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 1–10 of 95 papers
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
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | PriMaPs-EM+HP (DINO ViT-B/8) | Accuracy | 83.3 | — | Unverified |
| 2 | EAGLE (DINO, ViT-B/8) | Accuracy | 83.3 | — | Unverified |
| 3 | HP | Accuracy | 82.4 | — | Unverified |
| 4 | EQUSS | Accuracy | 82 | — | Unverified |
| 5 | PriMaPs-EM (DINO ViT-B/8) | Accuracy | 80.5 | — | Unverified |
| 6 | STEGO | Accuracy | 77 | — | Unverified |
| 7 | InfoSeg | Pixel Accuracy | 71.6 | — | Unverified |
| 8 | IIC | Accuracy | 45.4 | — | Unverified |