Fine-grained Image Classification on CUB-200
91.3Accuracy (All)VAN
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| VANBackbone=VAN-B4, ImageNet-22K pre-trained=true2022.02 | 91.3 | — | — | — | — | |
| ViTBackbone=ViT-B_16, ImageNet-22K pre-trained=true2022.02 | 90.3 | — | — | — | — | |
| DeiTBackbone=DeiT-B, ImageNet-22K pre-trained=true2022.02 | 90 | — | — | — | — | |
| S2-FracMixBackbone=ViT-B, Evaluation Protocol=Transfer Learning (Fine-tuned)2026.06 | 89.84 | — | — | — | — | |
| AutoMixBackbone=ViT-B, Evaluation Protocol=Transfer Learning (Fine-tuned)2026.06 | 88.91 | — | — | — | — | |
| AdAutoMixBackbone=ViT-B, Evaluation Protocol=Transfer Learning (Fine-tuned)2026.06 | 88.76 | — | — | — | — | |
| MixUpBackbone=ViT-B, Evaluation Protocol=Transfer Learning (Fine-tuned)2026.06 | 88.75 | — | — | — | — | |
| ConvNeXt-TinyBackbone=ConvNeXt-Tiny2026.02 | 88.4 | — | — | — | — | |
| PuzzleMixBackbone=ViT-B, Evaluation Protocol=Transfer Learning (Fine-tuned)2026.06 | 88.23 | — | — | — | — | |
| VanillaBackbone=ViT-B, Evaluation Protocol=Transfer Learning (Fine-tuned)2026.06 | 88 | — | — | — | — | |
| CutMixBackbone=ViT-B, Evaluation Protocol=Transfer Learning (Fine-tuned)2026.06 | 87.76 | — | — | — | — | |
| ProtoQuantBackbone=ConvNeXt-Tiny2026.02 | 87.6 | — | — | — | — | |
| DeiT-SmallBackbone=DeiT-Small2026.02 | 86.9 | — | — | — | — | |
| ProtoQuantBackbone=DeiT-Small2026.02 | 86.9 | — | — | — | — | |
| Q-SENNBackbone=ResNet-502026.02 | 85.9 | — | — | — | — | |
| PROTOPOOLBackbone=ResNet-502026.02 | 85.5 | — | — | — | — | |
| PROTOARGNETBackbone=ResNet-502026.02 | 85.4 | — | — | — | — | |
| PROTOVITBackbone=DeiT-Small2026.02 | 85.4 | — | — | — | — | |
| PROTOPFORMERBackbone=DeiT-Small2026.02 | 84.9 | — | — | — | — | |
| ResNet-50Backbone=ResNet-101, ImageNet-22K pre-trained=false2022.02 | 84.5 | — | — | — | — | |
| ResNet-50Backbone=ResNet-502026.02 | 84.5 | — | — | — | — | |
| S2-FracMixBackbone=ResNet-50, Evaluation Protocol=Transfer Learning (Fine-tuned)2026.06 | 84.42 | — | — | — | — | |
| PIP-NETBackbone=ConvNeXt-Tiny2026.02 | 84.3 | — | — | — | — | |
| INFODISENTBackbone=ConvNeXt-Tiny2026.02 | 84.1 | — | — | — | — | |
| MCPNETBackbone=ConvNeXt-Tiny2026.02 | 83.5 | — | — | — | — | |
| PromptSRCZero-shot=true, Training=mixed-dataset training2026.06 | 83.5 | — | 50.7 | — | 63.1 | |
| AdAutoMixBackbone=ResNet-50, Evaluation Protocol=Transfer Learning (Fine-tuned)2026.06 | 83.36 | — | — | — | — | |
| INFODISENTBackbone=ResNet-502026.02 | 83 | — | — | — | — | |
| ToolFGZero-shot=true, Training=mixed-dataset training2026.06 | 83 | — | 72.7 | — | 77.5 | |
| AutoMixBackbone=ResNet-50, Evaluation Protocol=Transfer Learning (Fine-tuned)2026.06 | 82.93 | — | — | — | — | |
| MixUpBackbone=ResNet-50, Evaluation Protocol=Transfer Learning (Fine-tuned)2026.06 | 82.79 | — | — | — | — | |
| PuzzleMixBackbone=ResNet-50, Evaluation Protocol=Transfer Learning (Fine-tuned)2026.06 | 82.59 | — | — | — | — | |
| ProtoQuantBackbone=ResNet-502026.02 | 82.5 | — | — | — | — | |
| PIP-NETBackbone=ResNet-502026.02 | 82 | — | — | — | — | |
| VanillaBackbone=ResNet-50, Evaluation Protocol=Transfer Learning (Fine-tuned)2026.06 | 81.76 | — | — | — | — | |
| CutMixBackbone=ResNet-50, Evaluation Protocol=Transfer Learning (Fine-tuned)2026.06 | 81.67 | — | — | — | — | |
| LUCID PPNBackbone=ConvNeXt-Tiny2026.02 | 81.5 | — | — | — | — | |
| MaPLeZero-shot=true, Training=mixed-dataset training2026.06 | 77.8 | — | 45.8 | — | 57.7 | |
| DiVE-kZero-shot=true, Training=mixed-dataset training2026.06 | 76.8 | — | 61.3 | — | 68.2 | |
| Fine-R1Zero-shot=true, Training=mixed-dataset training2026.06 | 73.2 | — | 55.3 | — | 63 | |
| ViRFTZero-shot=true, Training=mixed-dataset training2026.06 | 71.3 | — | 58.3 | — | 64.2 | |
| InfoSieve2023.10 | 69.4 | 77.9 | 65.2 | — | — | |
| DCCL2023.10 | 63.5 | 60.8 | 64.9 | — | — | |
| Qwen2.5-VL-7BZero-shot=true, Training=mixed-dataset training2026.06 | 63.3 | — | 48.2 | — | 54.7 | |
| PromptCAL2023.10 | 62.9 | 64.4 | 62.1 | — | — | |
| SimGCD2023.10 | 60.3 | 65.6 | 57.7 | — | — | |
| XCon2023.10 | 52.1 | 54.3 | 51 | — | — | |
| GPC2023.10 | 52 | 55.5 | 47.5 | — | — | |
| GCD2023.10 | 51.3 | 56.6 | 48.7 | — | — | |
| ORCA2023.10 | 36.3 | 43.8 | 32.6 | — | — | |
| UNO+2023.10 | 35.1 | 49 | 28.1 | — | — | |
| k-means2023.10 | 34.3 | 38.9 | 32.1 | — | — | |
| RankStats+2023.10 | 33.3 | 51.6 | 24.2 | — | — | |
| DiVE-kshot=42026.06 | — | — | — | 63.9 | — | |
| Fine-R1shot=42026.06 | — | — | — | 61.9 | — | |
| MaPLeshot=42026.06 | — | — | — | 60.9 | — | |
| PromptSRCshot=42026.06 | — | — | — | 70.3 | — | |
| Qwen2.5-VL-7Bshot=42026.06 | — | — | — | 51.6 | — | |
| ToolFGshot=42026.06 | — | — | — | 72 | — | |
| ViRFTshot=42026.06 | — | — | — | 60 | — |