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

Showing 150 of 95 papers

TitleStatusHype
Scene-Centric Unsupervised Panoptic SegmentationCode2
Urban1960SatSeg: Unsupervised Semantic Segmentation of Mid-20^th century Urban Landscapes with Satellite ImageriesCode2
What the DAAM: Interpreting Stable Diffusion Using Cross AttentionCode2
Deep Spectral Methods: A Surprisingly Strong Baseline for Unsupervised Semantic Segmentation and LocalizationCode2
Unsupervised Universal Image SegmentationCode2
DiffCut: Catalyzing Zero-Shot Semantic Segmentation with Diffusion Features and Recursive Normalized CutCode2
ReCLIP++: Learn to Rectify the Bias of CLIP for Unsupervised Semantic SegmentationCode2
Learn to Rectify the Bias of CLIP for Unsupervised Semantic SegmentationCode2
Unsupervised Semantic Segmentation by Distilling Feature CorrespondencesCode2
EAGLE: Eigen Aggregation Learning for Object-Centric Unsupervised Semantic SegmentationCode2
Discovering Object Masks with Transformers for Unsupervised Semantic SegmentationCode1
Drive&Segment: Unsupervised Semantic Segmentation of Urban Scenes via Cross-modal DistillationCode1
Unsupervised Portrait Shadow Removal via Generative PriorsCode1
Attention-based Transformation from Latent Features to Point CloudsCode1
Autoregressive Unsupervised Image SegmentationCode1
Boosting Unsupervised Semantic Segmentation with Principal Mask ProposalsCode1
Causal Unsupervised Semantic SegmentationCode1
Component-aware anomaly detection framework for adjustable and logical industrial visual inspectionCode1
CrOC: Cross-View Online Clustering for Dense Visual Representation LearningCode1
Deep Clustering for Unsupervised Learning of Visual FeaturesCode1
Deep ContourFlow: Advancing Active Contours with Deep LearningCode1
Dense Siamese Network for Dense Unsupervised LearningCode1
SERE: Exploring Feature Self-relation for Self-supervised TransformerCode1
GrowSP: Unsupervised Semantic Segmentation of 3D Point CloudsCode1
Invariant Information Clustering for Unsupervised Image Classification and SegmentationCode1
iSeg: An Iterative Refinement-based Framework for Training-free SegmentationCode1
Large-scale Unsupervised Semantic SegmentationCode1
Leveraging Hidden Positives for Unsupervised Semantic SegmentationCode1
LogoSP: Local-global Grouping of Superpoints for Unsupervised Semantic Segmentation of 3D Point CloudsCode1
Morphology-inspired Unsupervised Gland Segmentation via Selective Semantic GroupingCode1
Mumford-Shah Loss Functional for Image Segmentation with Deep LearningCode1
Multiple Fusion Adaptation: A Strong Framework for Unsupervised Semantic Segmentation AdaptationCode1
Optical Flow boosts Unsupervised Localization and SegmentationCode1
Peekaboo: Text to Image Diffusion Models are Zero-Shot SegmentorsCode1
Perceptual Grouping in Contrastive Vision-Language ModelsCode1
PiCIE: Unsupervised Semantic Segmentation using Invariance and Equivariance in ClusteringCode1
PointDC: Unsupervised Semantic Segmentation of 3D Point Clouds via Cross-Modal Distillation and Super-Voxel ClusteringCode1
Progressive Proxy Anchor Propagation for Unsupervised Semantic SegmentationCode1
ReCo: Retrieve and Co-segment for Zero-shot TransferCode1
Rectifying Pseudo Label Learning via Uncertainty Estimation for Domain Adaptive Semantic SegmentationCode1
Self-Supervised Learning of Object Parts for Semantic SegmentationCode1
Self-Supervised Visual Representation Learning with Semantic GroupingCode1
Semantic-Guided Zero-Shot Learning for Low-Light Image/Video EnhancementCode1
SmooSeg: Smoothness Prior for Unsupervised Semantic SegmentationCode1
Unsupervised Semantic Segmentation Through Depth-Guided Feature Correlation and SamplingCode1
Time Does Tell: Self-Supervised Time-Tuning of Dense Image RepresentationsCode1
TransFGU: A Top-down Approach to Fine-Grained Unsupervised Semantic SegmentationCode1
Uncovering the Inner Workings of STEGO for Safe Unsupervised Semantic SegmentationCode1
Unsupervised Hierarchical Semantic Segmentation with Multiview Cosegmentation and Clustering TransformersCode1
Unsupervised Object Localization: Observing the Background to Discover ObjectsCode1
Show:102550
← PrevPage 1 of 2Next →

