Object Detection on HRSC 2016 (test)
90.7mAP@0.07PKINet-S
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| PKINet-S#Params=30.8M2024.03 | 90.7 | — | 98.54 | — | |
| LSKNet-SPre-trained=ImageNet, Number of Parameters (#P)=31.0M, FLOPs=161G2023.03 | 90.65 | — | 98.46 | — | |
| LSKNet#Params=31.0M2024.03 | 90.65 | — | 98.46 | — | |
| LSKNet-SPre-training=IN, Number of Parameters=31.0M, FLOPs=161G2025.01 | 90.65 | — | 98.46 | — | |
| PKINet-SPre-training=IN, Number of Parameters=30.8M, FLOPs=190G2025.01 | 90.65 | — | 98.54 | — | |
| RTMDetPre-trained=COCO, Number of Parameters (#P)=52.3M, FLOPs=205G2023.03 | 90.6 | — | 97.1 | — | |
| RTMDetPre-training=CO, Number of Parameters=52.3M, FLOPs=205G2025.01 | 90.6 | — | 97.1 | — | |
| Strip R-CNN-SPre-training=IN, Number of Parameters=30.5M, FLOPs=159G2025.01 | 90.6 | — | 98.7 | — | |
| O-RCNNPre-trained=ImageNet, Number of Parameters (#P)=41.1M, FLOPs=199G2023.03 | 90.5 | — | 97.6 | — | |
| O-RCNN#Params=41.1M2024.03 | 90.5 | — | 97.6 | — | |
| O-RCNNPre-training=IN, Number of Parameters=41.1M, FLOPs=199G2025.01 | 90.5 | — | 97.6 | — | |
| ReDetPre-trained=ImageNet, Number of Parameters (#P)=31.6M2023.03 | 90.46 | — | 97.63 | — | |
| ReDet#Params=31.6M2024.03 | 90.46 | — | 97.63 | — | |
| O-RepPoints#Params=36.6M2024.03 | 90.38 | — | 97.26 | — | |
| AOPGPre-trained=ImageNet2023.03 | 90.34 | — | 96.22 | — | |
| AOPG2024.03 | 90.34 | — | 96.22 | — | |
| AOPGPre-training=IN2025.01 | 90.34 | — | 96.22 | — | |
| S2A-NETBackbone=R-1012021.01 | 90.17 | — | 95.01 | — | |
| S2ANetPre-trained=ImageNet, Number of Parameters (#P)=38.6M, FLOPs=198G2023.03 | 90.17 | — | 95.01 | — | |
| S2ANet#Params=38.6M2024.03 | 90.17 | — | 95.01 | — | |
| R³DET-GWDBackbone=R-1012021.01 | 89.85 | — | 97.37 | — | |
| GWDPre-trained=ImageNet, Number of Parameters (#P)=47.4M, FLOPs=456G2023.03 | 89.85 | — | 97.37 | — | |
| GWD#Params=47.4M2024.03 | 89.85 | — | 97.37 | — | |
| GWDPre-training=IN, Number of Parameters=47.4M, FLOPs=456G2025.01 | 89.85 | — | 97.37 | — | |
| DALBackbone=R-1012021.01 | 89.77 | — | — | — | |
| DALPre-trained=ImageNet, Number of Parameters (#P)=36.4M, FLOPs=216G2023.03 | 89.77 | — | — | — | |
| DALPre-training=IN, Number of Parameters=36.4M, FLOPs=216G2025.01 | 89.77 | — | — | — | |
| CSLBackbone=ResNet1012020.03 | 89.62 | — | 96.1 | — | |
| FPN-CSLBackbone=R-1012021.01 | 89.62 | — | 96.1 | — | |
| GCLBackbone=ResNet1012020.03 | 89.56 | — | 96.02 | — | |
| R³Det-DCLBackbone=ResNet1012020.11 | 89.46 | — | 96.41 | — | |
| BCLBackbone=ResNet1012020.03 | 89.46 | — | 96.41 | — | |
| R³DET-DCLBackbone=R-1012021.01 | 89.46 | — | 96.41 | — | |
| R³DetBackbone=ResNet1012020.11 | 89.26 | — | 96.01 | — | |
| R3DetBackbone=ResNet1012020.03 | 89.26 | — | 96.01 | — | |
| R³DETBackbone=R-1012021.01 | 89.26 | — | 96.01 | — | |
| R3DetPre-trained=ImageNet, Number of Parameters (#P)=41.9M, FLOPs=336G2023.03 | 89.26 | — | 96.01 | — | |
| R3Det#Params=41.9M2024.03 | 89.26 | — | 96.01 | — | |
| RetinaNet-RBackbone=ResNet1012020.11 | 89.18 | — | 95.21 | — | |
| RetinaNet-RBackbone=ResNet1012020.03 | 89.18 | — | 95.21 | — | |
| BBAVectorsBackbone=ResNet1012020.11 | 88.6 | — | — | — | |
| BBAVectorsBackbone=ResNet1012020.03 | 88.6 | — | — | — | |
| BBAVECTORSBackbone=R-1012021.01 | 88.6 | — | — | — | |
| OPLDBackbone=R-1012021.01 | 88.44 | — | — | — | |
| Gliding VertexBackbone=ResNet1012020.11 | 88.2 | — | — | — | |
| Gliding VertexBackbone=ResNet1012020.03 | 88.2 | — | — | — | |
| GLIDING VERTEXBackbone=R-1012021.01 | 88.2 | — | — | — | |
| G. V.Pre-trained=ImageNet, Number of Parameters (#P)=41.1M, FLOPs=198G2023.03 | 88.2 | — | — | — | |
| Gliding Vertex#Params=41.1M2024.03 | 88.2 | — | — | — | |
