Object Detection on IOD-Video (test)
42.19AP@0.5TEA (Spatio-temporal Aggregation)
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| TEA (Spatio-temporal Aggregation)Base Backbone=ResNet-50, Pretrained Model=ImageNet2022.06 | 42.19 | 8.66 | 22.97 | 9.95 | 15.69 | |
| MOCBase Backbone=DLA-34, Pretrained Model=COCO, Type=Video-based Detector2022.06 | 36.81 | 8.94 | 19.96 | 9.62 | 14.29 | |
| CRCNNBase Backbone=ResNet101&FPN, Pretrained Model=ImageNet, Type=Video-based Detector2022.06 | 36.15 | 7.46 | 18.84 | 9.14 | 13.52 | |
| MOC + FlowBase Backbone=DLA-34, Pretrained Model=COCO, Type=Video-based Detector2022.06 | 34.5 | 7.28 | 18.18 | 8.49 | 12.75 | |
| Faster RCNNBase Backbone=ResNet-50, Pretrained Model=ImageNet, Type=Frame-based Detector2022.06 | 33.49 | 6.76 | 16.52 | 8.82 | 12.31 | |
| SSDBase Backbone=ResNet-50, Pretrained Model=ImageNet, Type=Frame-based Detector2022.06 | 30.21 | 3.95 | 12.78 | 7.82 | 9.99 | |
| CenterNetBase Backbone=ResNet-50, Pretrained Model=ImageNet, Type=Frame-based Detector2022.06 | 24.8 | 4.32 | 11.21 | 6.65 | 8.5 |