Incremental Semantic Segmentation on Pascal VOC 10-5
79.1mIoU (1-10)Joint
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| JointSupervision (Sup)=Pixel-level (P), Backbone=ViT2026.06 | 79.1 | 77.3 | 78.2 | |
| JointSupervision (Sup)=Pixel-level (P), Backbone=ResNet2026.06 | 78.4 | 76.4 | 77.4 | |
| BARMSupervision (Sup)=Pixel-level (P), Backbone=ResNet2026.06 | 75.7 | 64.8 | 70.5 | |
| SASASupervision (Sup)=Image-level (I), Backbone=ViT2026.06 | 75.1 | 66.3 | 71.7 | |
| WILSONSupervision (Sup)=Image-level (I), Backbone=ViT2026.06 | 73.6 | 57.5 | 66.9 | |
| STARSupervision (Sup)=Pixel-level (P), Backbone=ResNet2026.06 | 73.5 | 64.7 | 69.3 | |
| ToCoSupervision (Sup)=Image-level (I), Backbone=ViT2026.06 | 72.9 | 56 | 65.7 | |
| TeddySupervision (Sup)=Image-level (I), Backbone=ResNet2026.06 | 68.9 | 51.7 | 61.7 | |
| DKDSupervision (Sup)=Pixel-level (P), Backbone=ResNet2026.06 | 68.8 | 57.6 | 63.4 | |
| ALIFESupervision (Sup)=Pixel-level (P), Backbone=ResNet2026.06 | 68.3 | 58.8 | 63.8 | |
| WILSONSupervision (Sup)=Image-level (I), Backbone=ResNet2026.06 | 66.8 | 46.5 | 58.1 |