Object Detection on DOTA v1.5 (test)
73.57mAPPKINet-v2-S
Evaluation results
| Method | Links | |||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| PKINet-v2-SScale=single-scale, Pretraining=ImageNet-1K (300 epochs), Framework=Oriented RCNN2026.03 | 73.57 | 80.4 | 85.15 | 56.33 | 75.14 | 52.7 | 82.69 | 89.12 | 90.89 | 79.53 | 68.96 | 65.73 | 73.72 | 77.76 | 76 | 72.79 | 50.15 | |
| LEGNet-SStrategy=single-scale, Base detector=O-RCNN2025.03 | 72.89 | 80.48 | 85.04 | 55.64 | 74.86 | 52.64 | 82.16 | 89.11 | 90.88 | 84.55 | 68.91 | 66.21 | 74.16 | 77.44 | 74.68 | 65.76 | 4,375 | |
| LSKNet+oursBackbone=LSKNet-S, Scale configuration=single scale training and testing, Fusion=FAAFusion, Head=FAA Head2026.02 | 72.28 | 80.69 | 85.34 | 55.42 | 77.91 | 52.64 | 82.26 | 88.48 | 90.85 | 86.38 | 69.32 | 62.04 | 74.11 | 75.83 | 74.7 | 73.18 | 27.3 | |
| Strip R-CNN-SPre-trained=ImageNet, Scale=Single-Scale2025.01 | 72.27 | 80.04 | 83.26 | 54.4 | 75.38 | 52.46 | 81.44 | 88.53 | 90.83 | 84.8 | 69.65 | 65.93 | 73.28 | 69.7 | 74.04 | — | 3,898 | |
| Strip R-CNN-SPre.=IN, Scale=Single-Scale2025.01 | 72.27 | 80.04 | 83.26 | 54.4 | 75.38 | 52.46 | 81.44 | 88.53 | 90.83 | 84.8 | 69.65 | 65.93 | 73.28 | 74.61 | 74.04 | 69.7 | 3,898 | |
| Strip R-CNN-Sscale=single-scale, backbone=StripNet-S, pre-training=ImageNet (300 epochs)2025.01 | 72.27 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Strip RCNN-SScale=single-scale, Pretraining=ImageNet-1K (300 epochs)2026.03 | 72.27 | 80.04 | 83.26 | 54.4 | 75.38 | 52.46 | 81.44 | 88.53 | 90.83 | 84.8 | 69.65 | 65.93 | 73.28 | 74.61 | 74.04 | 69.7 | 38.98 | |
| S-RCNN+oursBackbone=StripNet-S, Scale configuration=single scale training and testing, Fusion=FAAFusion, Head=FAA Head2026.02 | 71.57 | 80.14 | 82.02 | 51.89 | 76.42 | 52.52 | 81.45 | 88.21 | 90.88 | 85.76 | 68.52 | 62.34 | 72.68 | 74.01 | 74.97 | 71.2 | 32.1 | |
| PKINet-SPre-trained=ImageNet, Scale=Single-Scale2025.01 | 71.47 | 80.31 | 85 | 55.61 | 74.38 | 52.41 | 76.85 | 88.38 | 90.87 | 79.04 | 68.78 | 67.47 | 72.45 | 64.07 | 74.53 | — | 3,713 | |
| PKINet-SPre.=IN, Scale=Single-Scale2025.01 | 71.47 | 80.31 | 85 | 55.61 | 74.38 | 52.41 | 76.85 | 88.38 | 90.87 | 79.04 | 68.78 | 67.47 | 72.45 | 76.24 | 74.53 | 64.07 | 3,713 | |
| PKINet-Sscale=single-scale2025.01 | 71.47 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PKINet-SStrategy=single-scale2025.03 | 71.47 | 80.31 | 85 | 55.61 | 74.38 | 52.41 | 76.85 | 88.38 | 90.87 | 79.04 | 68.78 | 67.47 | 72.45 | 76.24 | 74.53 | 64.07 | 3,713 | |
| PKINetBackbone=PKINet-S, Scale configuration=single scale training and testing2026.02 | 71.47 | 80.31 | 85 | 55.61 | 74.38 | 52.41 | 76.85 | 88.38 | 90.87 | 79.04 | 68.78 | 67.47 | 72.45 | 76.24 | 74.53 | 64.07 | 37.13 | |
