Video Instance Segmentation on OVIS (test)
39.4APMinVIS
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| MinVISBackbone=Swin-L, Training=Full2022.08 | 39.4 | 61.5 | 41.3 | 18.1 | 43.3 | |
| MinVISBackbone=Swin-L, Training=10%2022.08 | 37.2 | 60.7 | 38 | 17.3 | 41.1 | |
| MinVISBackbone=Swin-L, Training=5%2022.08 | 35.7 | 60.1 | 35.8 | 17.3 | 39.9 | |
| MinVISBackbone=Swin-L, Training=1%2022.08 | 31.7 | 54.9 | 31.3 | 16.3 | 36.1 | |
| MaskTrack R-CNN*+SWABackbone=Swin-L, Training=Full2022.08 | 28.9 | 56.3 | 26.8 | 13.5 | 34 | |
| Mask2Former-VIS*Backbone=Swin-L, Training=Full2022.08 | 25.8 | 46.5 | 24.4 | 13.7 | 32.2 | |
| MinVISBackbone=R50, Training=Full2022.08 | 25 | 45.5 | 24 | 13.9 | 29.7 | |
| TeViTBackbone=MsgShifT, Training=Full2022.08 | 17.4 | 34.9 | 15 | 11.2 | 21.8 | |
| Mask2Former-VIS*Backbone=R50, Training=Full2022.08 | 17.3 | 37.3 | 15.1 | 10.5 | 23.5 | |
| STCBackbone=R50, Training=Full2022.08 | 15.5 | 33.5 | 13.4 | 11 | 20.8 | |
| CMaskTrack R-CNNBackbone=R50, Training=Full2022.08 | 15.4 | 33.9 | 13.1 | 9.3 | 20 | |
| CrossVISBackbone=R50, Training=Full2022.08 | 14.9 | 32.7 | 12.1 | 10.3 | 19.8 |