Oriented Object Detection on DOTA 1.5 (10% labeled)
48.63mAPSOOD
Evaluation results
| Method | Links | |
|---|---|---|
| SOODSetting=Semi-supervised, Detector=rotated-FCOS2023.04 | 48.63 | |
| Soft TeacherSetting=Semi-supervised, Detector=rotated-Faster-RCNN2023.04 | 48.46 | |
| Dense TeacherSetting=Semi-supervised, Detector=rotated-FCOS2023.04 | 46.9 | |
| Unbiased TeacherSetting=Semi-supervised, Detector=rotated-Faster-RCNN2023.04 | 44.51 | |
| Faster R-CNNSetting=Supervised, Detector=rotated-Faster-RCNN2023.04 | 43.43 | |
| FCOSSetting=Supervised, Detector=rotated-FCOS2023.04 | 42.78 |