- Run following command to build docker image:
docker build -t . crgnn
Code has been tested on Python 3.8, cuda 11.4 and PyTorch 1.10.0.
- Run docker container using following command:
docker run -it --rm --gpus=all --ipc=host -v `pwd`/<data-folder>:/data crgnn bash
- Run test.py script:
python3 test.py --data-root ./examples --checkpoint ./checkpoint/e20.pth
- Please see the real data in the above link.
- Please contact Tiantian Wang ([email protected]) if you need composited data.
If you find this work or code useful for your research, please cite:
@inproceedings{wang2021crgnn,
title={Video Matting via Consistency-Regularized Graph Neural Networks},
author={Wang, Tiantian and Liu, Sifei and Tian, Yapeng and Li, Kai and Yang, Ming-Hsuan},
booktitle={Proc. IEEE/CVF International Conference on Computer Vision (ICCV)},
year={2021}
}
This code is only for non-commercial purposes. As covered by the ADOBE IMAGE DATASET LICENSE AGREEMENT, the trained models included in this repository can only be used/distributed for non-commercial purposes. Anyone who violates this rule will be at his/her own risk.