This is the PyTorch implementation of the paper Channel Equilibrium Networks for Learning Deep Representation, ICML2020.

By Wenqi Shao, Shitao Tang, Xingang Pan, Ping Tan, Xiaogang Wang, Ping Luo.

[https://arxiv.org/pdf/2003.00214.pdf)

We design a novel neural building block, termed Channel Equilibrium (CE) block, which enables channels at the same layer to contribute equally to the learned representation.

Comparisons of top-1 accuracies on the validation set of ImageNet, by using ResNet50 trained with BN and CE.

• Install PyTorch
• Clone the repo:

  git clone https://github.com/Tangshitao/CENet.git
  

• python packages
  • pytorch>=0.4.0
  • torchvision>=0.2.1
  • tensorboardX
  • pyyaml

• Download the ImageNet dataset and put them into the {repo_root}/data/imagenet.

./train.sh

Number of GPUs and configuration file to use can be modified in train.sh

Download the pretrained models from Model Zoo and put them into the {repo_root}/model_zoo

./test.sh

Or you can specify the checkpoint path by modifying test.sh

--checkpoint_path model_zoo/ssn_8x2_75.848.pth \

We provide models pretrained with CE block on ImageNet.

*single-crop validation accuracy on ImageNet (a 224x224 center crop from resized image with shorter side=256)

In evaluation, download the above models and put them into the {repo_root}/model_zoo.

If you find this work helpful in your project or use our model zoo, please consider citing:

@article{shao2020channel,
  title={Channel equilibrium networks for learning deep representation},
  author={Shao, Wenqi and Tang, Shitao and Pan, Xingang and Tan, Ping and Wang, Xiaogang and Luo, Ping},
  journal={arXiv preprint arXiv:2003.00214},
  year={2020}
}