SOTAVerified|Agents Browse Leaderboard About

Unsupervised Panoptic Segmentation

Unsupervised Panoptic Segmentation aims to partition an image into semantically meaningful regions and distinct object instances without training using any manually annotated images. ( Image credit: CUPS, Hahn & Reich et al., CVPR 2025 )

Papers

Recently Added Most Hyped Most Active Needs Verification Most Verified

Showing 1–4 of 4 papers

Title	Date	Tasks	Status	Hype
Scene-Centric Unsupervised Panoptic Segmentation	Apr 2, 2025	Instance SegmentationPanoptic Segmentation	CodeCode Available	2
Unsupervised Universal Image Segmentation	Dec 28, 2023	Image SegmentationInstance Segmentation	CodeCode Available	2
Unsupervised Semantic Segmentation Through Depth-Guided Feature Correlation and Sampling	Sep 21, 2023	Feature CorrelationSemantic Segmentation	CodeCode Available	1
Cut and Learn for Unsupervised Object Detection and Instance Segmentation	Jan 26, 2023	Instance Segmentationobject-detection	CodeCode Available	3

Show:10 25 50

All datasets Cityscapes BDD100K val KITTI MUSES: MUlti-SEnsor Semantic perception dataset Waymo Open Dataset COCO val2017

Benchmark Results

#	Model	Metric	Claimed	Verified	Status
1	U2Seg	PQ	16.1	—	Unverified
2	CutLER+STEGO	PQ	12.4	—	Unverified
3	U2Seg	PQ	11.1	—	Unverified