SOTAVerified

Unsupervised Semantic Segmentation

Models that learn to segment each image (i.e. assign a class to every pixel) without seeing the ground truth labels.

( Image credit: SegSort: Segmentation by Discriminative Sorting of Segments )

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 1–10 of 95 papers

TitleStatusHype
Urban1960SatSeg: Unsupervised Semantic Segmentation of Mid-20^th century Urban Landscapes with Satellite ImageriesCode2
LogoSP: Local-global Grouping of Superpoints for Unsupervised Semantic Segmentation of 3D Point CloudsCode1
Federated Unsupervised Semantic Segmentation—0
Hierarchical Context Learning of object components for unsupervised semantic segmentationCode0
Scene-Centric Unsupervised Panoptic SegmentationCode2
COIN: Confidence Score-Guided Distillation for Annotation-Free Cell SegmentationCode0
NPSim: Nighttime Photorealistic Simulation From Daytime Images With Monocular Inverse Rendering and Ray Tracing—0
GraPix: Exploring Graph Modularity Optimization for Unsupervised Pixel ClusteringCode0
ULTra: Unveiling Latent Token Interpretability in Transformer Based Understanding—0
Unsupervised semantic segmentation of urban high-density multispectral point clouds—0
Show:102550
← PrevPage 1 of 10Next →
All datasetsCOCO-Stuff-27Cityscapes testPASCAL VOC 2012 valPotsdam-3COCO-Stuff-3ImageNet-S-50COCO-Stuff-171COCO-Stuff-81SUIMCOCO-Stuff-15ACDC (Adverse Conditions Dataset with Correspondences)Cityscapes val

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1DatUS (ViT-B/8) + OCPixel Accuracy69.98—Unverified
2GraPix + AUTPixel Accuracy65.48—Unverified
3DatUS (ViT-B/8)Pixel Accuracy64.67—Unverified
4GraPixPixel Accuracy64.06—Unverified