Unsupervised Panoptic Segmentation

8 benchmarks4 papers

Unsupervised Panoptic Segmentation aims to partition an image into semantically meaningful regions and distinct object instances without training using any manually annotated images. <span style="color:grey; opacity: 0.6">( Image credit: CUPS, Hahn & Reich et al., CVPR 2025 )</span>

Benchmarks

Unsupervised Panoptic Segmentation on Cityscapes

PQ

Unsupervised Panoptic Segmentation on BDD100K val

PQ

Unsupervised Panoptic Segmentation on KITTI

PQ

Unsupervised Panoptic Segmentation on MUSES: MUlti-SEnsor Semantic perception dataset

PQ

Unsupervised Panoptic Segmentation on Waymo Open Dataset

PQ

Unsupervised Panoptic Segmentation on COCO val2017

PQSQRQ