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1SGSegClustering [Accuracy]55.7Unverified
2DynaSeg - FSF (ResNet-18 FPN)Clustering [mIoU]54.1Unverified
3SANClustering [Accuracy]52Unverified
4DiffCutClustering [mIoU]49.1Unverified
5PiCIEClustering [Accuracy]48.1Unverified
6HCL (ViT-S/8)Linear Classifier [mIoU]47.4Unverified
7HCL (ViT-S/16)Linear Classifier [mIoU]45.8Unverified
8CAUSE (DINOv2, ViT-B/14)Clustering [mIoU]45.3Unverified
9Ours (SlotCon)Clustering [Accuracy]42.36Unverified
10CAUSE (ViT-B/8)Clustering [mIoU]41.9Unverified
#ModelMetricClaimedVerifiedStatus
1GraPixPixel Accuracy64.89Unverified
2CUPSmIoU26.8Unverified
3ViCEmIoU25.2Unverified
4EAGLE (DINO, ViT-B/8)mIoU22.1Unverified
5EQUSSmIoU22Unverified
6PriMaPs-EM + STEGO (DINO ViT-B/8)mIoU21.6Unverified
7STEGOmIoU21Unverified
8EAGLE (DINO, ViT-S/8)mIoU19.7Unverified
9PriMaPs-EM (DINO ViT-S/8)mIoU19.4Unverified
10HPmIoU18.4Unverified
#ModelMetricClaimedVerifiedStatus
1CAUSE (iBOT, ViT-B/16)Clustering [mIoU]53.4Unverified
2CAUSE (ViT-B/8)Clustering [mIoU]53.3Unverified
3CAUSE (DINOv2, ViT-B/14)Clustering [mIoU]53.2Unverified
4MaskDistill+CRFClustering [mIoU]48.9Unverified
5Leopart (ViT-B/8)Clustering [mIoU]47.2Unverified
6HCL (ViT-S/8)Clustering [mIoU]46.3Unverified
7MaskDistillClustering [mIoU]45.8Unverified
8MaskContrast (Saliency)Clustering [mIoU]44.2Unverified
9HCL (ViT-S/16)Clustering [mIoU]43.2Unverified
10Leopart (ViT-S/16)Clustering [mIoU]41.7Unverified
#ModelMetricClaimedVerifiedStatus
1PriMaPs-EM+HP (DINO ViT-B/8)Accuracy83.3Unverified
2EAGLE (DINO, ViT-B/8)Accuracy83.3Unverified
3HPAccuracy82.4Unverified
4EQUSSAccuracy82Unverified
5PriMaPs-EM (DINO ViT-B/8)Accuracy80.5Unverified
6STEGOAccuracy77Unverified
7InfoSegPixel Accuracy71.6Unverified
8IICAccuracy45.4Unverified
#ModelMetricClaimedVerifiedStatus
1SANPixel Accuracy80.3Unverified
2SGSegPixel Accuracy74.6Unverified
3InfoSegPixel Accuracy73.8Unverified
4InMARSPixel Accuracy73.1Unverified
5ACPixel Accuracy72.9Unverified
6IICPixel Accuracy72.3Unverified
#ModelMetricClaimedVerifiedStatus
1PASS (+Saliency map)mIoU (test)42.3Unverified
2PASSmIoU (test)32Unverified
3MaskContrast (+Saliency map)mIoU (test)24.2Unverified
4PiCIE (Supervised pretrain)mIoU (test)17.6Unverified
5MDC (Supervised pretrain)mIoU (test)14.3Unverified
#ModelMetricClaimedVerifiedStatus
1CAUSE-TR (ViT-S/8)mIoU15.2Unverified
2TransFGU (ViT-S/8)mIoU11.93Unverified
3PiCIE (ResNet-50)mIoU5.6Unverified
4IIC (ResNet-50)mIoU2.2Unverified
#ModelMetricClaimedVerifiedStatus
1CAUSE-TR (ViT-S/8)mIoU21.2Unverified
2CAUSE-MLP (ViT-S/8)mIoU19.1Unverified
3TransFGU (ViT-S/8)mIoU12.7Unverified
4MaskContrast (ResNet-50)mIoU3.7Unverified
#ModelMetricClaimedVerifiedStatus
1DatUS (ViT-B/8) + OCPixel Accuracy69.98Unverified
2GraPix + AUTPixel Accuracy65.48Unverified
3DatUS (ViT-B/8)Pixel Accuracy64.67Unverified
4GraPixPixel Accuracy64.06Unverified
#ModelMetricClaimedVerifiedStatus
1InfoSegPixel Accuracy38.8Unverified
2InMARSPixel Accuracy31Unverified
3IICPixel Accuracy27.7Unverified
#ModelMetricClaimedVerifiedStatus
1Segmenter ViT-S/16mIoU16.7Unverified
#ModelMetricClaimedVerifiedStatus
1Segmenter ViT-S/16mIoU21.8Unverified
#ModelMetricClaimedVerifiedStatus
1DenseSiammIoU16.4Unverified
#ModelMetricClaimedVerifiedStatus
1InfoSegPixel Accuracy69.6Unverified
#ModelMetricClaimedVerifiedStatus
1Segmenter ViT-S/16mIoU14.2Unverified
#ModelMetricClaimedVerifiedStatus
1PASSmIoU (test)11Unverified
#ModelMetricClaimedVerifiedStatus
1PASSmIoU (test)18.1Unverified
#ModelMetricClaimedVerifiedStatus
1Segmenter ViT-S/16mIoU18.9Unverified