Open-Vocabulary Semantic Segmentation on ADE-150
33.5mIoUSCAN
Evaluation results
| Method | Links | ||
|---|---|---|---|
| SCAN2023.12 | 33.5 | 34.2 | |
| CAT-SegPrompt Tuning=MMRL+VaMP2025.11 | 33.3 | — | |
| CAT-SegPrompt Tuning=MMRL2025.11 | 32.4 | — | |
| SAN2023.12 | 32.1 | 33.7 | |
| CAT-SegPrompt Tuning=-2025.11 | 31.8 | — | |
| OVSeg2023.12 | 29.6 | 30.5 | |
| MAFT2023.12 | 29.1 | 30.3 | |
| SANPrompt Tuning=-2025.11 | 27.5 | — | |
| OVSegPrompt Tuning=-2025.11 | 24.8 | — | |
| SimSeg2023.12 | 20.5 | 22.6 | |
| CoM-PTBackbone=ViT-L/16, PT Method=CoM-PT, Fine-tuning Protocol=SAN [70]2026.04 | 20.14 | — | |
| CoM-PTBackbone=ViT-B/16, PT Method=CoM-PT, Fine-tuning Protocol=SAN [70]2026.04 | 18.3 | — | |
| ZegFormerPrompt Tuning=-2025.11 | 18 | — | |
| BaselineBackbone=ViT-L/16, PT Method=Baseline, Fine-tuning Protocol=SAN [70]2026.04 | 13.58 | — | |
| BaselineBackbone=ViT-B/16, PT Method=Baseline, Fine-tuning Protocol=SAN [70]2026.04 | 11.4 | — |