hzhang57 / ibn-net

Instance-Batch Normalization Networks (ECCV2018)

Instance-Batch Normalization Network

Paper

Xingang Pan, Ping Luo, Jianping Shi, Xiaoou Tang. "Two at Once: Enhancing Learning and Generalization Capacities via IBN-Net", ECCV2018.

Introduction

• IBN-Net carefully unifies instance normalization and batch normalization in a single deep network.
• It provides an extremely simple way to increase both modeling and generalization capacity without adding model complexity.

Requirements

• Pytorch 0.3.1

Results

Top1/Top5 error on the ImageNet validation set are reported. You may get different results when training your models with different random seed.

Model	origin	re-implementation	IBN-Net
DenseNet-121	25.0/-	24.96/7.85	24.47/7.25
DenseNet-169	23.6/-	24.02/7.06	23.25/6.51
ResNet-50	24.7/7.8	24.27/7.08	22.54/6.32
ResNet-101	23.6/7.1	22.48/6.23	21.39/5.59
ResNeXt-101	21.2/5.6	21.31/5.74	20.88/5.42
SE-ResNet-101	22.38/6.07	21.68/5.88	21.25/5.51

Before Start

1. Clone the repository

   git clone https://github.com/XingangPan/IBN-Net.git

2. Download ImageNet dataset (if you need to test or train on ImageNet). You may follow the instruction at fb.resnet.torch to process the validation set.

Testing

1. Download our pre-trained models from Google Drive and save them to ./pretrained.
2. Edit test.sh. Modify model and data_path to yours.
   Options for model: densenet121_ibn_a, densenet169_ibn_a, resnet50_ibn_a_old, resnet50_ibn_a, resnet50_ibn_b, resnet101_ibn_a_old, resnet101_ibn_a, resnext101_ibn_a, se_resnet101_ibn_a.
   (Note: For IBN-Net version of ResNet-50 and ResNet-101, our results in the paper are reported based on an slower implementation, corresponding to resnet50_ibn_a_old and resnet101_ibn_a_old here. We also provide a faster implementation, and the models are resnet50_ibn_a, resnet101_ibn_a, and all the rest. The top1/top5 error for resnet50_ibn_a and resnet101_ibn_a are 22.76/6.41 and 5.61/21.29 respectively.)
3. Run test script

   sh test.sh

Training

1. Edit train.sh. Modify model and data_path to yours.
2. Run train script

   sh train.sh

This code is modified from bearpaw/pytorch-classification.

Citing IBN-Net

@inproceedings{pan2018IBN-Net,  
  author = {Xingang Pan, Ping Luo, Jianping Shi, and Xiaoou Tang},  
  title = {Two at Once: Enhancing Learning and Generalization Capacities via IBN-Net},  
  booktitle = {ECCV},  
  year = {2018}  
}