Image Classification on Flowers-102 (test)
98.8Top-1 AccuracyEfficientNet-B7
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| EfficientNet-B7Architecture=EfficientNet-B72021.06 | 98.8 | — | — | — | |
| DeiT-B/16Architecture=DeiT-B/16, distillation=true2021.06 | 98.8 | — | — | — | |
| ViT-B/16Pretraining + MaskSub=true, Finetuning + MaskSub=true2023.06 | 98.7 | — | — | — | |
| AdaBetBackbone=ViT2025.10 | 98.55 | — | — | — | |
| XCiT-L24/16Architecture=XCiT-L24/16, distillation=true2021.06 | 98.3 | — | — | — | |
| ViT-S/16Pretraining + MaskSub=true, Finetuning + MaskSub=true2023.06 | 98.3 | — | — | — | |
| XCiT-M24/16Architecture=XCiT-M24/16, distillation=true2021.06 | 98.2 | — | — | — | |
| PruneTrainBackbone=ViT2025.10 | 98.12 | — | — | — | |
| PytorchBackbone=ResNet-50, Pre-training recipe=[1]2021.10 | 97.9 | — | — | — | |
| A1Backbone=ResNet-502021.10 | 97.9 | — | — | — | |
| A2Backbone=ResNet-502021.10 | 97.9 | — | — | — | |
| TransBoostEvaluation Protocol=Transductive, Resolution=224, Backbone=ResNet50, Classes=102, Train (+Validation) size=7,169, Test size=1,0202022.05 | 97.85 | 2.99 | — | — | |
| Elastic TrainerBackbone=ViT2025.10 | 97.71 | — | — | — | |
| ViT-B/16Pretraining + MaskSub=true, Finetuning + MaskSub=false2023.06 | 97.7 | — | — | — | |
| CoOpBackbone=ViT-B/16, Shots=16-shots, Context length (M)=42023.03 | 97.63 | — | 81.23 | — | |
| A3Backbone=ResNet-502021.10 | 97.5 | — | — | — | |
| ViT-B/16Pretraining + MaskSub=false, Finetuning + MaskSub=false2023.06 | 97.5 | — | — | — | |
| XCiT-S24/16Architecture=XCiT-S24/16, distillation=true2021.06 | 97.4 | — | — | — | |
| CGC2021.10 | 96.18 | — | — | — | |
| Baseline2021.10 | 96.09 | — | — | — | |
| Last-K LayersBackbone=ViT2025.10 | 95.97 | — | — | — | |
| ProGradBackbone=ViT-B/16, Shots=16-shots, Context length (M)=42023.03 | 95.54 | — | 82.03 | — | |
| Transfer LearningBackbone=ViT2025.10 | 95.45 | — | — | — | |
| ViT-S/16Pretraining + MaskSub=true, Finetuning + MaskSub=false2023.06 | 95.2 | — | — | — | |
| SL-MLPEvaluation Protocol=Full-data finetuning, Pre-trained Dataset=ImageNet-1K, Pre-training Epochs=3002021.12 | 95.12 | — | — | — | |
| SupConEvaluation Protocol=Full-data finetuning, Pre-trained Dataset=ImageNet-1K, Pre-training Epochs=3002021.12 | 95.1 | — | — | — | |
| KgCoOpBackbone=ViT-B/16, Shots=16-shots, Context length (M)=42023.03 | 95 | — | 83.65 | — | |
| CoCoOpBackbone=ViT-B/16, Shots=16-shots, Context length (M)=42023.03 | 94.87 | — | 81.71 | — | |
| Standard InductiveEvaluation Protocol=Inductive, Resolution=224, Backbone=ResNet50, Classes=102, Train (+Validation) size=7,169, Test size=1,0202022.05 | 94.86 | — | — | — | |
| SupCon w/o MLPEvaluation Protocol=Full-data finetuning, Pre-trained Dataset=ImageNet-1K, Pre-training Epochs=3002021.12 | 94.6 | — | — | — | |
| ViT-S/16Pretraining + MaskSub=false, Finetuning + MaskSub=false2023.06 | 94.5 | — | — | — | |
