Audio-Visual Segmentation on AVS-Bench
58.4MS3AVSegFormer
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| AVSegFormerBackbone=PVT-v22025.03 | 58.4 | 24.9 | |
| CrabBackbone=ViT/L-142025.03 | 58.21 | 26.59 | |
| AVSBenchBackbone=ResNet-502025.03 | 54 | — | |
| TPAVIBackbone=PVT-v22025.03 | 54 | 29.8 |
Loading the SOTA2 catalog…