Image Classification on Aircraft
94.8AccuracyKNN
Evaluation Results
| Method | Links | |
|---|---|---|
| KNNSelection Method=KNN, Adaptation Strategy=Full, Training Epochs=4e, Backbone Architecture=Resnet-50-v22020.09 | 94.8 | |
| KLSelection Method=KL, Adaptation Strategy=Full, Training Epochs=2e, Backbone Architecture=Resnet-50-v22020.09 | 94.6 | |
| KNNSelection Method=KNN, Adaptation Strategy=Full, Training Epochs=2e, Backbone Architecture=Resnet-50-v22020.09 | 94.5 | |
| KLSelection Method=KL, Adaptation Strategy=Full, Training Epochs=4e, Backbone Architecture=Resnet-50-v22020.09 | 94.4 | |
| MetaFormerBackbone=MetaFormer-1, Pretrain=iNat212022.03 | 94.3 | |
| EPNSelection Method=EPN, Adaptation Strategy=Full, Training Epochs=4e, Backbone Architecture=Resnet-50-v22020.09 | 94.2 | |
| MetaFormerBackbone=MetaFormer-1, Pretrain=ImageNet-21k2022.03 | 94.2 | |
| Dom-Ad (In-v3)Backbone Architecture=Inception-v32020.09 | 94.1 | |
| CAPBackbone=Xception, Pretrain=ImageNet-1k2022.03 | 94.1 | |
| EPNSelection Method=EPN, Adaptation Strategy=Full, Training Epochs=2e, Backbone Architecture=Resnet-50-v22020.09 | 94 | |
| API-NetBackbone=DenseNet-161, Pretrain=ImageNet-1k2022.03 | 93.9 | |
| PMGBackbone=ResNet-50, Pretrain=ImageNet-1k2022.03 | 93.4 | |
| DCLBackbone=ResNet-50, Pretrain=ImageNet-1k2022.03 | 93 | |
| Dom-Ad (Am-B)Backbone Architecture=AmoebaNet-B2020.09 | 92.8 | |
| S3NBackbone=ResNet-50, Pretrain=ImageNet-1k2022.03 | 92.8 | |
| MetaFormerBackbone=MetaFormer-1, Pretrain=ImageNet-1k2022.03 | 92.8 | |
| GPipeBackbone=AmoebaNet-B, Pretrain=ImageNet-1k2022.03 | 92.7 | |
| KNNSelection Method=KNN, Adaptation Strategy=Adapters, Training Epochs=4e, Backbone Architecture=Resnet-50-v22020.09 | 92.5 | |
| KLSelection Method=KL, Adaptation Strategy=Adapters, Training Epochs=4e, Backbone Architecture=Resnet-50-v22020.09 | 92.1 | |
| EPNSelection Method=EPN, Adaptation Strategy=Adapters, Training Epochs=4e, Backbone Architecture=Resnet-50-v22020.09 | 92.1 | |
| BaselineBackbone Architecture=Resnet-50-v22020.09 | 91.4 | |
| Core-tuningBackbone=ResNet-50, Pre-training=MoCo-v22021.02 | 89.48 | |
| FNCEvaluation Protocol=Fine-tuned, Backbone=ResNet, Pre-training Dataset=ImageNet2020.11 | 88.5 | |
| BYOLArchitecture=ResNet-50, Pre-training Dataset=ImageNet, Evaluation Protocol=Fine-tuned2020.06 | 88.1 | |
| SimCLRArchitecture=ResNet-50, Pre-training Dataset=ImageNet, Evaluation Protocol=Fine-tuned2020.06 | 88.1 | |
| SimCLR v1Evaluation Protocol=Fine-tuned, Backbone=ResNet, Pre-training Dataset=ImageNet2020.11 | 88.1 | |
| BYOLEvaluation Protocol=Fine-tuned, Backbone=ResNet, Pre-training Dataset=ImageNet2020.11 | 88.1 | |
