| Benchmark | Model | Rank | Results |
|---|---|---|---|
| classification-on-indl | Darknet53 | #7 | Average Recall: 88.53% |
| object-detection-on-coco-o | YOLOv3 (DarkNet-53) | #41 | Average mAP: 14.8Effective Robustness: -0.37 |
| pedestrian-detection-on-dvtod | YOLOv3 (Thermal) | #2 | mAP: 82.7 |
| pedestrian-detection-on-dvtod | YOLOv3 (Visible) | #6 | mAP: 34.5 |
| real-time-object-detection-on-coco | YOLOv3-L | #74 | box AP: 33.0 |
| robust-object-detection-on-cityscapes-1 | Photometric distortion | #10 | mPC [AP]: 16.9 |