Object Detection on VOC Dark low-light (val)
71.7mAPLR-YOLOv8
Evaluation results
| Method | Links | ||
|---|---|---|---|
| LR-YOLOv8Detector=YOLOv82025.10 | 71.7 | 63.9 | |
| LR-YOLOv10Detector=YOLOv102025.10 | 70.6 | 62.7 | |
| GDIP→YOLOv8Detector=YOLOv8, Restoration Method=GDIP2025.10 | 68.9 | 61.2 | |
| FeatEnHancer→YOLOv8Detector=YOLOv8, Restoration Method=FeatEnHancer2025.10 | 68.7 | 60.8 | |
| Retinexformer→YOLOv8Detector=YOLOv8, Restoration Method=Retinexformer, Training Strategy=Alternating training2025.10 | 68.6 | 61 | |
| Retinexformer→YOLOv8Detector=YOLOv8, Restoration Method=Retinexformer, Training Strategy=Standard2025.10 | 67.8 | 59.5 | |
| Retinexformer→YOLOv8Detector=YOLOv8, Restoration Method=Retinexformer, Training Strategy=Adversarial training2025.10 | 67.7 | 60 | |
| FeatEnHancer→YOLOv10Detector=YOLOv10, Restoration Method=FeatEnHancer2025.10 | 67.6 | 59.9 | |
| Retinexformer→YOLOv10Detector=YOLOv10, Restoration Method=Retinexformer, Training Strategy=Alternating training2025.10 | 66.9 | 59.2 | |
| IA→YOLOv8Detector=YOLOv8, Restoration Method=IA2025.10 | 66.5 | 59.2 | |
| LLFormer→YOLOv10Detector=YOLOv10, Restoration Method=LLFormer, Training Strategy=Alternating training2025.10 | 66.3 | 59.2 | |
| Retinexformer→YOLOv10Detector=YOLOv10, Restoration Method=Retinexformer, Training Strategy=Standard2025.10 | 66.3 | 58.6 | |
| LLFormer→YOLOv8Detector=YOLOv8, Restoration Method=LLFormer, Training Strategy=Standard2025.10 | 66.2 | 58.7 | |
| LLFormer→YOLOv8Detector=YOLOv8, Restoration Method=LLFormer, Training Strategy=Adversarial training2025.10 | 66.2 | 58.8 | |
| LLFormer→YOLOv8Detector=YOLOv8, Restoration Method=LLFormer, Training Strategy=Alternating training2025.10 | 66.2 | 59.2 | |
| Retinexformer→YOLOv10Detector=YOLOv10, Restoration Method=Retinexformer, Training Strategy=Adversarial training2025.10 | 66 | 58.4 | |
| IA→YOLOv10Detector=YOLOv10, Restoration Method=IA2025.10 | 66 | 58.7 | |
| GDIP→YOLOv10Detector=YOLOv10, Restoration Method=GDIP2025.10 | 65.8 | 58.5 | |
| LLFormer→YOLOv10Detector=YOLOv10, Restoration Method=LLFormer, Training Strategy=Standard2025.10 | 65.6 | 58 | |
| LLFormer→YOLOv10Detector=YOLOv10, Restoration Method=LLFormer, Training Strategy=Adversarial training2025.10 | 64.7 | 57.5 | |
| YOLOv8Detector=YOLOv8, Restoration Method=None2025.10 | 63.4 | 55.8 | |
| YOLOv10Detector=YOLOv10, Restoration Method=None2025.10 | 62.1 | 55 |