| PKINet-v1-SScale=single-scale, Pretraining=ImageNet-1K (300 epochs)2026.03 | 71.47 | 80.31 | 85 | 55.61 | 74.38 | 52.41 | 76.85 | 88.38 | 90.87 | 79.04 | 68.78 | 67.47 | 72.45 | 76.24 | 74.53 | 64.07 | 37.13 | |
| SOODStrategy=single-scale2025.03 | 70.39 | 80.32 | 84.41 | 52.59 | 74.77 | 58.48 | 76.9 | 86.97 | 90.87 | 78.62 | 76.56 | 62.93 | 71.16 | 74.64 | 76.04 | 55.97 | 2,509 | |
| LSKNet-SPre-trained=ImageNet, Scale=Single-Scale2025.01 | 70.26 | 72.05 | 84.94 | 55.41 | 74.93 | 52.42 | 77.45 | 81.17 | 90.85 | 79.44 | 69 | 62.1 | 73.72 | 55.81 | 75.29 | — | 4,219 | |
| LSKNet-SPre.=IN, Scale=Single-Scale2025.01 | 70.26 | 72.05 | 84.94 | 55.41 | 74.93 | 52.42 | 77.45 | 81.17 | 90.85 | 79.44 | 69 | 62.1 | 73.72 | 77.49 | 75.29 | 55.81 | 4,219 | |
| LSKNet-Sscale=single-scale2025.01 | 70.26 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| LSKNet-SStrategy=single-scale2025.03 | 70.26 | 72.05 | 84.94 | 55.41 | 74.93 | 52.42 | 77.45 | 81.17 | 90.85 | 79.44 | 69 | 62.1 | 73.72 | 77.49 | 75.29 | 55.81 | 4,219 | |
| LSKNetBackbone=LSKNet-S, Scale configuration=single scale training and testing2026.02 | 70.26 | 72.05 | 84.94 | 55.41 | 74.93 | 52.42 | 77.45 | 81.17 | 90.85 | 79.44 | 69 | 62.1 | 73.72 | 77.49 | 75.29 | 55.81 | 42.19 | |
| LSKNet-SScale=single-scale2026.03 | 70.26 | 72.05 | 84.94 | 55.41 | 74.93 | 52.42 | 77.45 | 81.17 | 90.85 | 79.44 | 69 | 62.1 | 73.72 | 77.49 | 75.29 | 55.81 | 42.19 | |
| S-RCNNBackbone=StripNet-S, Scale configuration=single scale training and testing2026.02 | 69.84 | 80.31 | 82.24 | 53.91 | 76.61 | 52.39 | 81.55 | 87.82 | 90.89 | 82.44 | 66.32 | 61.58 | 73.27 | 75.04 | 65.84 | 73.42 | 13.85 | |
| DCFLscale=single-scale2025.01 | 67.37 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DCFLScale=single-scale, Pretraining=ImageNet-1K (300 epochs)2026.03 | 67.37 | — | — | — | — | 56.72 | — | 80.87 | — | 75.65 | — | — | — | — | — | — | — | |
| O-RCNN+oursBackbone=ResNet50, Scale configuration=single scale training and testing, Fusion=FAAFusion, Head=FAA Head2026.02 | 67.14 | 79.62 | 81.14 | 53.96 | 73.78 | 52.2 | 76.8 | 81.17 | 90.9 | 78.47 | 68.55 | 64.21 | 73.72 | 68.26 | 64.64 | 54.09 | 12.85 | |
| ReDetPre-trained=ImageNet, Scale=Single-Scale2025.01 | 66.86 | 79.2 | 82.81 | 51.92 | 71.41 | 52.38 | 75.73 | 80.92 | 90.83 | 75.81 | 68.64 | 49.29 | 72.03 | 63.33 | 70.55 | — | 1,153 | |
| ReDetPre.=IN, Scale=Single-Scale2025.01 | 66.86 | 79.2 | 82.81 | 51.92 | 71.41 | 52.38 | 75.73 | 80.92 | 90.83 | 75.81 | 68.64 | 49.29 | 72.03 | 73.36 | 70.55 | 63.33 | 1,153 | |
