Code for "Multi-scale Dynamic Feature Encoding Network for Image Demoireing" [ICCV2019 AIM Workshop][arxiv]
Team: MoePhoto Team Members: Xi Cheng, Zhenyong Fu, Jian Yang
- GPU>= 11GB Memory (We have tested on RTX2080Ti/Titan V/Tesla P40)
- Python 3.6.6
- Pytorch==1.2.0
- torchvision==0.4.0a0+6b959ee
- scikit-image==0.13.1
Put the images in ./data/TestingMoire
Run (Fidelity)
python test_all.py --cuda --model ./models/mddm.pth
Run (Perceptual)
python test_all.py --cuda --model ./models/mddmgan.pth
[WIP]