceptionv3

Rethinking the Inception Architecture for Computer Vision

Christian Szegedy, Vincent Vanhoucke, Sergey Ioffe, Jonathon Shlens, Zbigniew Wojna

pdf:https://arxiv.org/abs/1512.00567

code: torchvision :

https://github.com/pytorch/vision/blob/master/torchvision/models/inception.py

code: keras-applications :

https://github.com/keras-team/keras-applications/blob/master/keras_applications/inception_v3.py

7.Inceptionv4 && Inception-ResNetv2

Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning

Christian Szegedy, Sergey Ioffe, Vincent Vanhoucke, Alex Alemi

pdf:https://arxiv.org/abs/1602.07261

code: unofficial-keras :

https://github.com/kentsommer/keras-inceptionV4

code: unofficial-keras :

https://github.com/titu1994/Inception-v4

code: unofficial-keras :

https://github.com/yuyang-huang/keras-inception-resnet-v2

8. RIR

Resnet in Resnet: Generalizing Residual Architectures

Sasha Targ, Diogo Almeida, Kevin Lyman

pdf:https://arxiv.org/abs/1603.08029

code: unofficial-tensorflow :

https://github.com/SunnerLi/RiR-Tensorflow

code: unofficial-chainer :

https://github.com/nutszebra/resnet_in_resnet

9.Stochastic Depth ResNet

Deep Networks with Stochastic Depth

Gao Huang, Yu Sun, Zhuang Liu, Daniel Sedra, Kilian Weinberger

pdf:https://arxiv.org/abs/1603.09382

code: unofficial-torch :

https://github.com/yueatsprograms/Stochastic_Depth

code: unofficial-chainer :

https://github.com/yasunorikudo/chainer-ResDrop

code: unofficial-keras :

https://github.com/dblN/stochastic_depth_keras

10.WRN

Wide Residual Networks

Sergey Zagoruyko, Nikos Komodakis

pdf:https://arxiv.org/abs/1605.07146

code: official :

https://github.com/szagoruyko/wide-residual-networks

code: unofficial-pytorch :

https://github.com/xternalz/WideResNet-pytorch

code: unofficial-keras :

https://github.com/asmith26/wide_resnets_keras

code: unofficial-pytorch :

https://github.com/meliketoy/wide-resnet.pytorch

11.squeezenet

SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size?

Forrest N. Iandola, Song Han, Matthew W. Moskewicz, Khalid Ashraf, William J. Dally, Kurt Keutzer

pdf:https://arxiv.org/abs/1602.07360

code: torchvision :

https://github.com/pytorch/vision/blob/master/torchvision/models/squeezenet.py

code: unofficial-caffe :

https://github.com/DeepScale/SqueezeNet

code: unofficial-keras :

https://github.com/rcmalli/keras-squeezenet

code: unofficial-caffe :

https://github.com/songhan/SqueezeNet-Residual

12.GeNet

Genetic CNN

Lingxi Xie, Alan Yuille

pdf:https://arxiv.org/abs/1703.01513

code: unofficial-tensorflow :

https://github.com/aqibsaeed/Genetic-CNN

12.MetaQNN

Designing Neural Network Architectures using Reinforcement Learning

Bowen Baker, Otkrist Gupta, Nikhil Naik, Ramesh Raskar

pdf:https://arxiv.org/abs/1703.01513

code: official :https://github.com/bowenbaker/metaqnn

13.PyramidNet

Deep Pyramidal Residual Networks

Dongyoon Han, Jiwhan Kim, Junmo Kim

pdf:https://arxiv.org/abs/1610.02915

code: official :

https://github.com/jhkim89/PyramidNet

code: unofficial-pytorch :

https://github.com/dyhan0920/PyramidNet-PyTorch

14.DenseNet

Densely Connected Convolutional Networks

Gao Huang, Zhuang Liu, Laurens van der Maaten, Kilian Q. Weinberger

pdf:https://arxiv.org/abs/1608.06993

code: official :

https://github.com/liuzhuang13/DenseNet

code: unofficial-keras :

https://github.com/titu1994/DenseNet

code: unofficial-caffe :

https://github.com/shicai/DenseNet-Caffe

code: unofficial-tensorflow :

https://github.com/YixuanLi/densenet-tensorflow

code: unofficial-pytorch :

https://github.com/YixuanLi/densenet-tensorflow

code: unofficial-pytorch :

https://github.com/bamos/densenet.pytorch

code: unofficial-keras :

https://github.com/flyyufelix/DenseNet-Keras

15.FractalNet

FractalNet: Ultra-Deep Neural Networks without Residuals

Gustav Larsson, Michael Maire, Gregory Shakhnarovich

pdf:https://arxiv.org/abs/1605.07648

code: unofficial-caffe :

https://github.com/gustavla/fractalnet

code: unofficial-keras :

https://github.com/snf/keras-fractalnet

code: unofficial-tensorflow :

https://github.com/tensorpro/FractalNet

16.ResNext

Aggregated Residual Transformations for Deep Neural Networks

Saining Xie, Ross Girshick, Piotr Dollár, Zhuowen Tu, Kaiming He

pdf:https://arxiv.org/abs/1611.05431

code: official :

