Image Classification on Food101 (test)
96.38AccuracySDFed
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| SDFedBackbone=ViT-B162026.02 | 96.38 | — | — | — | — | — | — | — | |
| BiFTABackbone=ViT-L/14, Zero-shot=true2026.01 | 93.97 | — | — | — | — | — | — | — | |
| WCABackbone=ViT-L/14, Zero-shot=true2026.01 | 93.93 | — | — | — | — | — | — | — | |
| DINO.txt#param=304M, Evaluation Protocol=Zero-shot2026.04 | 93.4 | — | — | — | — | — | — | — | |
| CuPLBackbone=ViT-L/14, Zero-shot=true2026.01 | 93.37 | — | — | — | — | — | — | — | |
| WaffleBackbone=ViT-L/14, Zero-shot=true2026.01 | 93.33 | — | — | — | — | — | — | — | |
| CLIP-EBackbone=ViT-L/14, Zero-shot=true2026.01 | 93.07 | — | — | — | — | — | — | — | |
| CLIP-DBackbone=ViT-L/14, Zero-shot=true2026.01 | 93.03 | — | — | — | — | — | — | — | |
| UOPPBackbone=ViT-B162026.02 | 92.94 | — | — | — | — | — | — | — | |
| GPT-FLBackbone=ViT-B162026.02 | 92.85 | — | — | — | — | — | — | — | |
| CLIPBackbone=ViT-L/14, Zero-shot=true2026.01 | 92.55 | — | — | — | — | — | — | — | |
| SDFedBackbone=ResNet502026.02 | 92.15 | — | — | — | — | — | — | — | |
| FedOTPBackbone=ViT-B162026.02 | 90.71 | — | — | — | — | — | — | — | |
| OTTERBackbone=ViT-B/16, Setting=Zero-shot2024.04 | 89.9 | — | — | — | — | — | — | — | |
| HisTPTBackbone=ViT-B/16, Protocol=Continuous Test-time Prompt Tuning2024.10 | 89.3 | — | — | — | — | — | — | — | |
| UOPPBackbone=ResNet502026.02 | 89.12 | — | — | — | — | — | — | — | |
| PowderBackbone=ViT-B162026.02 | 88.41 | — | — | — | — | — | — | — | |
| Scalable Diffusion ModelMissing Type=Image, Zero-Shot=true2026.02 | 88.37 | — | — | — | — | — | — | — | |
| GPT-FLBackbone=ResNet502026.02 | 88.34 | — | — | — | — | — | — | — | |
| SyPMissing Type=Image, Zero-Shot=true2026.02 | 87.82 | — | — | — | — | — | — | — | |
| DCPMissing Type=Image, Zero-Shot=true2026.02 | 87.74 | — | — | — | — | — | — | — | |
| MDPMissing Type=Image, Zero-Shot=true2026.02 | 87.07 | — | — | — | — | — | — | — | |
| Prior MatchingBackbone=ViT-B/16, Setting=Zero-shot2024.04 | 86.8 | — | — | — | — | — | — | — | |
| MAPMissing Type=Image, Zero-Shot=true2026.02 | 86.72 | — | — | — | — | — | — | — | |
| FedTPGnum_prompts=8, alpha=0.32024.09 | 86.59 | — | — | — | — | — | — | — | |
| PromptFolionum_prompts=8, alpha=0.32024.09 | 86.5 | — | — | — | — | — | — | — | |
| DiffTPTBackbone=ViT-B/16, Protocol=Continuous Test-time Prompt Tuning2024.10 | 86.5 | — | — | — | — | — | — | — | |
| FedOTPBackbone=ResNet502026.02 | 85.67 | — | — | — | — | — | — | — | |
| Zero-shotBackbone=ViT-B/16, Setting=Zero-shot2024.04 | 85.6 | — | — | — | — | — | — | — | |
| PromptFL+FedProxnum_prompts=8, alpha=0.3, federated_optimizer=FedProx2024.09 | 85.48 | — | — | — | — | — | — | — | |
| PromptFLnum_prompts=8, alpha=0.32024.09 | 85.15 | — | — | — | — | — | — | — | |
| SHiRABackbone=ViT-Base, Evaluation Protocol=Fine-tuned2024.06 | 84.43 | — | — | — | — | — | — | — | |
| LoRABackbone=ViT-Base, Evaluation Protocol=Fine-tuned2024.06 | 84.27 | — | — | — | — | — | — | — | |
| TPTBackbone=ViT-B/16, Protocol=Continuous Test-time Prompt Tuning2024.10 | 84.2 | — | — | — | — | — | — | — | |
| REDEEMMissing Type=Image, Zero-Shot=true2026.02 | 83.78 | — | — | — | — | — | — | — | |
