Oriented Object Detection on DOTA 1.5 (30% labeled)
59.23mAPSOOD
Evaluation results
| Method | Links | |
|---|---|---|
| SOODSetting=Semi-supervised, Detector=rotated-FCOS2023.04 | 59.23 | |
| Dense TeacherSetting=Semi-supervised, Detector=rotated-FCOS2023.04 | 57.86 | |
| Soft TeacherSetting=Semi-supervised, Detector=rotated-Faster-RCNN2023.04 | 57.83 | |
| FCOSSetting=Supervised, Detector=rotated-FCOS2023.04 | 54.79 | |
| Unbiased TeacherSetting=Semi-supervised, Detector=rotated-Faster-RCNN2023.04 | 53.33 | |
| Faster R-CNNSetting=Supervised, Detector=rotated-Faster-RCNN2023.04 | 53.14 |