Skip to content

The codes for the work "Swin-Unet: Unet-like Pure Transformer for Medical Image Segmentation"

Notifications You must be signed in to change notification settings

Cufix/Swin-Unet

 
 

Repository files navigation

Swin-Unet

The codes for the work "Swin-Unet: Unet-like Pure Transformer for Medical Image Segmentation"(https://arxiv.org/abs/2105.05537). A validation for U-shaped Swin Transformer.

1. Download pre-trained swin transformer model (Swin-T)

2. Prepare data

  • The datasets we used are provided by TransUnet's authors. Please go to "./datasets/README.md" for details, or please send an Email to jienengchen01 AT gmail.com to request the preprocessed data. If you would like to use the preprocessed data, please use it for research purposes and do not redistribute it (following the TransUnet's License).

3. Environment

  • Please prepare an environment with python=3.7, and then use the command "pip install -r requirements.txt" for the dependencies.

4. Train/Test

  • Run the train script on synapse dataset. The batch size we used is 24. If you do not have enough GPU memory, the bacth size can be reduced to 12 or 6 to save memory.

  • Train

sh train.sh or python train.py --dataset Synapse --cfg configs/swin_tiny_patch4_window7_224_lite.yaml --root_path your DATA_DIR --max_epochs 150 --output_dir your OUT_DIR  --img_size 224 --base_lr 0.05 --batch_size 24
  • Test
sh test.sh or python test.py --dataset Synapse --cfg configs/swin_tiny_patch4_window7_224_lite.yaml --is_saveni --volume_path your DATA_DIR --output_dir your OUT_DIR --max_epoch 150 --base_lr 0.05 --img_size 224 --batch_size 24

References

Citation

@misc{cao2021swinunet,
      title={Swin-Unet: Unet-like Pure Transformer for Medical Image Segmentation}, 
      author={Hu Cao and Yueyue Wang and Joy Chen and Dongsheng Jiang and Xiaopeng Zhang and Qi Tian and Manning Wang},
      year={2021},
      eprint={2105.05537},
      archivePrefix={arXiv},
      primaryClass={eess.IV}
}

About

The codes for the work "Swin-Unet: Unet-like Pure Transformer for Medical Image Segmentation"

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 97.6%
  • Shell 2.4%