FreeSeg

Reported on 3 benchmarks across 2 tasks · 2 papers · 1 SOTA

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

Computer Vision3 results

Zero-Shot Semantic SegmentationonCOCO-Stuff
Transductive Setting hIoU· 2022-09-27
45.3
best: 49.8 (OTSeg+)
SOTA
FreeSeg: Free Mask from Interpretable Contrastive Language-Image Pretraining for Semantic Segmentation arXiv:2209.13558
Open Vocabulary Panoptic SegmentationonADE20K
PQ· 2023-03-30
16.3
best: 31.6 (UMG-CLIP-E/14)
FreeSeg: Unified, Universal and Open-Vocabulary Image Segmentation arXiv:2303.17225
Zero-Shot Semantic SegmentationonPASCAL VOC
Transductive Setting hIoU· 2022-09-27
86.9
best: 94.4 (OTSeg+)
FreeSeg: Free Mask from Interpretable Contrastive Language-Image Pretraining for Semantic Segmentation arXiv:2209.13558