Audio-Visual Instance Segmentation on AVISeg
54.65FSLASeaVIS
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| SeaVISBackbone=Swin-L, fps=19.392026.03 | 54.65 | 73.85 | 54.29 | |
| AVISMBackbone=Swin-L, fps=14.212026.03 | 52.49 | 71.13 | 53.46 | |
| SeaVISBackbone=Res-50, fps=34.652026.03 | 47.09 | 66.47 | 46.28 | |
| AVISMBackbone=Res-50, fps=20.462026.03 | 44.42 | 64.52 | 45.04 | |
| GenVIS-onlineBackbone=Res-50, fps=25.62026.03 | 12.31 | 24.97 | 25.16 | |
| IDOLBackbone=Res-50, fps=23.712026.03 | 7.66 | 46.42 | 45.18 | |
| MinVISBackbone=Res-50, fps=50.382026.03 | 2.45 | 35.13 | 44.66 | |
| DVIS-onlineBackbone=Res-50, fps=39.572026.03 | 0.03 | 29.97 | 44.29 |