Semantic Segmentation on Potsdam 3
82.4Pixel AccuracyHP
Evaluation results
| Method | Links | |
|---|---|---|
| HPBackbone=ViT-B/8, Pretrained=DINO, Evaluation Protocol=Unsupervised2023.03 | 82.4 | |
| EQUSSEvaluation Protocol=Unsupervised, Backbone=ViT-B2023.12 | 82 | |
| STEGOBackbone=ViT-B/8, Pretrained=DINO, Evaluation Protocol=Unsupervised2023.03 | 77 | |
| STEGOEvaluation Protocol=Unsupervised, Backbone=ViT-B2023.12 | 77 | |
| InfoSeg2021.10 | 71.6 | |
| InMARS2021.10 | 70.1 | |
| AC2021.10 | 66.5 | |
| IIC2021.10 | 65.1 | |
| IICBackbone=VGG11, Evaluation Protocol=Unsupervised2023.03 | 65.1 | |
| IICEvaluation Protocol=Unsupervised2023.12 | 65.1 | |
| IsolaMethodology=Clustering of features from non-segmentation method2021.10 | 63.9 | |
| CCBackbone=VGG11, Evaluation Protocol=Unsupervised2023.03 | 63.9 | |
| Isola et al.Evaluation Protocol=Unsupervised2023.12 | 63.9 | |
| DINOBackbone=ViT-B/8, Evaluation Protocol=Unsupervised2023.03 | 53 | |
| DoerschMethodology=Clustering of features from non-segmentation method2021.10 | 49.6 | |
| ContextPredictionBackbone=VGG11, Evaluation Protocol=Unsupervised2023.03 | 49.6 | |
| Doersch et al.Evaluation Protocol=Unsupervised2023.12 | 49.6 | |
| K-Means2021.10 | 45.7 | |
| K-MeansBackbone=VGG11, Evaluation Protocol=Unsupervised2023.03 | 45.7 | |
| DeepClusterBackbone=VGG11, Evaluation Protocol=Unsupervised2023.03 | 41.7 | |
| Deep ClusterEvaluation Protocol=Unsupervised2023.12 | 41.7 | |
| Random CNN2021.10 | 38.2 | |
| Random CNNBackbone=VGG11, Evaluation Protocol=Unsupervised2023.03 | 38.2 | |
| SIFTBackbone=VGG11, Evaluation Protocol=Unsupervised2023.03 | 38.2 | |
| SIFTEvaluation Protocol=Unsupervised2023.12 | 38.2 |