Pedestrian Detection on Caltech
5.92MRNMS-Ped
Evaluation results
| Method | Links | |
|---|---|---|
| NMS-PedBackbone=ResNet-502021.06 | 5.92 | |
| SSA-CNNBackbone=VGG-162021.06 | 6.27 | |
| W2NetBackbone=ResNet-502021.06 | 6.37 | |
| AR-PedBackbone=VGG-162021.06 | 6.45 | |
| TFAN+TDEM+PRMBackbone=ResNet-1012021.06 | 6.5 | |
| Our baselineBackbone=ResNet-502021.06 | 6.61 | |
| MGANBackbone=VGG-162021.06 | 6.83 | |
| SDS-RCNNBackbone=VGG-162021.06 | 7.36 | |
| F-DNNBackbone=ResNet-502021.06 | 8.65 | |
| RPN+BFBackbone=VGG-162021.06 | 9.58 | |
| PATHBackbone=ViT-B, Evaluation Protocol=Partial FT2023.03 | 28.3 | |
| PATHBackbone=ViT-L, Evaluation Protocol=Partial FT2023.03 | 28.7 | |
| SoTA †Note=multitask or stronger models2023.03 | 28.8 | |
| PATHBackbone=ViT-B, Evaluation Protocol=FT2023.03 | 30.1 | |
| PATHBackbone=ViT-B, Evaluation Protocol=Head FT2023.03 | 31.1 | |
| SoTA2023.03 | 46.6 | |
| MAEBackbone=ViT-B2023.03 | 48.1 |