| CLIPBackbone=ViT-B/16, Protocol=Zero-shot2024.10 | 83.6 | — | — | — | — | — | — | — | |
| PromptFL+FedAMPnum_prompts=8, alpha=0.3, federated_strategy=FedAMP2024.09 | 83.24 | — | — | — | — | — | — | — | |
| PowderBackbone=ResNet502026.02 | 83.21 | — | — | — | — | — | — | — | |
| CoOpnum_prompts=8, alpha=0.32024.09 | 83.09 | — | — | — | — | — | — | — | |
| Scalable Diffusion ModelMissing Type=Both, Zero-Shot=true2026.02 | 83.01 | — | — | — | — | — | — | — | |
| APArchitecture=CLIP (ViT-B/32)2026.04 | 82.43 | — | — | — | — | — | — | — | |
| RAGPTMissing Type=Image, Zero-Shot=true2026.02 | 82.42 | — | — | — | — | — | — | — | |
| PromptFL+FedPernum_prompts=8, alpha=0.3, federated_strategy=FedPer2024.09 | 82.2 | — | — | — | — | — | — | — | |
| DCPMissing Type=Both, Zero-Shot=true2026.02 | 81.83 | — | — | — | — | — | — | — | |
| AdaVFM (LARGE)#param=50M, Evaluation Protocol=Zero-shot2026.04 | 81.7 | — | — | — | — | — | — | — | |
| CoOpBackbone=ViT-B162026.02 | 81.65 | — | — | — | — | — | — | — | |
| APArchitecture=Swin-Transformer (Swin-B)2026.04 | 81.38 | — | — | — | — | — | — | — | |
| HisTPTBackbone=ResNet-50, Protocol=Continuous Test-time Prompt Tuning2024.10 | 81.3 | — | — | — | — | — | — | — | |
| SyPMissing Type=Both, Zero-Shot=true2026.02 | 81.27 | — | — | — | — | — | — | — | |
| PromptFL+FTnum_prompts=8, alpha=0.3, tuning=fine-tuning2024.09 | 80.85 | — | — | — | — | — | — | — | |
| AdaVFM (BASE)#param=36M, Evaluation Protocol=Zero-shot2026.04 | 80.5 | — | — | — | — | — | — | — | |
| MDPMissing Type=Both, Zero-Shot=true2026.02 | 79.62 | — | — | — | — | — | — | — | |
| VPArchitecture=CLIP (ViT-B/32)2026.04 | 79.61 | — | — | — | — | — | — | — | |
| AdaVFM (SMALL)#param=29M, Evaluation Protocol=Zero-shot2026.04 | 79.2 | — | — | — | — | — | — | — | |
| REDEEMMissing Type=Both, Zero-Shot=true2026.02 | 78.77 | — | — | — | — | — | — | — | |
| Scalable Diffusion ModelMissing Type=Text, Zero-Shot=true2026.02 | 78.71 | — | — | — | — | — | — | — | |
| BOFTBackbone=ViT-L, m=4, r=4, #Params=2.95M2024.05 | 78.42 | — | — | — | — | — | — | — | |
| MAPMissing Type=Both, Zero-Shot=true2026.02 | 78.42 | — | — | — | — | — | — | — | |
| SVFTBackbone=ViT-L, Variant=SVFT-B, d=8, #Params=2.5M2024.05 | 78.36 | — | — | — | — | — | — | — | |
| DiffTPTBackbone=ResNet-50, Protocol=Continuous Test-time Prompt Tuning2024.10 | 78.3 | — | — | — | — | — | — | — | |
| AdaVFM (TINY)#param=16M, Evaluation Protocol=Zero-shot2026.04 | 78.3 | — | — | — | — | — | — | — | |
| SyPMissing Type=Text, Zero-Shot=true2026.02 | 78.01 | — | — | — | — | — | — | — | |
| SVFTBackbone=ViT-L, Variant=SVFT-P, d=2, #Params=0.74M2024.05 | 77.94 | — | — | — | — | — | — | — | |
| Full-FTBackbone=ViT-L, #Params=303.3M2024.05 | 77.83 | — | — | — | — | — | — | — | |
| DCPMissing Type=Text, Zero-Shot=true2026.02 | 77.82 | — | — | — | — | — | — | — | |
| LoRA_mul + VPT_add + PACEShot=16, Backbone=ViT-B/16, Pre-trained=ImageNet-21K2024.09 | 77.8 | — | — | — | — | — | — | — | |
| RAGPTMissing Type=Both, Zero-Shot=true2026.02 | 77.54 | — | — | — | — | — | — | — | |
| REDEEMMissing Type=Text, Zero-Shot=true2026.02 | 77.52 | — | — | — | — | — | — | — | |
