| --- |
| license: other |
| license_name: s-lab-1.0 |
| license_link: https://huggingface.co/AEmotionStudio/matanyone2/blob/main/LICENSE.txt |
| tags: |
| - video-matting |
| - alpha-matting |
| - comfyui |
| - ffmpega |
| --- |
| |
| # MatAnyone2 — SafeTensors Mirror |
|
|
| This is a **SafeTensors conversion** of the [MatAnyone2](https://github.com/pq-yang/MatAnyone2) video matting model (CVPR 2026) for use with [ComfyUI-FFMPEGA](https://github.com/AEmotionStudio/ComfyUI-FFMPEGA). |
|
|
| ## Model Details |
|
|
| | Property | Value | |
| |----------|-------| |
| | Architecture | MatAnyone2 (temporal video matting) | |
| | Parameters | 562 tensors, ~135 MB | |
| | Format | SafeTensors | |
| | Original Format | PyTorch (.pth) | |
| | Source | [pq-yang/MatAnyone2](https://github.com/pq-yang/MatAnyone2) | |
|
|
| ## ⚠️ License Notice |
|
|
| This model is released under the **NTU S-Lab License 1.0** which is strictly **non-commercial**. |
|
|
| > Redistribution and use for non-commercial purpose in source and binary forms, with or without modification, are permitted provided that the following conditions are met... |
|
|
| See [LICENSE.txt](LICENSE.txt) for the full license text. **Commercial use requires explicit written permission from the original authors.** |
|
|
| ## Usage |
|
|
| This model is automatically downloaded by ComfyUI-FFMPEGA when the `video_matting` no-LLM mode is selected. No manual setup is required. |
|
|
| ## Citation |
|
|
| ```bibtex |
| @article{yang2025matanyone2, |
| title={MatAnyone2: Stable Video Matting with Consistent Memory Propagation}, |
| author={Yang, Peiqing and Shang, Shangchen and others}, |
| year={2025} |
| } |
| ``` |
|
|