| ReDetscale=single-scale2025.01 | 66.86 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ReDetStrategy=single-scale2025.03 | 66.86 | 79.2 | 82.81 | 51.92 | 71.41 | 52.38 | 75.73 | 80.92 | 90.83 | 75.81 | 68.64 | 49.29 | 72.03 | 73.36 | 70.55 | 63.33 | 1,153 | |
| ReDetBackbone=ReResNet, Scale configuration=single scale training and testing2026.02 | 66.86 | 79.2 | 82.81 | 51.92 | 71.41 | 52.38 | 75.73 | 80.92 | 90.83 | 75.81 | 68.64 | 49.29 | 72.03 | 73.36 | 70.55 | 63.33 | 11.53 | |
| ReDetScale=single-scale, Pretraining=ImageNet-1K (300 epochs)2026.03 | 66.86 | 79.2 | 82.81 | 51.92 | 71.41 | 54.08 | 75.73 | 80.92 | 90.83 | 75.81 | 68.64 | 49.29 | 72.03 | 73.36 | 70.55 | 63.33 | 11.53 | |
| O-RCNNBackbone=ResNet50, Scale configuration=single scale training and testing2026.02 | 66.77 | 79.68 | 81.72 | 53.36 | 72.17 | 52.26 | 76.47 | 81 | 90.86 | 78.93 | 68.18 | 65.05 | 72.17 | 67.68 | 65.23 | 56.11 | 7.42 | |
| RoI Trans.Backbone=ResNet50, Scale configuration=single scale training and testing2026.02 | 65.5 | 71.7 | 82.7 | 53 | 71.5 | 51.3 | 74.6 | 80.6 | 90.4 | 78 | 68.3 | 53.1 | 73.4 | 73.9 | 65.6 | 56.9 | 3 | |
| HTCPre-trained=ImageNet, Scale=Single-Scale2025.01 | 63.4 | 77.8 | 73.67 | 51.4 | 63.99 | 51.54 | 73.31 | 80.31 | 90.48 | 75.12 | 67.34 | 48.51 | 70.63 | 55.87 | 64.48 | — | 515 | |
| HTCPre.=IN, Scale=Single-Scale2025.01 | 63.4 | 77.8 | 73.67 | 51.4 | 63.99 | 51.54 | 73.31 | 80.31 | 90.48 | 75.12 | 67.34 | 48.51 | 70.63 | 64.84 | 64.48 | 55.87 | 515 | |
| HTCscale=single-scale2025.01 | 63.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| HTCStrategy=single-scale2025.03 | 63.4 | 77.8 | 73.67 | 51.4 | 63.99 | 51.54 | 73.31 | 80.31 | 90.48 | 75.12 | 67.34 | 48.51 | 70.63 | 64.84 | 64.48 | 55.87 | 515 | |
| HTCScale configuration=single scale training and testing2026.02 | 63.4 | 77.8 | 73.67 | 51.4 | 63.99 | 51.54 | 73.31 | 80.31 | 90.48 | 75.12 | 67.34 | 48.51 | 70.63 | 64.84 | 64.48 | 55.87 | 5.15 | |
| HTCScale=single-scale2026.03 | 63.4 | 77.8 | 73.67 | 51.4 | 63.99 | 51.54 | 73.31 | 80.31 | 90.48 | 75.12 | 67.34 | 48.51 | 70.63 | 64.84 | 64.48 | 55.87 | 5.15 | |
| Mask R-CNNPre-trained=ImageNet, Scale=Single-Scale2025.01 | 62.67 | 76.84 | 73.51 | 49.9 | 57.8 | 51.31 | 71.34 | 79.75 | 90.46 | 74.21 | 66.07 | 46.21 | 70.61 | 57.81 | 64.46 | — | 942 | |
| Mask R-CNNPre.=IN, Scale=Single-Scale2025.01 | 62.67 | 76.84 | 73.51 | 49.9 | 57.8 | 51.31 | 71.34 | 79.75 | 90.46 | 74.21 | 66.07 | 46.21 | 70.61 | 63.07 | 64.46 | 57.81 | 942 | |
