Visual Anomaly Segmentation on RSDD (AP, F1-Max, IoUano)
56.2APAG-VAS
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| AG-VASBase Model=LLaVA-OneVision-7B2026.03 | 56.2 | 57.8 | 36.9 | |
| AG-VASBase Model=LLaVA-13B2026.03 | 54.3 | 57 | 35.9 | |
| LISA*Base Model=LLaVA-13B, Retrained=true2026.03 | 52.1 | 57.1 | 35.2 | |
| LISA*Base Model=LLaVA-OneVision-7B, Retrained=true2026.03 | 48.8 | 54.9 | 34.6 | |
| Bayes-PFL*Base Model=CLIP(ViT-L-14-336), Retrained=true2026.03 | 41.9 | 46.9 | 7.7 | |
| Bayes-PFLBase Model=CLIP(ViT-L-14-336)2026.03 | 39.1 | 45.2 | 3.7 | |
| LISABase Model=LLaVA-OneVision-7B2026.03 | 31.6 | 37.6 | 22.6 | |
| APRIL-GANBase Model=CLIP(ViT-L-14-336)2026.03 | 30.6 | 40.2 | 24.3 | |
| AnomalyCLIP*Base Model=CLIP(ViT-L-14-336), Retrained=true2026.03 | 25.5 | 34.4 | 8.5 | |
| AnomalyCLIPBase Model=CLIP(ViT-L-14-336)2026.03 | 19.1 | 30.6 | 8.8 | |
| LISABase Model=LLaVA-7B2026.03 | 6.9 | 12.5 | 2.9 | |
| PixelLMBase Model=LLaVA-13B2026.03 | 5.3 | 9.4 | 4.9 | |
| LISABase Model=LLaVA-13B2026.03 | 5.2 | 11.4 | 2.6 | |
| WinCLIPBase Model=CLIP(ViT-B-16-240)2026.03 | 2.1 | 6.6 | 0.1 | |
| PaDTBase Model=Qwen2.5-VL-7B2026.03 | 0.9 | 1.8 | 1.7 |