Image Matting On P3M 10K
평가 지표
MAD
MSE
SAD
평가 결과
이 벤치마크에서 각 모델의 성능 결과
모델 이름 | MAD | MSE | SAD | Paper Title | Repository |
|---|---|---|---|---|---|
| P3M-Net (v) | 0.0036 | 0.0015 | 6.24 | Rethinking Portrait Matting with Privacy Preserving | |
| P3M-Net (r) | 0.0051 | 0.0026 | 8.73 | Privacy-Preserving Portrait Matting | |
| LF | 0.0250 | 0.0191 | 42.95 | A Late Fusion CNN for Digital Matting | |
| GFM | 0.0080 | 0.0050 | 13.20 | Bridging Composite and Real: Towards End-to-end Deep Image Matting | |
| HATT | 0.0152 | 0.0054 | 25.99 | Attention-Guided Hierarchical Structure Aggregation for Image Matting | |
| SHM | 0.0125 | 0.0100 | 21.56 | Semantic Human Matting | |
| StyleMatte | 0.004 | 0.0019 | 6.97 | Adversarially-Guided Portrait Matting |
0 of 7 row(s) selected.