Semantic Segmentation on PASCAL VOC with background class
84mIoULSMSeg
Evaluation Results
| Method | Links | |
|---|---|---|
| LSMSegVLM=CLIP ViT-L/14, Training Dataset=COCO-Stuff2024.11 | 84 | |
| CAT-SegVLM=CLIP ViT-L/14, Training Dataset=COCO-Stuff2024.11 | 82.5 | |
| FC-CLIPVLM=ConvNeXt-L, Training Dataset=COCO Panoptic2024.11 | 81.8 | |
| LSMSegVLM=CLIP ViT-B/16, Training Dataset=COCO-Stuff2024.11 | 81.1 | |
| CAT-SegVLM=CLIP ViT-B/16, Training Dataset=COCO-Stuff2024.11 | 77.3 | |
| OpenSegVLM=ALIGN EN-B7, Training Dataset=COCO Panoptic2024.11 | 70.2 | |
| ZegFormerVLM=CLIP ViT-B/16, Training Dataset=COCO-Stuff2024.11 | 65.5 | |
| TCLEvaluation Mode=Zero-shot, Backbone=CLIP ViT-B/162022.12 | 55 | |
| GroupViT (RedCaps)Training Data=RedCaps + CC12M, Evaluation Mode=Zero-shot2022.12 | 50.4 | |
| GroupViT (YFCC)Training Data=YFCC + CC12M, Evaluation Mode=Zero-shot2022.12 | 49.5 | |
| MaskCLIPRefinement=With heuristic refinement, Evaluation Mode=Zero-shot2022.12 | 38.8 | |
| MaskCLIP (Baseline)Refinement=None, Evaluation Mode=Zero-shot2022.12 | 29.3 | |
| ReCoEvaluation Mode=Zero-shot2022.12 | 25.1 |