| SimCLR (repro)Architecture=ResNet-50, Pre-training Dataset=ImageNet, Evaluation Protocol=Fine-tuned2020.06 | 87.6 | |
| RIFLEBackbone=ResNet-50, Pre-training=MoCo-v22021.02 | 87.6 | |
| SimCLR v2Evaluation Protocol=Fine-tuned, Backbone=ResNet, Pre-training Dataset=ImageNet2020.11 | 87.6 | |
| M&MBackbone=ResNet-50, Pre-training=MoCo-v22021.02 | 87.45 | |
| SCLBackbone=ResNet-50, Pre-training=MoCo-v22021.02 | 87.44 | |
| Bi-tuningBackbone=ResNet-50, Pre-training=MoCo-v22021.02 | 87.39 | |
| BSSBackbone=ResNet-50, Pre-training=MoCo-v22021.02 | 87.18 | |
| DELTABackbone=ResNet-50, Pre-training=MoCo-v22021.02 | 87.05 | |
| SL-CE-tuningBackbone=ResNet-50, Pre-training=Supervised ImageNet2021.02 | 87.03 | |
| CE-tuningBackbone=ResNet-50, Pre-training=MoCo-v22021.02 | 86.87 | |
| L2SPBackbone=ResNet-50, Pre-training=MoCo-v22021.02 | 86.55 | |
| Supervised-INArchitecture=ResNet-50, Pre-training Dataset=ImageNet, Evaluation Protocol=Fine-tuned2020.06 | 86 | |
| Random initArchitecture=ResNet-50, Pre-training Dataset=ImageNet, Evaluation Protocol=Fine-tuned2020.06 | 85.9 | |
| LRWOpt2024.03 | 81.78 | |
| LRW-Hard2024.03 | 81.28 | |
| LRW-Easy2024.03 | 80.87 | |
| LRW-Random2024.03 | 80.69 | |
| MAPLE2024.03 | 80.58 | |
| FSR2024.03 | 80.55 | |
| BILAW2024.03 | 80.37 | |
| RHO-LOSS2024.03 | 80.34 | |
| ERM2024.03 | 80.22 | |
| LOUPEevaluation=linear probing2022.08 | 80.2 | |
| MBW2024.03 | 80.12 | |
| MWN2024.03 | 80.11 | |
| ZLaPBatch Size=1000, Backbone=ViT-B/16, Keff=Medium (5-25), Protocol=Batch test-time adaptation2025.01 | 75.4 | |
| HSMLWay=5, Shot=52019.05 | 73.49 | |
| CLIPevaluation=linear probing2022.08 | 69.4 | |
| MUMOMAMLWay=5, Shot=52019.05 | 67.31 | |
| CoCoOpBackbone=ViT-B/16, Evaluation Protocol=16-shot ImageNet tuned2022.09 | 66.89 | |
| Meta-SGDWay=5, Shot=52019.05 | 66.84 | |
| MAMLWay=5, Shot=52019.05 | 66.18 | |
| BMAMLWay=5, Shot=52019.05 | 65.74 | |
| Prompt EnsembleBackbone=ViT-B/16, Evaluation Protocol=Zero-shot2022.09 | 65.63 | |
| TPTBackbone=ViT-B/16, Evaluation Protocol=Zero-shot2022.09 | 65.5 | |
| CoOpBackbone=ViT-B/16, Evaluation Protocol=16-shot ImageNet tuned2022.09 | 64.15 | |
| MT-NetWay=5, Shot=52019.05 | 63.03 | |
| StableRepPre-training Data Source=Synthetic, Evaluation Protocol=5-way, 5-shot2023.06 | 62.6 | |
| StableRepPre-training dataset=CC12M, Pre-training data source=Syn, Evaluation protocol=Few-shot2023.06 | 62.6 | |
| CLIPBackbone=ViT-B/16, Evaluation Protocol=Zero-shot2022.09 | 62.59 | |
| CLIPPre-training Data Source=Real, Evaluation Protocol=5-way, 5-shot2023.06 | 62 | |
| CLIPPre-training dataset=CC12M, Pre-training data source=Real, Evaluation protocol=Few-shot2023.06 | 62 | |
| TPTBackbone=ResNet-50, Evaluation Protocol=Zero-shot2022.09 | 61.46 | |
