Image Classification on Places-LT
39.36Accuracy (Overall)SGD-DualOpt_Fine-tune
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| SGD-DualOpt_Fine-tuneBackbone=ConvNeXt-B2026.04 | 39.36 | — | — | — | — | |
| FullBackbone=ConvNeXt-B2026.04 | 38.61 | — | — | — | — | |
| L2-SPBackbone=ConvNeXt-B2026.04 | 38.52 | — | — | — | — | |
| Adam-DualOpt_Fine-tuneBackbone=ViT-B2026.04 | 38.07 | — | — | — | — | |
| LoRABackbone=ConvNeXt-B2026.04 | 37.11 | — | — | — | — | |
| VPTBackbone=ViT-B2026.04 | 37.02 | — | — | — | — | |
| LinearBackbone=ConvNeXt-B2026.04 | 36.41 | — | — | — | — | |
| VPTBackbone=ConvNeXt-B2026.04 | 36.32 | — | — | — | — | |
| OLTRBackbone Net=ResNet-152, Evaluation Setting=closed-set2019.04 | 35.9 | 44.7 | 37 | 25.3 | — | |
| Lifted LossBackbone Net=ResNet-152, Evaluation Setting=closed-set2019.04 | 35.2 | 41.1 | 35.4 | 24 | — | |
| Range LossBackbone Net=ResNet-152, Evaluation Setting=closed-set2019.04 | 35.1 | 41.1 | 35.4 | 23.2 | — | |
| FSLwFBackbone Net=ResNet-152, Evaluation Setting=closed-set2019.04 | 34.9 | 43.9 | 29.9 | 29.5 | — | |
| Focal LossBackbone Net=ResNet-152, Evaluation Setting=closed-set2019.04 | 34.6 | 41.1 | 34.8 | 22.4 | — | |
| LoRABackbone=ViT-B2026.04 | 33.3 | — | — | — | — | |
| LinearBackbone=ViT-B2026.04 | 31.95 | — | — | — | — | |
| FullBackbone=ViT-B2026.04 | 31.59 | — | — | — | — | |
| L2-SPBackbone=ViT-B2026.04 | 31.22 | — | — | — | — | |
| Plain ModelBackbone Net=ResNet-152, Evaluation Setting=closed-set2019.04 | 27.2 | 45.9 | 22.4 | 0.36 | — | |
| Focal LossBackbone Net=ResNet-152, Evaluation Setting=open-set2019.04 | — | 41 | 34.8 | 22.3 | 45.3 | |
| FSLwFBackbone Net=ResNet-152, Evaluation Setting=open-set2019.04 | — | 38.1 | 19.5 | 14.8 | 37.5 | |
| Lifted LossBackbone Net=ResNet-152, Evaluation Setting=open-set2019.04 | — | 41 | 35.2 | 23.8 | 45.9 | |
| OLTRBackbone Net=ResNet-152, Evaluation Setting=open-set2019.04 | — | 44.6 | 36.8 | 25.2 | 46.4 | |
| Plain ModelBackbone Net=ResNet-152, Evaluation Setting=open-set2019.04 | — | 45.9 | 22.4 | 0.36 | 36.6 | |
| Range LossBackbone Net=ResNet-152, Evaluation Setting=open-set2019.04 | — | 41 | 35.3 | 23.1 | 45.7 | |
| Range Loss + OpenMaxBackbone Net=ResNet-152, Evaluation Setting=open-set2019.04 | — | 41.1 | 35.4 | 23.2 | 45.8 |