huyuejingling / E2E-HIM

End-to-End Human Instance Matting (E2E-HIM)

Official repository for the paper End-to-End Human Instance Matting

Description

E2E-HIM is a human instance matting network.

Demovideo

Requirements

Hardware:

GPU memory >= 6GB.

Packages:

• torch >= 1.10
• numpy >= 1.16
• opencv-python >= 4.0
• einops >= 0.3.2
• timm >= 0.4.12

Models

The model can only be used and distributed for noncommercial purposes.

Quantitative results on HIM-100K.

Model Name	Size	EMSE	EMAD
E2E-HIM	270MiB	5.33	6.62

Evaluation

We provide the script eval_swintiny.py for evaluation. Note that, current E2E-HIM cannot be applied to high-resolution images.

Citation

If you use this model in your research, please cite this project to acknowledge its contribution.

@article{liu2023end,
  title={End-to-end human instance matting},
  author={Liu, Qinglin and Zhang, Shengping and Meng, Quanling and Zhong, Bineng and Liu, Peiqiang and Yao, Hongxun},
  journal={IEEE Transactions on Circuits and Systems for Video Technology},
  year={2023},
  publisher={IEEE}
}