jackweiwang / VideoMatting-CRGNN

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Video Matting via Consistency-Regularized Graph Neural Networks

Project Page | Real Data | Paper


Our code has been tested on Python 3.7, cuda 10.1 and PyTorch 1.4.0.

pip install -r requirements.txt
# install dcn
cd models/archs/dcn
python setup.py develop


Run the following command to do inference of CRGNN on the video matting dataset:

python test.py


  1. Please see the real data in the above link.
  2. Please contact Tiantian Wang (tiantianwang.ice@gmail.com) if you need composited data.


If you find this work or code useful for your research, please cite:

  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)},

Permission and Disclaimer

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.



Language:Python 75.3%Language:Cuda 14.6%Language:C++ 10.1%