This is the official Pytorch implementation of "Hybrid Ladder Transformers with Efficient Parallel-Cross Attention for Medical Image Segmentation", Haozhe Luo et al. 2021
- Train the hybrid ladder transformer from scratch
- Train some of the other baseline models described in the paper
- v1.0
- Basic Pytorch dependency
- Tested on Pytorch 1.3, Python 3.6
- Download the data from: https://warwick.ac.uk/fac/cross_fac/tia/data/glascontest/ https://monuseg.grand-challenge.org/Data/
- Prepare the data:
--project
train.py
--data
--Train_Folder
--Test_Folder
- Train the model from scratch: python train.py
- Kindly cite our publication if you use any part of the code
@misc{haozhe2021LadderTransformer,
title={Hybrid Ladder Transformers with Efficient Parallel-Cross Attention for Medical Image Segmentation},
author={Haozhe Luo and Yu Changdong and Raghavendra Selvan},
howpublished={Openreview},
month={December},
note={Preprint on Openreview},
url={https://openreview.net/forum?id=swvVpnzro9q},
year={2021}}
- Haozhe Luo: luohaozhe@stu.scu.edu.cn
Baseline implementations were based on the following repositories:
