CUPS

Reported on 6 benchmarks across 3 tasks · 1 paper · 1 SOTA

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

Computer Vision2 results

Unsupervised Semantic SegmentationonCityscapes test
mIoU· 2025-04-02
26.8
SOTA
Scene-Centric Unsupervised Panoptic Segmentation arXiv:2504.01955
Unsupervised Semantic SegmentationonCityscapes test
Accuracy· 2025-04-02
83.2
best: 84.3 (ViCE)
Scene-Centric Unsupervised Panoptic Segmentation arXiv:2504.01955

Medical2 results

Semantic SegmentationonCityscapes test
Accuracy· 2025-04-02
83.2
best: 84.3 (ViCE)
Scene-Centric Unsupervised Panoptic Segmentation arXiv:2504.01955
Semantic SegmentationonCityscapes test
mIoU· 2025-04-02
26.8
best: 67.3 (EDANet)
Scene-Centric Unsupervised Panoptic Segmentation arXiv:2504.01955

Audio2 results

10-shot image generationonCityscapes test
Accuracy· 2025-04-02
83.2
best: 84.3 (ViCE)
Scene-Centric Unsupervised Panoptic Segmentation arXiv:2504.01955
10-shot image generationonCityscapes test
mIoU· 2025-04-02
26.8
best: 67.3 (EDANet)
Scene-Centric Unsupervised Panoptic Segmentation arXiv:2504.01955