ZegCLIP

Reported on 4 benchmarks across 1 task · 1 paper · 4 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Computer Vision4 results

Zero-Shot Semantic SegmentationonPASCAL VOC
Inductive Setting hIoU· 2022-12-07
84.3
best: 88.4 (CLIP-RC)
SOTA
ZegCLIP: Towards Adapting CLIP for Zero-shot Semantic Segmentation arXiv:2212.03588
Zero-Shot Semantic SegmentationonPASCAL VOC
Transductive Setting hIoU· 2022-12-07
91.1
best: 94.4 (OTSeg+)
SOTA
ZegCLIP: Towards Adapting CLIP for Zero-shot Semantic Segmentation arXiv:2212.03588
Zero-Shot Semantic SegmentationonCOCO-Stuff
Inductive Setting hIoU· 2022-12-07
40.8
best: 41.5 (OTSeg+)
SOTA
ZegCLIP: Towards Adapting CLIP for Zero-shot Semantic Segmentation arXiv:2212.03588
Zero-Shot Semantic SegmentationonCOCO-Stuff
Transductive Setting hIoU· 2022-12-07
48.5
best: 49.8 (OTSeg+)
SOTA
ZegCLIP: Towards Adapting CLIP for Zero-shot Semantic Segmentation arXiv:2212.03588