Audio-Visual Segmentation on AVS-Bench S4 (test)
82.1mIoUAVSegFormer
Evaluation results
| Method | Links | ||
|---|---|---|---|
| AVSegFormerVisual Backbone=PVT-v22025.03 | 82.1 | 89.9 | |
| TPAVIVisual Backbone=PVT-v22025.03 | 78.7 | 87.9 | |
| AVSBenchVisual Backbone=PVT-v22025.03 | 78.7 | 87.9 | |
| BAVSVisual Backbone=ResNet-502025.03 | 78 | 85.3 | |
| LGVTVisual Backbone=Swin-T2025.03 | 74.9 | 87.3 | |
| CATRVisual Backbone=ResNet-502025.03 | 74.8 | 86.6 | |
| CrabVisual Backbone=ViT-L/142025.03 | 73.25 | 86.81 | |
| TPAVIVisual Backbone=ResNet-502025.03 | 72.8 | 84.8 | |
| iGANVisual Backbone=Swin-T2025.03 | 61.6 | 77.8 |