Object Detection on SUN RGB-D (test)
55.7mAPMCTNet
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| MCTNetInput modality=RGB-D, RGB backbone=ResNet-50, Depth backbone=ResNet-182023.09 | 55.7 | — | — | — | — | — | — | — | — | |
| FetNetInput modality=RGB-D, RGB backbone=ResNet-50, Depth backbone=ResNet-1522023.09 | 54.5 | — | — | — | — | — | — | — | — | |
| RBF Weighted Hyper-InvolutionInput modality=RGB-D, RGB backbone=YOLO, Depth backbone=Depth-aware Modulation2023.09 | 53.3 | — | — | — | — | — | — | — | — | |
| Faster R-CNNInput modality=RGB, RGB backbone=ResNet-50, Depth backbone=-2023.09 | 52.3 | — | — | — | — | — | — | — | — | |
| Dynamic R-CNNInput modality=RGB, RGB backbone=ResNet-50, Depth backbone=-2023.09 | 52.2 | — | — | — | — | — | — | — | — | |
| Cascade R-CNNInput modality=RGB, RGB backbone=ResNet-50, Depth backbone=-2023.09 | 51.5 | — | — | — | — | — | — | — | — | |
| SABLInput modality=RGB, RGB backbone=ResNet-50, Depth backbone=-2023.09 | 50.4 | — | — | — | — | — | — | — | — | |
| GFLInput modality=RGB, RGB backbone=ResNet-50, Depth backbone=-2023.09 | 47.5 | — | — | — | — | — | — | — | — | |
| CMACInput modality=RGB-D, RGB backbone=VGG-16, Depth backbone=AlexNet2023.09 | 47.5 | — | — | — | — | — | — | — | — | |
| ATSSInput modality=RGB, RGB backbone=ResNet-50, Depth backbone=-2023.09 | 47.3 | — | — | — | — | — | — | — | — | |
| AC-CNNInput modality=RGB-D, RGB backbone=VGG-16, Depth backbone=AlexNet2023.09 | 45.4 | — | — | — | — | — | — | — | — | |
| Super TransferInput modality=RGB-D, RGB backbone=VGG-16, Depth backbone=AlexNet2023.09 | 43.8 | — | — | — | — | — | — | — | — | |
| Sparse R-CNNInput modality=RGB, RGB backbone=ResNet-50, Depth backbone=-2023.09 | 43.4 | — | — | — | — | — | — | — | — | |
| RGB-D R-CNNInput modality=RGB-D, RGB backbone=AlexNet, Depth backbone=AlexNet2023.09 | 35.2 | — | — | — | — | — | — | — | — | |
| YOLOv11Input modality=RGB, RGB backbone=YOLOv11x, Depth backbone=-2023.09 | 32.9 | — | — | — | — | — | — | — | — | |
| YOLOv12Input modality=RGB, RGB backbone=YOLOv12x, Depth backbone=-2023.09 | 32.7 | — | — | — | — | — | — | — | — | |
| YOLOv8Input modality=RGB, RGB backbone=YOLOv8x, Depth backbone=-2023.09 | 32.4 | — | — | — | — | — | — | — | — | |
| DLWBackbone=EfficientDet, alpha=1.02026.02 | 19.1 | 33.6 | 1.5 | 9.8 | 27 | 35.3 | 3.5 | 24.3 | 44.6 | |
| DLSBackbone=EfficientDet, lambda_CLOSE=1, lambda_DISTANT=5, beta=25%2026.02 | 18.7 | 32.8 | 1.6 | 9.4 | 26.6 | 35.2 | 3.5 | 24.2 | 45.1 | |
| BWBackbone=EfficientDet2026.02 | 18.5 | 33.2 | 1.3 | 9.3 | 26.6 | 36.4 | 3.2 | 25 | 47.3 | |
| IWBackbone=EfficientDet2026.02 | 18.3 | 32.3 | 1.3 | 9.5 | 25.8 | 35.3 | 3 | 24.2 | 45.6 | |
| BaselineBackbone=EfficientDet2026.02 | 17.6 | 31.3 | 2 | 8.9 | 24.7 | 32 | 3.8 | 20.8 | 43.7 | |
| LBackbone=EfficientDet2026.02 | 12.7 | 23.2 | 0.5 | 5.5 | 18.6 | 27.7 | 1.4 | 16.5 | 37.2 | |
| CLBackbone=EfficientDet2026.02 | 10.2 | 19.7 | 0.5 | 4.3 | 14.9 | 24.4 | 0.9 | 13.3 | 33.6 | |
| PDBackbone=EfficientDet2026.02 | 0.5 | 1.9 | 0 | 0.1 | 0.7 | 2 | 0.2 | 0.8 | 2.9 |