Object Detection on DOTA v1.5
68.92mAPGGHL
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| GGHLSpeed (fps)=41.072021.09 | 68.92 | — | — | — | — | — | |
| Faster R-CNN OBB + RTSpeed (fps)=12.402021.09 | 65.03 | — | — | — | — | — | |
| Cascade Mask R-CNNSpeed (fps)=7.202021.09 | 63.41 | — | — | — | — | — | |
| Hybrid Task MaskSpeed (fps)=7.902021.09 | 63.4 | — | — | — | — | — | |
| Mask R-CNNSpeed (fps)=9.702021.09 | 62.67 | — | — | — | — | — | |
| Faster R-CNN H-OBBSpeed (fps)=13.702021.09 | 62.57 | — | — | — | — | — | |
| Faster R-CNN OBB + DpoolSpeed (fps)=12.102021.09 | 62.2 | — | — | — | — | — | |
| Faster R-CNN OBBSpeed (fps)=14.102021.09 | 62 | — | — | — | — | — | |
| RetinaNet OBBSpeed (fps)=12.102021.09 | 59.16 | — | — | — | — | — | |
| SDCoNetBackbone=Swin-T, Type=Remote Sensing Object Detectors2026.01 | 41.9 | 67.4 | 45.5 | 25 | 45.1 | 56.2 | |
| RoI Trans. + COBB-ln-lnbackbone=RoI Trans., variant=ln-ln2024.02 | 40.96 | 66.66 | 43.29 | — | — | — | |
| RoI Trans. + COBB-sig-sigbackbone=RoI Trans., variant=sig-sig2024.02 | 40.85 | 65.88 | 42.76 | — | — | — | |
| RoI Trans. + COBB-ln-sigbackbone=RoI Trans., variant=ln-sig2024.02 | 40.8 | 67.18 | 41.75 | — | — | — | |
| RoI Trans.baseline=true2024.02 | 40.36 | 65.69 | 41.76 | — | — | — | |
| DINO (Baseline)Backbone=Swin-T, Type=General Object Detectors2026.01 | 40.2 | 65.9 | 42.7 | 19.2 | 43.4 | 54 | |
| Oriented R-CNN + COBB-sigbackbone=Oriented R-CNN, variant=sig2024.02 | 40.04 | 66.25 | 41.34 | — | — | — | |
| Oriented R-CNN + COBB-lnbackbone=Oriented R-CNN, variant=ln2024.02 | 40.01 | 66.18 | 41.42 | — | — | — | |
| FFCA-YOLOBackbone=CSPDarknet, Type=Remote Sensing Object Detectors2026.01 | 39.9 | 64.2 | 42.5 | 22.1 | 42.5 | 47 | |
| Oriented R-CNNbaseline=true2024.02 | 39.31 | 65.47 | 40.35 | — | — | — | |
| Rotated Faster R-CNN + COBB-lnbackbone=Rotated Faster R-CNN, variant=ln2024.02 | 37.3 | 64.35 | 37.62 | — | — | — | |
| Rotated Faster R-CNN + COBB-sigbackbone=Rotated Faster R-CNN, variant=sig2024.02 | 37.17 | 64.03 | 36.88 | — | — | — | |
| DINOBackbone=ResNet50, Type=General Object Detectors2026.01 | 36.7 | 59.7 | 39.7 | 16 | 39.4 | 50.6 | |
| Gliding Vertexbaseline=true2024.02 | 36.32 | 63.12 | 36.98 | — | — | — | |
| Rotated Faster R-CNNbaseline=true2024.02 | 35.96 | 63.52 | 35.36 | — | — | — | |
| SR4IRBackbone=MobileNet-V3, Type=General Object Detectors2026.01 | 32.9 | 51.1 | 34 | 14.7 | 35.8 | 49.5 | |
| LEGNet-T (Faster RCNN)Backbone=LEGNet-T, Type=Remote Sensing Object Detectors2026.01 | 31.6 | 50.8 | 34.3 | 8.4 | 36.1 | 43.4 | |
| EESRGANBackbone=ResNet50, Type=Remote Sensing Object Detectors2026.01 | 31.4 | 51.2 | 34.2 | 16 | 36 | 48.1 | |
| SuperYOLO (RGB)Backbone=CSPDarknet, Type=Remote Sensing Object Detectors2026.01 | 30.7 | 49.5 | 39.7 | 19.1 | 33.5 | 42.9 | |
| Faster RCNNBackbone=ResNet50, Type=General Object Detectors2026.01 | 30.3 | 48.7 | 32.7 | 9.3 | 33.8 | 43.4 | |
| FSANetBackbone=Swin-T, Type=Remote Sensing Object Detectors2026.01 | 28.3 | 55.7 | 27.1 | 10.3 | 31.2 | 38.4 | |
| DN-DETRBackbone=ResNet50, Type=General Object Detectors2026.01 | 27.4 | 46.2 | 25.4 | 3 | 28.9 | 42.5 | |
| YOLOX-sBackbone=CSPDarknet, Type=General Object Detectors2026.01 | 26.9 | 49.8 | 27.7 | 11.9 | 29.2 | 34.7 | |
| Sparse R-CNNBackbone=ResNet50, Type=General Object Detectors2026.01 | 26.6 | 44.6 | 27.9 | 6 | 26.3 | 42.2 | |
| Cascade R-CNNBackbone=ResNet50, Type=General Object Detectors2026.01 | 26.5 | 43.4 | 27.5 | 6.5 | 27.9 | 38.8 | |
| Deformable-DETRBackbone=Swin-T, Type=General Object Detectors2026.01 | 24.8 | 46.4 | 22.9 | 4.4 | 25.6 | 43.4 | |
| RetinaNetBackbone=ResNet50, Type=General Object Detectors2026.01 | 24.3 | 41.3 | 24.7 | 2.1 | 23.9 | 39 | |
| DAB-DETRBackbone=ResNet50, Type=General Object Detectors2026.01 | 23.3 | 44 | 21.5 | 3.5 | 21.1 | 39.4 |