Google colab link: https://colab.research.google.com/drive/1rlk-nOgIJa14UDNQLJyRxh3McoRLo3b0?usp=sharing
Download the MPV dataset from Image-based Multi-pose Virtual Try On and put the dataset under "./dataset/images/".
Select postive perspective images, create dataset split file 'data_pair.txt', and put it under "./dataset/".
Cloth mask. Use the removebg api or holistically-nested-edge-detection for the cloth mask, and put the mask in "./dataset/cloth_mask".
sh train.sh