Audio-Visual Semantic Segmentation on AVSBench-Semantic (AVSS) (test)
48.5Jaccard Index (J)AVSStone
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| AVSStoneBackbone=Swin-base, Method Stage=Two-stage2026.03 | 48.5 | 53.2 | 50.9 | |
| SDAVSBackbone=Swin-base, Method Stage=One-stage2026.03 | 46 | 51.1 | 48.9 | |
| SDAVSBackbone=PVT-v2, Method Stage=One-stage2026.03 | 44.8 | 50.2 | 47.5 | |
| BiasAVSBackbone=Swin-base, Method Stage=One-stage2026.03 | 44.4 | 49.9 | 47.2 | |
| COMBOBackbone=PVT-v2, Method Stage=One-stage2026.03 | 42.1 | 46.1 | 44.1 | |
| SelMBackbone=PVT-v2, Method Stage=One-stage2026.03 | 41.3 | 46.9 | 44.1 | |
| AVS-MambaBackbone=PVT-v2, Method Stage=One-stage2026.03 | 39.7 | 45.1 | 42.4 | |
| CQFormerBackbone=PVT-v2, Method Stage=One-stage2026.03 | 38.1 | 41 | 39.55 | |
| AVSegFormerBackbone=PVT-v2, Method Stage=One-stage2026.03 | 37.3 | 42.8 | 40.1 | |
| AVSACBackbone=PVT-v2, Method Stage=One-stage2026.03 | 37 | 42.4 | 39.7 | |
| BAVSBackbone=Swin-base, Method Stage=One-stage2026.03 | 33.6 | 37.5 | 35.6 | |
| CATRBackbone=PVT-v2, Method Stage=One-stage2026.03 | 32.8 | 38.5 | 35.7 | |
| TPAVIBackbone=PVT-v2, Method Stage=One-stage2026.03 | 29.8 | 35.2 | 32.5 |