List your company

Image Classification on ImageNet-C (val)

33.8mCEMAE
31.60446.42761.2576.073Jul 5, 2021Apr 5, 2022Jan 5, 2023Oct 7, 2023Jul 8, 2024Apr 9, 2025Jan 9, 2026

Evaluation results

MethodLinks
MAEModel=ViT-H, Resolution=224, Training Objective=MAE pre-training2021.1133.8————————————————
MAEModel=ViT-H, Resolution=448, Training Objective=MAE pre-training2021.1136.8————————————————
Source only2025.0938.724.326.125.122.223.235.443.248.449.356.97026.845.143.741.4—
LAME2025.0938.824.326.125.122.223.235.443.248.449.356.97026.845.143.741.4—
ETA2025.0940.826.428.427.223.524.637.245.15150.858.870.629.147.746.945—
Supervised BaselineModel=ViT-H, Resolution=224, Training Objective=Supervised2021.1141.3————————————————
SAR2025.0941.526.529.22824.525.337.445.151.75159.170.531.548.248.646.3—
SHOT2025.0941.72830.128.825263845.751.55159.170.630.248.447.845.8—
Tent2025.0941.727.83028.824.925.93845.551.35159.170.63048.247.845.7—
MAEModel=ViT-L, Resolution=224, Training Objective=MAE pre-training2021.1141.8————————————————
Supervised BaselineModel=ViT-L, Resolution=224, Training Objective=Supervised2021.1142.3————————————————
DeepAugmentNotes=Previous Best2021.1142.5————————————————
STL (FAN-L-Hybrid)Params (M)=77.32024.0142.5————————————————
FAN-L-HybridParams (M)=76.82024.0143————————————————
TransNeXt-BaseResolution=224x224, Parameters=89.7M, FLOPs=18.4G2023.1143.5————————————————
STL (FAN-B-Hybrid)Params (M)=50.92024.0143.6————————————————
MaxViT-BaseResolution=224x224, Parameters=119.5M, FLOPs=24.0G2023.1143.6————————————————
TransNeXt-SmallResolution=224x224, Parameters=49.7M, FLOPs=10.3G2023.1143.9————————————————
DAT-AugReg-ViTParams (M)=862024.0144.7————————————————
FAN-B-HybridParams (M)=50.42024.0145.2————————————————
Supervised BaselineModel=ViT-B, Resolution=224, Training Objective=Supervised2021.1145.8————————————————
PVTv2-B5Resolution=224x224, Parameters=82.0M, FLOPs=11.8G2023.1145.9————————————————
MaxViT-SmallResolution=224x224, Parameters=68.9M, FLOPs=11.7G2023.1146.4————————————————
TransNeXt-TinyResolution=224x224, Parameters=28.2M, FLOPs=5.7G2023.1146.5————————————————
PVTv2-B4Resolution=224x224, Parameters=62.6M, FLOPs=10.1G2023.1146.5————————————————
MSNArchitecture=ViT-B/16, Pre-training method=MSN2022.0446.6————————————————
ConvNext-BParams (M)=88.62024.0146.8————————————————
RVT-BParams (M)=91.82024.0146.8————————————————
ConvNeXt-BResolution=224x224, Parameters=88.6M, FLOPs=15.4G2023.1146.8————————————————
BiFormer-BResolution=224x224, Parameters=56.8M, FLOPs=9.8G2023.1147.2————————————————
STL (FAN-S-Hybrid)Params (M)=26.52024.0147.3————————————————
FAN-S-HybridParams (M)=26.32024.0147.8————————————————
PVTv2-B3Resolution=224x224, Parameters=45.2M, FLOPs=6.9G2023.1148————————————————
iBOTBackbone=ViT-S/16, Pre-training epochs=800, Evaluation protocol=Linear Probing2021.1148.1————————————————
BiFormer-SResolution=224x224, Parameters=25.5M, FLOPs=4.5G2023.1148.5————————————————
DeiT-Base/16Resolution=224x224, Parameters=86.6M, FLOPs=17.6G2023.1148.5————————————————
XCIT-S24Params (M)=47.72024.0149.4————————————————
ConvNeXt-SResolution=224x224, Parameters=50.2M, FLOPs=8.7G2023.1149.5————————————————
FocalNet-B (LRF)Resolution=224x224, Parameters=88.7M, FLOPs=15.4G2023.1149.5————————————————
MaxViT-TinyResolution=224x224, Parameters=30.9M, FLOPs=5.6G2023.1149.6————————————————
LV-ViT-MParams (M)=562024.0150.5————————————————
TransNeXt-MicroResolution=224x224, Parameters=12.8M, FLOPs=2.7G2023.1150.8————————————————
FocalNet-S (LRF)Resolution=224x224, Parameters=50.3M, FLOPs=8.7G2023.1151————————————————
