DA-TransUNet: Integrating Positional and Channel Dual Attention with Transformer-Based U-Net for Enhanced Medical Image Segmentation (https://arxiv.org/abs/2310.12570)
Please use the preprocessed data for research purposes.
Please prepare an environment with python=3.7(conda create -n envir python=3.7.12), and then use the command "pip install -r requirements.txt" for the dependencies.
Run the train script on synapse dataset. The batch size can be reduced to 12 or 16 to save memory(please also decrease the base_lr linearly), and both can reach similar performance.
CUDA_VISIBLE_DEVICES=0 python train.py --dataset Synapse --vit_name R50-ViT-B_16
python test.py --dataset Synapse --vit_name R50-ViT-B_16
@article{sun2024transunet, title={DA-TransUNet: integrating spatial and channel dual attention with transformer U-net for medical image segmentation}, author={Sun, Guanqun and Pan, Yizhi and Kong, Weikun and Xu, Zichang and Ma, Jianhua and Racharak, Teeradaj and Nguyen, Le-Minh and Xin, Junyi}, journal={Frontiers in Bioengineering and Biotechnology}, volume={12}, pages={1398237}, year={2024}, publisher={Frontiers Media SA} }