Project page / Paper / Demo
Semantically Multi-modal Image Synthesis(CVPR2020).
Zhen Zhu, Zhiliang Xu, Ansheng You, Xiang Bai
- torch>=1.0.0
- torchvision
- dominate
- dill
- scikit-image
- tqdm
- opencv-python
DeepFashion
Note: We provide an example of DeepFashion Dataset. For some reason, that is slight different from the DeepFashion used in ours paper.
Cityscapes
Cityscapes can be downloaded at here
ADE20K
This dataset can be downloaded at here
Download the tar of the pretrained models from the Google Drive Folder. There are deepfashion.sh, cityscapes.sh and ade20k.sh in the scripts folder. Change the parameters like dataroot and so on, then comment or uncomment some code to test/train model. And you can specify the --test_mask for SMIS test.
Our code is based on the popular SPADE