ConvNext-SParams (M)=50.22024.0151.2————————————————
RVT-SParams (M)=23.32024.0151.4————————————————
XCIT-S12Params (M)=26.32024.0151.5————————————————
MAEModel=ViT-B, Resolution=224, Training Objective=MAE pre-training2021.1151.7————————————————
DINOBackbone=ViT-S/16, Pre-training epochs=800, Evaluation protocol=Linear Probing2021.1151.7————————————————
MAE-ViT-BParams (M)=862024.0151.7————————————————
MAEArchitecture=ViT-B/16, Pre-training method=MAE2022.0451.7————————————————
PVTv2-B2Resolution=224x224, Parameters=25.4M, FLOPs=4.0G2023.1152.6————————————————
Swin-SParams (M)=502024.0152.7————————————————
LV-ViT-SParams (M)=262024.0152.9————————————————
ConvNext-TParams (M)=28.62024.0153.2————————————————
ConvNeXt-TResolution=224x224, Parameters=28.6M, FLOPs=4.5G2023.1153.2————————————————
TRUST naive2025.0953.443.445.644.938.336.65354.960.257.16672.250.25961.158.5—
Swin-BParams (M)=87.82024.0154.4————————————————
Swin-BResolution=224x224, Parameters=87.8M, FLOPs=15.4G2023.1154.5————————————————
DeiT-Small/16Resolution=224x224, Parameters=22.1M, FLOPs=4.6G2023.1154.6————————————————
Swin-SResolution=224x224, Parameters=49.6M, FLOPs=8.7G2023.1154.9————————————————
FocalNet-T (LRF)Resolution=224x224, Parameters=28.6M, FLOPs=4.5G2023.1155————————————————
CvT*-LS-21Params (M)=32.12021.0755.2————————————————
BiFormer-TResolution=224x224, Parameters=13.1M, FLOPs=2.2G2023.1155.7————————————————
TRUST2025.0956.146.849.448.542.840.857.157.961.757.366.871.954.961.463.660.2—
CvT-21Params (M)=322021.0756.2————————————————
DeiT-SParams (M)=22.12021.0757.1————————————————
CvT*-LS-13Params (M)=20.32021.0758.7————————————————
LSNet-BFLOPS=1.3G2025.0359.3————————————————
CvT-13Params (M)=202021.0759.6————————————————
Swin-TParams (M)=28.32024.0159.6————————————————
Swin-TResolution=224x224, Parameters=28.3M, FLOPs=4.5G2023.1162————————————————
PVTv2-B1FLOPS=2.1G2025.0362.2————————————————
PVTv2-B1Resolution=224x224, Parameters=14.0M, FLOPs=2.1G2023.1162.6————————————————
FastViT-T12FLOPS=1.4G2025.0364.3————————————————
LSNet-SFLOPS=0.5G2025.0365.7————————————————
UniRepLKNet-AFLOPS=0.6G2025.0367————————————————
PoolFormer-S12FLOPS=1.8G2025.0367.7————————————————
LSNet-TFLOPS=0.3G2025.0368.2————————————————
FasterNet-T2FLOPS=1.9G2025.0370.8————————————————
EfficientViT-M3FLOPS=0.3G2025.0371.1————————————————
FastViT-T8FLOPS=0.7G2025.0372.1————————————————
EdgeNeXt-SFLOPS=1.3G2025.0372.1————————————————
RN101Backbone=ResNet-1012026.0173.168.369.167.874.270.571.872.373.572.175.879.276.473.974.677.1—
PVTv2-B0FLOPS=0.6G2025.0375.4————————————————
ViTBackbone=ViT2026.0175.771.27270.576.873.474.575.176.274.878.381.578.976.577.379.6—
ResNet50Architecture=ResNet50, Pre-training method=Supervised2022.0476.7————————————————
StarNet-S1FLOPS=0.4G2025.0377.5————————————————
SwinBackbone=Swin2026.0178.474.575.273.879.176.277.37878.977.580.683.281.379.279.882—
ResNet-50Params (M)=25.62021.0778.9————————————————
PVT-TinyResolution=224x224, Parameters=13.2M, FLOPs=1.9G2023.1179.6————————————————
PVT-TinyFLOPS=1.9G2025.0379.6————————————————
HVTBackbone=HVT2026.0181.678.979.377.682.479.880.781.281.880.483.185.68482.182.584.8—
MoCo-v2 + Patch-base NSalpha=2, k=327682021.1087.2————————————————
MoCo-v2 + Patch-base NSalpha=2, k=163842021.1087.5————————————————
Patch-base NSalpha=2, base_method=MoCo-v22021.1087.6————————————————
MoCo-v2 + Patch-base NSalpha=2, k=655362021.1087.6————————————————
MoCo-v22021.1087.7————————————————
MoCo-v22021.1087.7————————————————
EdgeNeXt-XSFLOPS=0.5G2025.0388.4————————————————
MoCHibase_method=MoCo-v22021.1088.7————————————————