Weakly Supervised Object Localization on CUB-200-2011
86.6MaxBoxAccV2FSL baseline
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| FSL baselineBackbone=InceptionV32021.09 | 86.6 | — | |
| LayerCAM + oursBackbone=VGG162021.09 | 84.3 | — | |
| Best WSOL + oursBackbone=VGG162021.09 | 84.3 | — | |
| GradCAM++ + oursBackbone=VGG162021.09 | 84.1 | — | |
| Smooth-GradCAM++ + oursBackbone=VGG162021.09 | 83.1 | — | |
| GradCAM++ + oursBackbone=ResNet502021.09 | 82.7 | — | |
| LayerCAM + oursBackbone=ResNet502021.09 | 82.7 | — | |
| Best WSOL + oursBackbone=ResNet502021.09 | 82.7 | — | |
| FSL baselineBackbone=ResNet502021.09 | 82.4 | — | |
| Smooth-GradCAM++ + oursBackbone=ResNet502021.09 | 81.6 | — | |
| GradCAM + oursBackbone=ResNet502021.09 | 80.8 | — | |
| XGradCAM + oursBackbone=VGG162021.09 | 80.1 | — | |
| XGradCAM + oursBackbone=ResNet502021.09 | 80 | — | |
| GradCAM + oursBackbone=VGG162021.09 | 79.5 | — | |
| CAM* + oursBackbone=ResNet502021.09 | 79.4 | — | |
| CAM* + oursBackbone=VGG162021.09 | 79.1 | — | |
| GradCAM + oursBackbone=InceptionV32021.09 | 76.2 | — | |
| Best WSOL + oursBackbone=InceptionV32021.09 | 76.2 | — | |
| LayerCAMBackbone=VGG162021.09 | 74.1 | — | |
| Smooth-GradCAM++ + oursBackbone=InceptionV32021.09 | 74 | — | |
| LayerCAM + oursBackbone=InceptionV32021.09 | 73.9 | — | |
| GradCAM++Backbone=VGG162021.09 | 73.8 | — | |
| GradCAM++ + oursBackbone=InceptionV32021.09 | 73.1 | — | |
| LayerCAMBackbone=ResNet502021.09 | 72.6 | — | |
| FSL baselineBackbone=VGG162021.09 | 71.6 | — | |
| CAM* + oursBackbone=InceptionV32021.09 | 71.2 | — | |
| XGradCAM + oursBackbone=InceptionV32021.09 | 70.6 | — | |
| GradCAM++Backbone=ResNet502021.09 | 70.2 | — | |
| ICLBackbone=VGG162021.09 | 66.7 | — | |
| Smooth-GradCAM++Backbone=ResNet502021.09 | 66.6 | — | |
| ACoLBackbone=ResNet502021.09 | 66.5 | — | |
| Best WSOLBackbone=ResNet502021.09 | 66.5 | — | |
| ADLBackbone=VGG162021.09 | 66.3 | — | |
| Best WSOLBackbone=VGG162021.09 | 66.3 | — | |
| HaSBackbone=ResNet502021.09 | 64.7 | — | |
| Smooth-GradCAM++Backbone=VGG162021.09 | 64.1 | — | |
| CAMBackbone=VGG162021.09 | 63.7 | — | |
| HaSBackbone=VGG162021.09 | 63.7 | — | |
| GradCAMBackbone=ResNet502021.09 | 63.3 | — | |
| ICLBackbone=ResNet502021.09 | 63.2 | — | |
| XGradCAMBackbone=ResNet502021.09 | 63.2 | — | |
| CAMBackbone=ResNet502021.09 | 63 | — | |
| CutMixBackbone=ResNet502021.09 | 62.8 | — | |
| XGradCAMBackbone=VGG162021.09 | 62.8 | — | |
| GradCAMBackbone=VGG162021.09 | 62.7 | — | |
| LayerCAMBackbone=InceptionV32021.09 | 62.6 | — | |
| CutMixBackbone=VGG162021.09 | 62.3 | — | |
| CAM*Backbone=ResNet502021.09 | 62.1 | — | |
| GradCAM++Backbone=InceptionV32021.09 | 60.7 | — | |
| SPGBackbone=ResNet502021.09 | 60.4 | — | |
| ICLBackbone=InceptionV32021.09 | 60.3 | — | |
| Center baselineBackbone=VGG162021.09 | 59.7 | — | |
| Center baselineBackbone=InceptionV32021.09 | 59.7 | — | |
| Center baselineBackbone=ResNet502021.09 | 59.7 | — | |
| Smooth-GradCAM++Backbone=InceptionV32021.09 | 59.7 | — | |
| ADLBackbone=InceptionV32021.09 | 58.8 | — | |
| Best WSOLBackbone=InceptionV32021.09 | 58.8 | — | |
| ADLBackbone=ResNet502021.09 | 58.4 | — | |
| CutMixBackbone=InceptionV32021.09 | 57.5 | — | |
| ACoLBackbone=VGG162021.09 | 57.4 | — | |
| GradCAMBackbone=InceptionV32021.09 | 57.1 | — | |
| CAM*Backbone=VGG162021.09 | 57 | — | |
| CAMBackbone=InceptionV32021.09 | 56.7 | — | |
| XGradCAMBackbone=InceptionV32021.09 | 56.7 | — | |
| SPGBackbone=VGG162021.09 | 56.3 | — | |
| ACoLBackbone=InceptionV32021.09 | 56.2 | — | |
| SPGBackbone=InceptionV32021.09 | 55.9 | — | |
| CAM*Backbone=InceptionV32021.09 | 54.4 | — | |
| HaSBackbone=InceptionV32021.09 | 53.4 | — | |
| ACoLBackbone=Mean2021.09 | — | 60 | |
| ADLBackbone=Mean2021.09 | — | 61.1 | |
| Best WSOLBackbone=Mean2021.09 | — | 61.1 | |
| Best WSOL + oursBackbone=Mean2021.09 | — | 80.3 | |
| CAMBackbone=Mean2021.09 | — | 61.1 | |
| CAM*Backbone=Mean2021.09 | — | 57.8 | |
| CAM* + oursBackbone=Mean2021.09 | — | 76.5 | |
| Center baselineBackbone=Mean2021.09 | — | 59.7 | |
| CutMixBackbone=Mean2021.09 | — | 60.8 | |
| FSL baselineBackbone=Mean2021.09 | — | 80.2 | |
| GradCAMBackbone=Mean2021.09 | — | 61 | |
| GradCAM + oursBackbone=Mean2021.09 | — | 78.8 | |
| GradCAM++Backbone=Mean2021.09 | — | 68.2 | |
| GradCAM++ + oursBackbone=Mean2021.09 | — | 79.9 | |
| HaSBackbone=Mean2021.09 | — | 60.6 | |
| ICLBackbone=Mean2021.09 | — | 63.4 | |
| LayerCAMBackbone=Mean2021.09 | — | 69.7 | |
| LayerCAM + oursBackbone=Mean2021.09 | — | 80.3 | |
| Smooth-GradCAM++Backbone=Mean2021.09 | — | 63.4 | |
| Smooth-GradCAM++ + oursBackbone=Mean2021.09 | — | 79.5 | |
| SPGBackbone=Mean2021.09 | — | 57.5 | |
| XGradCAMBackbone=Mean2021.09 | — | 60.9 | |
| XGradCAM + oursBackbone=Mean2021.09 | — | 76.9 |