MatAnyone: Stable Video Matting with Consistent Memory Propagation
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866915212773294080 |
|---|---|
| author | Yang, Peiqing Zhou, Shangchen Zhao, Jixin Tao, Qingyi Loy, Chen Change |
| author_facet | Yang, Peiqing Zhou, Shangchen Zhao, Jixin Tao, Qingyi Loy, Chen Change |
| contents | Auxiliary-free human video matting methods, which rely solely on input frames, often struggle with complex or ambiguous backgrounds. To address this, we propose MatAnyone, a robust framework tailored for target-assigned video matting. Specifically, building on a memory-based paradigm, we introduce a consistent memory propagation module via region-adaptive memory fusion, which adaptively integrates memory from the previous frame. This ensures semantic stability in core regions while preserving fine-grained details along object boundaries. For robust training, we present a larger, high-quality, and diverse dataset for video matting. Additionally, we incorporate a novel training strategy that efficiently leverages large-scale segmentation data, boosting matting stability. With this new network design, dataset, and training strategy, MatAnyone delivers robust and accurate video matting results in diverse real-world scenarios, outperforming existing methods. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_14677 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | MatAnyone: Stable Video Matting with Consistent Memory Propagation Yang, Peiqing Zhou, Shangchen Zhao, Jixin Tao, Qingyi Loy, Chen Change Computer Vision and Pattern Recognition Auxiliary-free human video matting methods, which rely solely on input frames, often struggle with complex or ambiguous backgrounds. To address this, we propose MatAnyone, a robust framework tailored for target-assigned video matting. Specifically, building on a memory-based paradigm, we introduce a consistent memory propagation module via region-adaptive memory fusion, which adaptively integrates memory from the previous frame. This ensures semantic stability in core regions while preserving fine-grained details along object boundaries. For robust training, we present a larger, high-quality, and diverse dataset for video matting. Additionally, we incorporate a novel training strategy that efficiently leverages large-scale segmentation data, boosting matting stability. With this new network design, dataset, and training strategy, MatAnyone delivers robust and accurate video matting results in diverse real-world scenarios, outperforming existing methods. |
| title | MatAnyone: Stable Video Matting with Consistent Memory Propagation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2501.14677 |