Unsupervised Segmentation on COCO-Things (val)
19.2mIoUDenseCL
Evaluation results
| Method | Links | |
|---|---|---|
| DenseCLBackbone=ResNet50, Pre-training dataset=IN2023.03 | 19.2 | |
| CrOCBackbone=ViT-S/16, Pre-training dataset=CC2023.03 | 17.2 | |
| CrOCBackbone=ViT-S/16, Pre-training dataset=CC+2023.03 | 17.1 | |
| SoCoBackbone=ResNet50, Pre-training dataset=IN2023.03 | 16.3 | |
| ReSimBackbone=ResNet50, Pre-training dataset=IN2023.03 | 15.9 | |
| PixProBackbone=ResNet50, Pre-training dataset=IN2023.03 | 15.2 | |
| CP2Backbone=ViT-S/16, Pre-training dataset=IN2023.03 | 12.9 | |
| ORLBackbone=ResNet50, Pre-training dataset=CC+2023.03 | 12 | |
| VICRegLBackbone=ResNet50, Pre-training dataset=IN2023.03 | 11.2 | |
| BYOLBackbone=ResNet50, Pre-training dataset=CC+2023.03 | 9.4 | |
| DINOBackbone=ViT-S/16, Pre-training dataset=CC2023.03 | 9.4 | |
| MAEBackbone=ViT-S/16, Pre-training dataset=CC2023.03 | 7.5 | |
| CrOCBackbone=ViT-S/16, Pre-training dataset=IN2023.03 | 5.4 |