| Mask RCNNscale=single-scale2025.01 | 62.67 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Mask R-CNNStrategy=single-scale2025.03 | 62.67 | 76.84 | 73.51 | 49.9 | 57.8 | 51.31 | 71.34 | 79.75 | 90.46 | 74.21 | 66.07 | 46.21 | 70.61 | 63.07 | 64.46 | 57.81 | 942 | |
| Mask R-CNNBackbone=ResNet50, Scale configuration=single scale training and testing2026.02 | 62.67 | 76.84 | 73.51 | 49.9 | 57.8 | 51.31 | 71.34 | 79.75 | 90.46 | 74.21 | 66.07 | 46.21 | 70.61 | 63.07 | 64.46 | 57.81 | 9.42 | |
| Mask RCNNScale=single-scale2026.03 | 62.67 | 76.84 | 73.51 | 49.9 | 57.8 | 51.31 | 71.34 | 79.75 | 90.46 | 74.21 | 66.07 | 46.21 | 70.61 | 63.07 | 64.46 | 57.81 | 9.42 | |
| FR-OPre-trained=ImageNet, Scale=Single-Scale2025.01 | 62 | 71.89 | 74.47 | 44.45 | 59.87 | 51.28 | 68.98 | 79.37 | 90.78 | 77.38 | 67.5 | 47.75 | 69.72 | 60.47 | 65.28 | — | 154 | |
| FR-OPre.=IN, Scale=Single-Scale2025.01 | 62 | 71.89 | 74.47 | 44.45 | 59.87 | 51.28 | 68.98 | 79.37 | 90.78 | 77.38 | 67.5 | 47.75 | 69.72 | 61.22 | 65.28 | 60.47 | 154 | |
| FR-Oscale=single-scale2025.01 | 62 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FR-OStrategy=single-scale2025.03 | 62 | 71.89 | 74.47 | 44.45 | 59.87 | 51.28 | 68.98 | 79.37 | 90.78 | 77.38 | 67.5 | 47.75 | 69.72 | 61.22 | 65.28 | 60.47 | 154 | |
| R. F-RCNNBackbone=ResNet50, Scale configuration=single scale training and testing2026.02 | 62 | 71.89 | 74.47 | 44.45 | 59.87 | 51.28 | 68.98 | 79.37 | 90.78 | 77.38 | 67.5 | 47.75 | 69.72 | 61.22 | 65.28 | 60.47 | 1.54 | |
| FR-OScale=single-scale2026.03 | 62 | 71.89 | 74.47 | 44.45 | 59.87 | 51.28 | 68.98 | 79.37 | 90.78 | 77.38 | 67.5 | 47.75 | 69.72 | 61.22 | 65.28 | 60.47 | 1.54 | |
| RetinaNet-OPre-trained=ImageNet, Scale=Single-Scale2025.01 | 59.16 | 71.43 | 77.64 | 42.12 | 64.65 | 44.53 | 56.79 | 73.31 | 90.84 | 76.02 | 59.96 | 46.95 | 69.24 | 48.06 | 64.52 | — | 83 | |
| RetinaNet-OPre.=IN, Scale=Single-Scale2025.01 | 59.16 | 71.43 | 77.64 | 42.12 | 64.65 | 44.53 | 56.79 | 73.31 | 90.84 | 76.02 | 59.96 | 46.95 | 69.24 | 59.65 | 64.52 | 48.06 | 83 | |
| RetinaNet-Oscale=single-scale2025.01 | 59.16 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RetinaNet-OStrategy=single-scale2025.03 | 59.16 | 71.43 | 77.64 | 42.12 | 64.65 | 44.53 | 56.79 | 73.31 | 90.84 | 76.02 | 59.96 | 46.95 | 69.24 | 59.65 | 64.52 | 48.06 | 83 | |
| RetinaNet-OBackbone=ResNet50, Scale configuration=single scale training and testing2026.02 | 59.16 | 71.43 | 77.64 | 42.12 | 64.65 | 44.53 | 56.79 | 73.31 | 90.84 | 76.02 | 59.96 | 46.95 | 69.24 | 59.65 | 64.52 | 48.06 | 0.83 | |
| RetinaNet-OScale=single-scale2026.03 | 59.16 | 71.43 | 77.64 | 42.12 | 64.65 | 44.53 | 56.79 | 73.31 | 90.84 | 76.02 | 59.96 | 46.95 | 69.24 | 59.65 | 64.52 | 48.06 | 0.83 |