| SLEvaluation Protocol=Full-data finetuning, Pre-trained Dataset=ImageNet-1K, Pre-training Epochs=3002021.12 | 94.31 | — | — | — | |
| AiRPrompting Strategy=Text Prompt, Learning Paradigm=TRZSL2025.04 | 92.3 | — | — | — | |
| AiRPrompting Strategy=Text Prompt, Learning Paradigm=SSL2025.04 | 91.2 | — | — | — | |
| ViT-L/16Architecture=ViT-L/162021.06 | 89.7 | — | — | — | |
| CPLYear=2024, Prompting Strategy=Text Prompt, Learning Paradigm=SSL2025.04 | 89.6 | — | — | — | |
| ViT-B/16Architecture=ViT-B/162021.06 | 89.5 | — | — | — | |
| SimSiam + DIFFUSEMIXFramework=Self-Supervised Learning, Augmentation=DIFFUSEMIX2024.04 | 89.24 | — | — | — | |
| CPLYear=2024, Prompting Strategy=Text Prompt, Learning Paradigm=TRZSL2025.04 | 87.3 | — | — | — | |
| SimSiamFramework=Self-Supervised Learning2024.04 | 86.93 | — | — | — | |
| GRIPYear=2023, Prompting Strategy=Text Prompt, Learning Paradigm=TRZSL2025.04 | 86.8 | — | — | — | |
| US3LBackbone=ResNet-18, Width=1.0x, Params=11.69M, MACS=1.82G, Evaluation Protocol=linear evaluation2023.03 | 84.9 | — | — | — | |
| GRIPYear=2023, Prompting Strategy=Text Prompt, Learning Paradigm=SSL2025.04 | 83.6 | — | — | — | |
| US3LBackbone=ResNet-18, Width=0.75x, Params=6.68M, MACS=1.05G, Evaluation Protocol=linear evaluation2023.03 | 83.5 | — | — | — | |
| BYOLBackbone=ResNet-18, Width=1.0x, Params=11.69M, MACS=1.82G, Evaluation Protocol=linear evaluation2023.03 | 83 | — | — | — | |
| BYOLBackbone=ResNet-18, Width=0.75x, Params=6.68M, MACS=1.05G, Evaluation Protocol=linear evaluation2023.03 | 82.6 | — | — | — | |
| MoCo v2 + DIFFUSEMIXFramework=Self-Supervised Learning, Augmentation=DIFFUSEMIX2024.04 | 82.15 | — | — | — | |
| US3LBackbone=ResNet-18, Width=0.5x, Params=3.06M, MACS=0.49G, Evaluation Protocol=linear evaluation2023.03 | 82 | — | — | — | |
| AiRPrompting Strategy=Visual Prompt, Learning Paradigm=TRZSL2025.04 | 81.9 | — | — | — | |
| BYOLBackbone=ResNet-18, Width=0.5x, Params=3.06M, MACS=0.49G, Evaluation Protocol=linear evaluation2023.03 | 81.4 | — | — | — | |
| AdaBetBackbone=ResNet502025.10 | 81.39 | — | — | — | |
| FPLYear=2023, Prompting Strategy=Text Prompt, Learning Paradigm=TRZSL2025.04 | 80.9 | — | — | — | |
| Elastic TrainerBackbone=ResNet502025.10 | 80.6 | — | — | — | |
| Full TrainingBackbone=ResNet502025.10 | 80.56 | — | — | — | |
| PruneTrainBackbone=ResNet502025.10 | 80.45 | — | — | — | |
| MoCo v2Framework=Self-Supervised Learning2024.04 | 80.31 | — | — | — | |
| CPLYear=2024, Prompting Strategy=Visual Prompt, Learning Paradigm=TRZSL2025.04 | 80.1 | — | — | — | |
| Last-K LayersBackbone=ResNet502025.10 | 79.16 | — | — | — | |
| Fisher InformationBackbone=ResNet502025.10 | 79.05 | — | — | — | |
| Full TrainingBackbone=ViT2025.10 | 78.66 | — | — | — | |
| AiRPrompting Strategy=Visual Prompt, Learning Paradigm=SSL2025.04 | 78.5 | — | — | — | |
| CEConvTraining Augmentation=AugMix2023.10 | 78.1 | — | — | — | |
| Transfer LearningBackbone=ResNet502025.10 | 77.18 | — | — | — | |