| FNCEvaluation Protocol=Linear eval, Backbone=ResNet, Pre-training Dataset=ImageNet2020.11 | 61.3 | |
| CLIPPre-training Data Source=Synthetic, Evaluation Protocol=5-way, 5-shot2023.06 | 61.3 | |
| CLIPPre-training dataset=CC12M, Pre-training data source=Syn, Evaluation protocol=Few-shot2023.06 | 61.3 | |
| Supervised-INArchitecture=ResNet-50, Pre-training Dataset=ImageNet, Evaluation Protocol=Linear evaluation2020.06 | 61 | |
| Prompt EnsembleBackbone=ResNet-50, Evaluation Protocol=Zero-shot2022.09 | 60.85 | |
| BYOLArchitecture=ResNet-50, Pre-training Dataset=ImageNet, Evaluation Protocol=Linear evaluation2020.06 | 60.6 | |
| BYOLEvaluation Protocol=Linear eval, Backbone=ResNet, Pre-training Dataset=ImageNet2020.11 | 60.6 | |
| CoCoOpBackbone=ResNet-50, Evaluation Protocol=16-shot ImageNet tuned2022.09 | 59.61 | |
| CLIPLinear evaluation=true, Model size=Base, Patch size=16*16, Resolution=224*2242023.01 | 59.5 | |
| CLIPBackbone=ResNet-50, Evaluation Protocol=Zero-shot2022.09 | 58.8 | |
| CoOpBackbone=ResNet-50, Evaluation Protocol=16-shot ImageNet tuned2022.09 | 58.15 | |
| HSMLWay=5, Shot=12019.05 | 57.38 | |
| StableRepPre-training dataset=RedCaps, Pre-training data source=Syn, Evaluation protocol=Few-shot2023.06 | 57.3 | |
| CLIPPre-training dataset=RedCaps, Pre-training data source=Syn, Evaluation protocol=Few-shot2023.06 | 55.2 | |
| CLIPPre-training dataset=RedCaps, Pre-training data source=Real, Evaluation protocol=Few-shot2023.06 | 54.5 | |
| BMAMLWay=5, Shot=12019.05 | 53.63 | |
| MUMOMAMLWay=5, Shot=12019.05 | 53.14 | |
| PyramidCLIPTask=Linear Probe, Pretrain Dataset=143M, Image Encoder=ResNet502022.04 | 53.1 | |
| Meta-SGDWay=5, Shot=12019.05 | 52.99 | |
| IOTAShot=16-shot, Backbone=ViT-B/162026.01 | 52.84 | |
| MAMLWay=5, Shot=12019.05 | 51.37 | |
| SimCLR v2Evaluation Protocol=Linear eval, Backbone=ResNet, Pre-training Dataset=ImageNet2020.11 | 51.1 | |
| SimCLRArchitecture=ResNet-50, Pre-training Dataset=ImageNet, Evaluation Protocol=Linear evaluation2020.06 | 50.3 | |
| SimCLR v1Evaluation Protocol=Linear eval, Backbone=ResNet, Pre-training Dataset=ImageNet2020.11 | 50.3 | |
| SimCLR (repro)Architecture=ResNet-50, Pre-training Dataset=ImageNet, Evaluation Protocol=Linear evaluation2020.06 | 49.8 | |
| DaVinciLinear evaluation=true, Model size=Base, Patch size=16*16, Resolution=224*2242023.01 | 49.6 | |
| CLIPTask=Linear Probe, Pretrain Dataset=400M, Image Encoder=ResNet502022.04 | 49.1 | |
| X+OS_SimCorepretrain=Target + SimCore, p=1%2023.03 | 48.45 | |
| DeCLIPTask=Linear Probe, Pretrain Dataset=88M, Image Encoder=ResNet502022.04 | 48.4 | |
| X+OS_SimCorepretrain=Target + SimCore, p=Adaptive, Criterion=Stopping Criterion2023.03 | 48.27 |