Image Classification on Pascal VOC 2007 (val)
96.7mAPML-Decoder + AAM
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ML-Decoder + AAMBackbone=TResNet-L, Input resolution=448x4482022.09 | 96.7 | 35.78 | |
| GAT re-weightingBackbone=TResNet-L, Input resolution=448x4482022.09 | 96.67 | 35.2 | |
| ML-DecoderBackbone=TResNet-L, Input resolution=448x4482022.09 | 96.6 | 35.78 | |
| GATNBackbone=ResNeXt-101, Input resolution=448x4482022.09 | 96.3 | 36 | |
| Q2LBackbone=TResNet-L, Input resolution=448x4482022.09 | 96.1 | 57.43 | |
| ML-Decoder + AAMBackbone=EfficientNet-V2-L, Input resolution=448x4482022.09 | 96.05 | 49.92 | |
| GAT re-weightingBackbone=EfficientNet-V2-s, Input resolution=448x4482022.09 | 96 | 10.83 | |
| ML-Decoder + AAMBackbone=EfficientNet-V2-s, Input resolution=448x4482022.09 | 95.86 | 12 | |
| ML-DecoderBackbone=EfficientNet-V2-s, Input resolution=448x448, Training strategy=ours2022.09 | 95.54 | 12 | |
| ML-GCNBackbone=EfficientNet-V2-s, Input resolution=448x448, Training strategy=ours2022.09 | 95.25 | 10.83 | |
| Q2LBackbone=EfficientNet-V2-s, Input resolution=448x448, Training strategy=ours2022.09 | 94.94 | 15.4 | |
| ASLBackbone=TResNet-L, Input resolution=448x4482022.09 | 94.6 | 43.5 | |
| ASLBackbone=EfficientNet-V2-s, Input resolution=448x448, Training strategy=ours2022.09 | 94.24 | 10.83 |