MCL

This code implements MCL model, described in the paper Multi-feature Contrastive Learning for Unpaired Image-to-Image Translation, Yao Gou, Min Li, Yu Song, Yujie He, LiTao Wang, CAIS, 2022.

The code borrows heavily from the PyTorch implementation of CycleGAN and CUT:

https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix

https://github.com/taesungp/contrastive-unpaired-translation

Our MCL is inspired by ContraD

https://github.com/jh-jeong/ContraD

Train the MCL model: python train.py --dataroot ./datasets/yourdata --name yourname_MCL --IMCL_mode imcl

Train the sinmcl model: python train.py --dataroot ./datasets/yourdata --name yourname_MCL --model sinmcl