DilateFormer

Official PyTorch implementation of IEEE Transaction on Multimedia 2023 paper “DilateFormer: Multi-Scale Dilated Transformer for Visual Recognition” . [paper] [Project Page]

We currenent release the pytorch version code for:

• ImageNet-1K training

Image classification

Our repository is built base on the DeiT repository, but we add some useful features:

1. Calculating accurate FLOPs and parameters with fvcore (see check_model.py).
2. Auto-resuming.
3. Saving best models and backup models.
4. Generating training curve (see generate_tensorboard.py).

Installation

• Install PyTorch 1.7.0+ and torchvision 0.8.1+

  conda install -c pytorch pytorch torchvision

• Install other packages

  pip install timm
  pip install fvcore

Training

Simply run the training scripts as followed, and take dilateformer_tiny as example:

bash dist_train.sh dilateformer_tiny [other prams]

If the training was interrupted abnormally, you can simply rerun the script for auto-resuming. Sometimes the checkpoint may not be saved properly, you should set the resumed model via --reusme ${work_path}/ckpt/backup.pth.

Generate curves

You can generate the training curves as followed:

python3 generate_tensoboard.py

Note that you should install tensorboardX.

Calculating FLOPs and Parameters

You can calculate the FLOPs and parameters via:

python3 check_model.py

Acknowledgement

This repository is built using the timm library and the DeiT repository.

Citation

If you use this code for a paper, please cite:

DilateFormer

@article{jiao2023dilateformer,
title = {DilateFormer: Multi-Scale Dilated Transformer for Visual Recognition},
author = {Jiao, Jiayu and Tang, Yu-Ming and Lin, Kun-Yu and Gao, Yipeng and Ma, Jinhua and Wang, Yaowei and Zheng, Wei-Shi},
journal = {{IEEE} Transaction on Multimedia},
year = {2023}
}