https://github.com/gbup-group/IEBN

The official implementation of paper "Instance Enhancement Batch Normalization: an Adaptive Regulator of Batch Noise".
https://github.com/gbup-group/IEBN

Last synced: about 1 month ago
JSON representation

The official implementation of paper "Instance Enhancement Batch Normalization: an Adaptive Regulator of Batch Noise".

• Host: GitHub
• URL: https://github.com/gbup-group/IEBN
• Owner: gbup-group
• License: mit
• Created: 2019-08-08T14:01:46.000Z (over 5 years ago)
• Default Branch: master
• Last Pushed: 2023-06-16T02:38:03.000Z (over 1 year ago)
• Last Synced: 2024-08-03T19:08:17.700Z (5 months ago)
• Language: Python
• Homepage:
• Size: 152 KB
• Stars: 40
• Watchers: 4
• Forks: 8
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

• awesome-normalization-techniques - [Python Reference

README

        
# Instance Enhancement Batch Normalization: an Adaptive Regulator of Batch Noise

![GitHub](https://img.shields.io/github/license/gbup-group/DIANet.svg)

![GitHub](https://img.shields.io/badge/gbup-%E7%A8%B3%E4%BD%8F-blue.svg)

By [Senwei Liang*](https://leungsamwai.github.io), [Zhongzhan Huang*](https://github.com/dedekinds) (* contribute equally), [Mingfu Liang](https://wuyujack.github.io/) and [Haizhao Yang](https://haizhaoyang.github.io/).

This repository is the implementation of "Instance Enhancement Batch Normalization: an Adaptive Regulator of Batch Noise" [[paper]](https://arxiv.org/abs/1908.04008)  on CIFAR-100 dataset. Our paper has been accepted for presentation at the Thirty-Fourth AAAI Conference on Artificial Intelligence (AAAI-20). You can also check with the [AAAI proceeding version](https://aaai.org/ojs/index.php/AAAI/article/view/5917).

## Introduction

Instance Enhancement Batch Normalization (IEBN) is an attention-based version of BN which recalibrates channel information of BN by a simple linear transformation.



  



## Requirement

* Python 3.6 and [PyTorch 1.0](http://pytorch.org/)

## Usage

  ```

python cifar.py -a iebn_resnet --dataset cifar100 --block-name bottleneck --depth 164 --epochs 164 --schedule 81 122 --gamma 0.1 --wd 1e-4 --checkpoint checkpoints/cifar100/resnet-164-iebn

  ```

## Results

|                 | original |  IEBN  |

|:---------------:|:--------:|:------:|

|    ResNet164    |   74.29  |  77.09 |

**Notes:**

- Training on 2 GPUs

## Citing IEBN

```

@inproceedings{liang2020instance,

  title={Instance Enhancement Batch Normalization: An Adaptive Regulator of Batch Noise.},

  author={Liang, Senwei and Huang, Zhongzhan and Liang, Mingfu and Yang, Haizhao},

  booktitle={AAAI},

  pages={4819--4827},

  year={2020}

}

```

## Acknowledgments

Many thanks to [bearpaw](https://github.com/bearpaw) for his simple and clean [Pytorch framework](https://github.com/bearpaw/pytorch-classification) for image classification task.