| GRIPYear=2023, Prompting Strategy=Visual Prompt, Learning Paradigm=TRZSL2025.04 | 77.1 | — | — | — | |
| US3LBackbone=ResNet-18, Width=0.25x, Params=0.83M, MACS=0.14G, Evaluation Protocol=linear evaluation2023.03 | 77 | — | — | — | |
| CoOpYear=2022, Prompting Strategy=Text Prompt, Learning Paradigm=SSL2025.04 | 76.8 | — | — | — | |
| FPLYear=2023, Prompting Strategy=Text Prompt, Learning Paradigm=SSL2025.04 | 75.9 | — | — | — | |
| Transfer LearningBackbone=VGG162025.10 | 75.83 | — | — | — | |
| BYOLBackbone=ResNet-18, Width=0.25x, Params=0.83M, MACS=0.14G, Evaluation Protocol=linear evaluation2023.03 | 75.5 | — | — | — | |
| BaselineTraining Augmentation=AugMix2023.10 | 75.49 | — | — | — | |
| CIConv-WTraining Augmentation=Color Jitter2023.10 | 75.05 | — | — | — | |
| Elastic TrainerBackbone=VGG162025.10 | 74.86 | — | — | — | |
| AdaBetBackbone=VGG162025.10 | 74.71 | — | — | — | |
| AiRPrompting Strategy=Text Prompt, Learning Paradigm=UL2025.04 | 74.3 | — | — | — | |
| Fisher InformationBackbone=VGG162025.10 | 74.21 | — | — | — | |
| CEConvTraining Augmentation=Color Jitter2023.10 | 74.17 | — | — | — | |
| DIFFUSEMIXBackbone=ResNet-18, Images per class=102024.04 | 74.12 | — | — | — | |
| DIFFUSEMIXBackbone=ResNet-18, Data Regime=10 images per class2024.04 | 74.12 | — | — | — | |
| CPLYear=2024, Prompting Strategy=Visual Prompt, Learning Paradigm=SSL2025.04 | 73.5 | — | — | — | |
| Last-K LayersBackbone=VGG162025.10 | 73.5 | — | — | — | |
| CPLYear=2024, Prompting Strategy=Text Prompt, Learning Paradigm=UL2025.04 | 72.9 | — | — | — | |
| DePTBaseline=CoCoOp2023.09 | 72.17 | — | — | — | |
| CLIPBackbone=ViT-B/16, Shots=16-shots, Context length (M)=42023.03 | 72.08 | — | 74.83 | — | |
| FPLYear=2023, Prompting Strategy=Visual Prompt, Learning Paradigm=TRZSL2025.04 | 71.9 | — | — | — | |
| CoCoOp2023.09 | 71.8 | — | — | — | |
| CEConv-2Training Augmentation=Color Jitter2023.10 | 71.72 | — | — | — | |
| HisTPTBackbone=ViT-B/16, Protocol=Continuous Test-time Prompt Tuning2024.10 | 71.2 | — | — | — | |
| DePTBaseline=PromptSRC2023.09 | 70.93 | — | — | — | |
| PromptSRC2023.09 | 70.8 | — | — | — | |
| DePTBaseline=KgCoOp2023.09 | 70.57 | — | — | — | |
| DePTBaseline=CoOp2023.09 | 70.5 | — | — | — | |
| GuidedMixupBackbone=ResNet-18, Images per class=102024.04 | 70.44 | — | — | — | |
| GuidedMixupBackbone=ResNet-18, Data Regime=10 images per class2024.04 | 70.44 | — | — | — | |
| KgCoOp2023.09 | 70.3 | — | — | — | |
| DePTBaseline=MaPLe2023.09 | 70.1 | — | — | — | |
| MaPLe2023.09 | 70.03 | — | — | — | |
| GRIPYear=2023, Prompting Strategy=Text Prompt, Learning Paradigm=UL2025.04 | 69.8 | — | — | — | |
| DiffTPTBackbone=ViT-B/16, Protocol=Continuous Test-time Prompt Tuning2024.10 | 69.4 | — | — | — | |
| Guided-SRBackbone=ResNet-18, Images per class=102024.04 | 69.31 | — | — | — | |
| AiRPrompting Strategy=Visual Prompt, Learning Paradigm=UL2025.04 | 69.2 | — | — | — |