Semantic Segmentation on Cityscapes 27 classes (val)
61.47mIoUJAFAR
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| JAFARBackbone=DINOv2-ViT-S/14, Method Category=Task-Agnostic, Resolution=448x448, Evaluation Protocol=Linear Probing2025.06 | 61.47 | 93.42 | |
| DySampleBackbone=DINOv2-ViT-S/14, Method Category=Task-Dependent, Resolution=448x448, Evaluation Protocol=Linear Probing2025.06 | 59.71 | 92.69 | |
| BilinearBackbone=DINOv2-ViT-S/14, Method Category=Training-free, Resolution=448x448, Evaluation Protocol=Linear Probing2025.06 | 59.37 | 92.47 | |
| StridedBackbone=DINOv2-ViT-S/14, Method Category=Training-free, Resolution=448x448, Evaluation Protocol=Linear Probing2025.06 | 59.26 | 92.57 | |
| LIFTBackbone=DINOv2-ViT-S/14, Method Category=Task-Agnostic, Resolution=448x448, Evaluation Protocol=Linear Probing2025.06 | 58.75 | 92.6 | |
| FeatUpBackbone=DINOv2-ViT-S/14, Method Category=Task-Agnostic, Resolution=448x448, Evaluation Protocol=Linear Probing2025.06 | 56.06 | 91.86 | |
| CARAFEBackbone=DINOv2-ViT-S/14, Method Category=Task-Dependent, Resolution=448x448, Evaluation Protocol=Linear Probing2025.06 | 56.05 | 91.83 | |
| ReSFuBackbone=DINOv2-ViT-S/14, Method Category=Task-Dependent, Resolution=448x448, Evaluation Protocol=Linear Probing2025.06 | 55.53 | 91.62 | |
| NearestBackbone=DINOv2-ViT-S/14, Method Category=Training-free, Resolution=448x448, Evaluation Protocol=Linear Probing2025.06 | 54.05 | 90.36 | |
| SAPABackbone=DINOv2-ViT-S/14, Method Category=Task-Dependent, Resolution=448x448, Evaluation Protocol=Linear Probing2025.06 | 50.12 | 90.02 | |
| Large Image (x8)Backbone=DINOv2-ViT-S/14, Method Category=Training-free, Resolution=Large Image (x8), Evaluation Protocol=Linear Probing2025.06 | 47.72 | 92.49 |