| LoRABackbone=ViT-L, r=8, #Params=3.54M2024.05 | 77.13 | — | — | — | — | — | — | — | |
| CoOpBackbone=ResNet502026.02 | 76.84 | — | — | — | — | — | — | — | |
| RAGPTMissing Type=Text, Zero-Shot=true2026.02 | 76.51 | — | — | — | — | — | — | — | |
| DoRABackbone=ViT-L, r=8, #Params=3.76M2024.05 | 76.41 | — | — | — | — | — | — | — | |
| LoRA_mul + VPT_addShot=16, Backbone=ViT-B/16, Pre-trained=ImageNet-21K2024.09 | 76.4 | — | — | — | — | — | — | — | |
| VPT_add + PACEShot=16, Backbone=ViT-B/16, Pre-trained=ImageNet-21K2024.09 | 76.1 | — | — | — | — | — | — | — | |
| LoRABackbone=ViT-L, r=1, #Params=0.44M2024.05 | 75.97 | — | — | — | — | — | — | — | |
| VeRABackbone=ViT-L, r=256, #Params=0.06M2024.05 | 75.97 | — | — | — | — | — | — | — | |
| SVFTBackbone=ViT-L, Variant=SVFT-P, #Params=0.05M2024.05 | 75.95 | — | — | — | — | — | — | — | |
| DoRABackbone=ViT-L, r=1, #Params=0.66M2024.05 | 75.9 | — | — | — | — | — | — | — | |
| VPArchitecture=Swin-Transformer (Swin-B)2026.04 | 75.72 | — | — | — | — | — | — | — | |
| HeadBackbone=ViT-L, #Params=null2024.05 | 75.57 | — | — | — | — | — | — | — | |
| VPT_addShot=16, Backbone=ViT-B/16, Pre-trained=ImageNet-21K2024.09 | 75.4 | — | — | — | — | — | — | — | |
| LoRA_add + PACEShot=16, Backbone=ViT-B/16, Pre-trained=ImageNet-21K2024.09 | 75.2 | — | — | — | — | — | — | — | |
| MDPMissing Type=Text, Zero-Shot=true2026.02 | 75 | — | — | — | — | — | — | — | |
| TPTBackbone=ResNet-50, Protocol=Continuous Test-time Prompt Tuning2024.10 | 74.4 | — | — | — | — | — | — | — | |
| MAPMissing Type=Text, Zero-Shot=true2026.02 | 74.16 | — | — | — | — | — | — | — | |
| CLIPBackbone=ResNet-50, Protocol=Zero-shot2024.10 | 73.9 | — | — | — | — | — | — | — | |
| AdaVFM (MIN)#param=6M, Evaluation Protocol=Zero-shot2026.04 | 73.8 | — | — | — | — | — | — | — | |
| LoRA_addShot=16, Backbone=ViT-B/16, Pre-trained=ImageNet-21K2024.09 | 71.3 | — | — | — | — | — | — | — | |
| LoRA_add + PACEShot=1, Backbone=ViT-B/16, Pre-trained=ImageNet-21K2024.09 | 40.6 | — | — | — | — | — | — | — | |
| VPT_add + PACEShot=1, Backbone=ViT-B/16, Pre-trained=ImageNet-21K2024.09 | 39.9 | — | — | — | — | — | — | — | |
| LoRA_mul + VPT_add + PACEShot=1, Backbone=ViT-B/16, Pre-trained=ImageNet-21K2024.09 | 39.3 | — | — | — | — | — | — | — | |
| LoRA_mul + VPT_addShot=1, Backbone=ViT-B/16, Pre-trained=ImageNet-21K2024.09 | 35.4 | — | — | — | — | — | — | — | |
| VPT_addShot=1, Backbone=ViT-B/16, Pre-trained=ImageNet-21K2024.09 | 34.3 | — | — | — | — | — | — | — | |
| LoRA_addShot=1, Backbone=ViT-B/16, Pre-trained=ImageNet-21K2024.09 | 33.9 | — | — | — | — | — | — | — | |
| Prototypical CurriculumShot=5-shot, Protocol=linear probing, Training epochs=4002024.11 | 19.11 | — | — | — | — | — | — | — | |
| BaselineShot=5-shot, Protocol=linear probing, Training epochs=16002024.11 | 16.97 | — | — | — | — | — | — | — | |
| Prototypical CurriculumShot=5-shot, Protocol=linear probing, Training epochs=2002024.11 | 16.64 | — | — | — | — | — | — | — | |
| Any-shift PromptingShots=162024.02 | — | — | 90.87 | 91.35 | 91.11 | — | — | — | |
| Any-shift PromptingShots=162024.02 | — | — | 90.87 | 91.35 | 91.11 | — | — | — | |
| APP2025.11 | — | — | 84.6 | 86.1 | 85.4 | — | — | — |