Outlier-aware object detection on Youtube-VIS (ID) MS-COCO (OD) 2021 (test)
81.14FPR@95STUD
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| STUDBackbone=RegNetX-4.0GF, Detector=Faster R-CNN2023.02 | 81.14 | 74.82 | 27.2 | |
| CSIBackbone=RegNetX-4.0GF, Detector=Faster R-CNN2023.02 | 82.43 | 71.81 | 24.2 | |
| FFSBackbone=RegNetX-4.0GF, Detector=Faster R-CNN2023.02 | 83.06 | 76.37 | 27.6 | |
| GANBackbone=RegNetX-4.0GF, Detector=Faster R-CNN2023.02 | 85.75 | 72.95 | 25.5 | |
| EnergyBackbone=RegNetX-4.0GF, Detector=Faster R-CNN2023.02 | 88.54 | 67.83 | 26.7 |