Open-vocabulary Segmentation on PASCAL Context
36.8mIoU+CS & AMF (GCLIP)
Evaluation Results
| Method | Links | |
|---|---|---|
| +CS & AMF (GCLIP)Pub. & Year=Ours, Setting=TF-OVSS2025.02 | 36.8 | |
| +CSPub. & Year=Ours, Setting=TF-OVSS2025.02 | 36.2 | |
| GEM‡Pub. & Year=CVPR’24, Setting=TF-OVSS2025.02 | 35.9 | |
| ClearCLIPPub. & Year=ECCV’24, Setting=TF-OVSS2025.02 | 35.9 | |
| CLIPtrasePub. & Year=ECCV’24, Setting=TF-OVSS2025.02 | 34.9 | |
| SCLIPPub. & Year=ECCV’24, Setting=TF-OVSS2025.02 | 34.2 | |
| ReCLIPPub. & Year=CVPR’24, Setting=USS2025.02 | 33.8 | |
| CLIP-S4Pub. & Year=CVPR’23, Setting=USS2025.02 | 33.6 | |
| INSIGHT_thresParadigm=Use Frozen CLIP with Interpretability, Background Prompt=Yes2026.01 | 33.2 | |
| CLIP-DINOiserParadigm=Use Frozen CLIP, Background Prompt=Yes, Protocol=Rerun (*)2026.01 | 32.5 | |
| MaskCLIP+†Pub. & Year=ECCV’22, Setting=USS2025.02 | 31.1 | |
| ProMaCVenue=Ours, Segmentation Annotation=-, Extra Text pairs=-2024.08 | 30.7 | |
| TCLPub. & Year=CVPR’23, Setting=T-OVSS2025.02 | 30.3 | |
| OVDiffParadigm=Build prototypes per class, Extra Backbones=DINO & SD, Background Prompt=Yes2026.01 | 30.1 | |
| INSIGHTParadigm=Use Frozen CLIP with Interpretability, Background Prompt=Yes2026.01 | 28.9 | |
| SegCLIPVenue=ICML23, Segmentation Annotation=COCO [35], Extra Text pairs=CC [48]2024.08 | 24.7 | |
| SegCLIPParadigm=Text/image alignment training with captions, Background Prompt=Yes2026.01 | 24.7 | |
| TCLVenue=CVPR23, Segmentation Annotation=-, Extra Text pairs=CC3M [48], CC12M [7]2024.08 | 24.3 | |
| TCLParadigm=Text/image alignment training with captions, Background Prompt=Yes2026.01 | 24.3 | |
| MaskCLIPParadigm=Use Frozen CLIP, Extra Backbones=ref[9], Background Prompt=Yes2026.01 | 24 | |
| MaskCLIPVenue=ECCV22, Segmentation Annotation=-, Extra Text pairs=-2024.08 | 23.6 | |
| GroupViT‡Pub. & Year=CVPR’22, Setting=T-OVSS2025.02 | 23.4 | |
| ViewCoVenue=ICLR23, Segmentation Annotation=-, Extra Text pairs=CC12M [7], YFCC14M [53]2024.08 | 23 | |
| MaskCLIPParadigm=Use Frozen CLIP, Background Prompt=Yes2026.01 | 22.9 | |
| GroupViTVenue=CVPR22, Segmentation Annotation=-, Extra Text pairs=CC12M [7], YFCC14M [53]2024.08 | 22.4 | |
| ZeroSegParadigm=Text/image alignment training with captions, Background Prompt=Yes2026.01 | 21.8 | |
| MaskCLIP†Pub. & Year=ECCV’22, Setting=TF-OVSS2025.02 | 21.7 | |
| OVSegmentorVenue=CVPR23, Segmentation Annotation=-, Extra Text pairs=CC12M [7]2024.08 | 20.4 | |
| OVSegmentorParadigm=Text/image alignment training with captions, Background Prompt=Yes2026.01 | 20.4 | |
| ReCoParadigm=Build prototypes per class, Background Prompt=Yes2026.01 | 19.9 | |
| CLIP-DIYParadigm=Use Frozen CLIP, Extra Backbones=DINO, Background Prompt=Yes2026.01 | 19.7 | |
| GroupViTParadigm=Text/image alignment training with captions, Background Prompt=Yes2026.01 | 18.7 | |
| CLIP‡Pub. & Year=ICML’21, Setting=TF-OVSS2025.02 | 9.2 |