Object Detection on NS-FB (balanced)
35.5APRRPN + ResNet-101
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| RRPN + ResNet-101Proposal Method=RRPN, Backbone=ResNet-1012019.05 | 35.5 | 59 | 37 | 42.1 | 21.1 | 39.1 | 51.4 | |
| RRPN + ResNeXt-101Proposal Method=RRPN, Backbone=ResNeXt-1012019.05 | 35.4 | 59.2 | 36.9 | 42 | 20.2 | 39.1 | 51 | |
| Selective Search + ResNet-101Proposal Method=Selective Search, Backbone=ResNet-1012019.05 | 33.6 | 54.8 | 35.7 | 38.5 | 0.1 | 29.1 | 59.1 | |
| Selective Search + ResNeXt-101Proposal Method=Selective Search, Backbone=ResNeXt-1012019.05 | 33.2 | 54.5 | 35.2 | 38.2 | 0.1 | 29.1 | 58.5 |