https://github.com/facebookresearch/ResNeXt

code: keras-applications :

https://github.com/keras-team/keras-applications/blob/master/keras_applications/resnext.py

code: unofficial-pytorch :

https://github.com/prlz77/ResNeXt.pytorch

code: unofficial-keras :

https://github.com/titu1994/Keras-ResNeXt

code: unofficial-tensorflow :

https://github.com/taki0112/ResNeXt-Tensorflow

code: unofficial-tensorflow :

https://github.com/wenxinxu/ResNeXt-in-tensorflow

17.IGCV1

Interleaved Group Convolutions for Deep Neural Networks

Ting Zhang, Guo-Jun Qi, Bin Xiao, Jingdong Wang

pdf:https://arxiv.org/abs/1707.02725

code official :

https://github.com/hellozting/InterleavedGroupConvolutions

18.Residual Attention Network

Residual Attention Network for Image Classification

Fei Wang, Mengqing Jiang, Chen Qian, Shuo Yang, Cheng Li, Honggang Zhang, Xiaogang Wang, Xiaoou Tang

pdf:https://arxiv.org/abs/1704.06904

code: official :

https://github.com/fwang91/residual-attention-network

code: unofficial-pytorch :

https://github.com/tengshaofeng/ResidualAttentionNetwork-pytorch

code: unofficial-gluon :

https://github.com/PistonY/ResidualAttentionNetwork

code: unofficial-keras :

https://github.com/koichiro11/residual-attention-network

19.Xception

Xception: Deep Learning with Depthwise Separable Convolutions

Fran?ois Chollet

pdf:https://arxiv.org/abs/1610.02357

code: unofficial-pytorch :

https://github.com/jfzhang95/pytorch-deeplab-xception/blob/master/modeling/backbone/xception.py

code: unofficial-tensorflow :

https://github.com/kwotsin/TensorFlow-Xception

code: unofficial-caffe :

https://github.com/yihui-he/Xception-caffe

code: unofficial-pytorch :

https://github.com/tstandley/Xception-PyTorch

code: keras-applications :

https://github.com/keras-team/keras-applications/blob/master/keras_applications/xception.py

20.MobileNet

MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications

Andrew G. Howard, Menglong Zhu, Bo Chen, Dmitry Kalenichenko, Weijun Wang, Tobias Weyand, Marco Andreetto, Hartwig Adam

pdf:https://arxiv.org/abs/1704.04861

code: unofficial-tensorflow :

https://github.com/Zehaos/MobileNet

code: unofficial-caffe :

https://github.com/shicai/MobileNet-Caffe

code: unofficial-pytorch :

https://github.com/marvis/pytorch-mobilenet

code: keras-applications :

https://github.com/keras-team/keras-applications/blob/master/keras_applications/mobilenet.py

21.PolyNet

PolyNet: A Pursuit of Structural Diversity in Very Deep Networks

Xingcheng Zhang, Zhizhong Li, Chen Change Loy, Dahua Lin

pdf:https://arxiv.org/abs/1611.05725

code: official :

https://github.com/open-mmlab/polynet

22.DPN

Dual Path Networks

Yunpeng Chen, Jianan Li, Huaxin Xiao, Xiaojie Jin, Shuicheng Yan, Jiashi Feng

pdf:https://arxiv.org/abs/1707.01629

code: official :

https://github.com/cypw/DPNs

code: unoffical-keras :

https://github.com/titu1994/Keras-DualPathNetworks

code: unofficial-pytorch :

https://github.com/oyam/pytorch-DPNs

code: unofficial-pytorch :

https://github.com/rwightman/pytorch-dpn-pretrained

23.Block-QNN

Practical Block-wise Neural Network Architecture Generation

Zhao Zhong, Junjie Yan, Wei Wu, Jing Shao, Cheng-Lin Liu

pdf:https://arxiv.org/abs/1708.05552

24.CRU-Net

Sharing Residual Units Through Collective Tensor Factorization in Deep Neural Networks

Chen Yunpeng, Jin Xiaojie, Kang Bingyi, Feng Jiashi, Yan Shuicheng

pdf:https://arxiv.org/abs/1703.02180

code official :

https://github.com/cypw/CRU-Net

code unofficial-mxnet :

https://github.com/bruinxiong/Modified-CRUNet-and-Residual-Attention-Network.mxnet