| RSDETBackbone=R-502021.01 | 86.5 | — | — | — | |
| RoI-TransformerBackbone=ResNet1012020.11 | 86.2 | — | — | — | |
| RoI-TransformerBackbone=ResNet1012020.03 | 86.2 | — | — | — | |
| ROI-TRANS.Backbone=R-1012021.01 | 86.2 | — | — | — | |
| RoI Trans.Pre-trained=ImageNet, Number of Parameters (#P)=55.1M, FLOPs=200G2023.03 | 86.2 | — | — | — | |
| RoI Trans.#Params=55.1M2024.03 | 86.2 | — | — | — | |
| RRDBackbone=VGG162020.11 | 84.3 | — | — | — | |
| RRDBackbone=VGG162020.03 | 84.3 | — | — | — | |
| RRDBackbone=VGG162021.01 | 84.3 | — | — | — | |
| RetinaNet-HBackbone=ResNet1012020.11 | 82.89 | — | 89.27 | — | |
| RetinaNet-HBackbone=ResNet1012020.03 | 82.89 | — | 89.27 | — | |
| R2PNBackbone=VGG162020.11 | 79.6 | — | — | — | |
| R2PNBackbone=VGG162020.03 | 79.6 | — | — | — | |
| R2PNBackbone=VGG162021.01 | 79.6 | — | — | — | |
| TOSOBackbone=R-1012021.01 | 79.29 | — | — | — | |
| RRPNBackbone=ResNet1012020.11 | 79.08 | — | 85.64 | — | |
| RRPNBackbone=ResNet1012020.03 | 79.08 | — | 85.64 | — | |
| AXIS LEARNINGBackbone=R-1012021.01 | 78.15 | — | — | — | |
| RC1 & RC2Backbone=VGG162020.11 | 75.7 | — | — | — | |
| RC1 & RC2Backbone=VGG162020.03 | 75.7 | — | — | — | |
| RC1 & RC2Backbone=VGG162021.01 | 75.7 | — | — | — | |
| R2CNNBackbone=ResNet1012020.11 | 73.07 | — | 79.73 | — | |
| R2CNNBackbone=ResNet1012020.03 | 73.07 | — | 79.73 | — | |
| CenterMapEvaluation Protocol=VOC2012 metrics2021.03 | — | 92.8 | — | — | |
| CenterMapPre-trained=ImageNet, Number of Parameters (#P)=41.1M, FLOPs=198G2023.03 | — | — | 92.8 | — | |
| CenterMap#Params=41.1M2024.03 | — | — | 92.8 | — | |
| CENTERMAPBackbone=R-502021.01 | — | — | 92.8 | — | |
| CenterMap OBBBackbone=ResNet502020.11 | — | — | 92.8 | — | |
| CenterMap OBBBackbone=ResNet502020.03 | — | — | 92.8 | — | |
| CFC-NetArchitecture=Single-stage, Backbone=ResNet50, Size=416×416, NA=1, Augmentation=false, Multi-scale=false2021.01 | — | 86.3 | — | — | |
| CFC-Net (aug + ms)Architecture=Single-stage, Backbone=ResNet101, Size=800×800, NA=1, Augmentation=true, Multi-scale=true2021.01 | — | 89.7 | — | — | |
| CFC-Net (aug)Architecture=Single-stage, Backbone=ResNet50, Size=800×800, NA=1, Augmentation=true, Multi-scale=false2021.01 | — | 88.6 | — | — | |
| CFC-Net (aug)Architecture=Single-stage, Backbone=ResNet101, Size=800×800, NA=1, Augmentation=true, Multi-scale=false2021.01 | — | 89.5 | — | — | |
| CSLEvaluation Protocol=VOC2007 metrics2021.03 | — | 89.62 | — | — | |
| DRNEvaluation Protocol=VOC2012 metrics2021.03 | — | 92.7 | — | — | |
| DRNBackbone=Hourglass 1042020.11 | — | — | 92.7 | — | |
| DRNBackbone=Hourglass1042020.03 | — | — | 92.7 | — | |
| DRNBackbone=H-1042021.01 | — | — | 92.7 | — | |
| DRNPre-trained=ImageNet2023.03 | — | — | 92.7 | — | |
| DRN2024.03 | — | — | 92.7 | — | |
| DRNPre-training=IN2025.01 | — | — | 92.7 | — | |
| fully-supervisedSource=N/A, Target Train=rotated2023.04 | — | — | — | 90.3 | |
| Gliding vertexBackbone=Light-head R-CNN2019.11 | — | 87.4 | — | — | |
| Gliding vertexBackbone=ResNet-101 + FPN2019.11 | — | 88.2 | — | — | |
| Gliding VertexEvaluation Protocol=VOC2007 metrics2021.03 | — | 88.2 | — | — | |
| Gliding VertexArchitecture=Two-stage, Backbone=ResNet101, Size=512×800, NA=52021.01 | — | 88.2 | — | — | |
| GrabcutSource=N/A, Target Train=axis-aligned2023.04 | — | — | — | 24 | |
| Grabcut TrainSource=N/A, Target Train=axis-aligned2023.04 | — | — | — | 62.9 | |
| KCRSource=COCO [13], Target Train=axis-aligned2023.04 | — | — | — | 57.9 | |
| KCRSource=COCO Aug, Target Train=axis-aligned2023.04 | — | — | — | 78.3 | |
| KCRSource=DOTA [32], Target Train=axis-aligned2023.04 | — | — | — | 79.1 |