Official repository for the paper End-to-End Human Instance Matting
E2E-HIM is a human instance matting network.
GPU memory >= 6GB.
- torch >= 1.10
- numpy >= 1.16
- opencv-python >= 4.0
- einops >= 0.3.2
- timm >= 0.4.12
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 |
We provide the script eval_swintiny.py
for evaluation. Note that, current E2E-HIM cannot be applied to high-resolution images. Due to the commercial value of the method, we only provide inference code and do not intend to release the training code.
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}
}