25.ShuffleNet

ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices

Xiangyu Zhang, Xinyu Zhou, Mengxiao Lin, Jian Sun

pdf:https://arxiv.org/abs/1707.01083

code: unofficial-tensorflow :

https://github.com/MG2033/ShuffleNet

code: unofficial-pytorch :

https://github.com/jaxony/ShuffleNet

code: unofficial-caffe :

https://github.com/farmingyard/ShuffleNet

code: unofficial-keras :

https://github.com/scheckmedia/keras-shufflenet

26.CondenseNet

CondenseNet: An Efficient DenseNet using Learned Group Convolutions

Gao Huang, Shichen Liu, Laurens van der Maaten, Kilian Q. Weinberger

pdf:https://arxiv.org/abs/1711.09224

code: official :

https://github.com/ShichenLiu/CondenseNet

code: unofficial-tensorflow :

https://github.com/markdtw/condensenet-tensorflow

27.NasNet

Learning Transferable Architectures for Scalable Image Recognition

Barret Zoph, Vijay Vasudevan, Jonathon Shlens, Quoc V. Le

pdf:https://arxiv.org/abs/1707.07012

code: unofficial-keras :

https://github.com/titu1994/Keras-NASNet

code: keras-applications :

https://github.com/keras-team/keras-applications/blob/master/keras_applications/nasnet.py

code: unofficial-pytorch :

https://github.com/wandering007/nasnet-pytorch

code: unofficial-tensorflow :

https://github.com/yeephycho/nasnet-tensorflow

28.MobileNetV2

MobileNetV2: Inverted Residuals and Linear Bottlenecks

Mark Sandler, Andrew Howard, Menglong Zhu, Andrey Zhmoginov, Liang-Chieh Chen

pdf:https://arxiv.org/abs/1801.04381

code: unofficial-keras :

https://github.com/xiaochus/MobileNetV2

code: unofficial-pytorch :

https://github.com/Randl/MobileNetV2-pytorch

code: unofficial-tensorflow :

https://github.com/neuleaf/MobileNetV2

29.IGCV2

IGCV2: Interleaved Structured Sparse Convolutional Neural Networks

Guotian Xie, Jingdong Wang, Ting Zhang, Jianhuang Lai, Richang Hong, Guo-Jun Qi

pdf:https://arxiv.org/abs/1804.06202

30.hier

Hierarchical Representations for Efficient Architecture Search

Hanxiao Liu, Karen Simonyan, Oriol Vinyals, Chrisantha Fernando, Koray Kavukcuoglu

pdf:https://arxiv.org/abs/1711.00436

31.PNasNet

Progressive Neural Architecture Search

Chenxi Liu, Barret Zoph, Maxim Neumann, Jonathon Shlens, Wei Hua, Li-Jia Li, Li Fei-Fei, Alan Yuille, Jonathan Huang, Kevin Murphy

pdf:https://arxiv.org/abs/1712.00559

code: tensorflow-slim :

https://github.com/tensorflow/models/blob/master/research/slim/nets/nasnet/pnasnet.py

code: unofficial-pytorch :

https://github.com/chenxi116/PNASNet.pytorch

code: unofficial-tensorflow :

https://github.com/chenxi116/PNASNet.TF

32.AmoebaNet

Regularized Evolution for Image Classifier Architecture Search

Esteban Real, Alok Aggarwal, Yanping Huang, Quoc V Le

pdf:https://arxiv.org/abs/1802.01548

code: tensorflow-tpu :

https://github.com/tensorflow/tpu/tree/master/models/official/amoeba_net

33.SENet

Squeeze-and-Excitation Networks

Jie Hu, Li Shen, Samuel Albanie, Gang Sun, Enhua Wu

pdf:https://arxiv.org/abs/1709.01507

code: official :

https://github.com/hujie-frank/SENet

code: unofficial-pytorch :

https://github.com/moskomule/senet.pytorch

code: unofficial-tensorflow :

https://github.com/taki0112/SENet-Tensorflow

code: unofficial-caffe :

https://github.com/shicai/SENet-Caffe

code: unofficial-mxnet :

https://github.com/bruinxiong/SENet.mxnet

34.ShuffleNetV2

ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design

Ningning Ma, Xiangyu Zhang, Hai-Tao Zheng, Jian Sun

pdf:https://arxiv.org/abs/1807.11164

code: unofficial-pytorch :

https://github.com/Randl/ShuffleNetV2-pytorch

code: unofficial-keras :

https://github.com/opconty/keras-shufflenetV2

code: unofficial-pytorch :

https://github.com/Bugdragon/ShuffleNet_v2_PyTorch

code: unofficial-caff2:

https://github.com/wolegechu/ShuffleNetV2.Caffe2

35.IGCV3

IGCV3: Interleaved Low-Rank Group Convolutions for Efficient Deep Neural Networks

Ke Sun, Mingjie Li, Dong Liu, Jingdong Wang

pdf:https://arxiv.org/abs/1806.00178

code: official :

https://github.com/homles11/IGCV3

code: unofficial-pytorch :

https://github.com/xxradon/IGCV3-pytorch

code: unofficial-tensorflow :

https://github.com/ZHANG-SHI-CHANG/IGCV3

36.MNasNet

MnasNet: Platform-Aware Neural Architecture Search for Mobile

Mingxing Tan, Bo Chen, Ruoming Pang, Vijay Vasudevan, Quoc V. Le

pdf:https://arxiv.org/abs/1807.11626

code: unofficial-pytorch :

https://github.com/AnjieZheng/MnasNet-PyTorch

code: unofficial-caffe :

https://github.com/LiJianfei06/MnasNet-caffe

code: unofficial-MxNet :

https://github.com/chinakook/Mnasnet.MXNet

code: unofficial-keras :

https://github.com/Shathe/MNasNet-Keras-Tensorflow

責(zé)任編輯：lq