List your company
ResearchFormerly WizWand

Image Classification on CIFAR-100 (Top-1 Accuracy Focus)

95.6Top-1 AccuracyDINOv2
68.164875.287482.4189.5326Mar 23, 2022Dec 7, 2022Aug 23, 2023May 8, 2024Jan 22, 2025Oct 8, 2025Jun 24, 2026
Updated 4mo ago

Evaluation results

MethodLinks
DINOv2architecture=ViT-g/142024.1195.6
AIMv2architecture=ViT-3B/14, Resolution=448px2024.1194.5
TF-TAS-S#Param=23M, Fine-tuned=true, Resolution=384x3842022.0391.2
AutoFormer-S#Param=23M, Fine-tuned=true, Resolution=384x3842022.0391.1
DeiT-B#Param=86M, Fine-tuned=true, Resolution=384x3842022.0390.8
ViT-B/16#Param=86M, Fine-tuned=false2022.0387.1
ResNet-110 + KCLABackbone=ResNet-110, Params(M)=1.242026.0676.98
ResNet-110 + DLA-LBackbone=ResNet-110, Params(M)=1.412026.0676.82
ResNet-110 + DIANetBackbone=ResNet-110, Params(M)=1.392026.0676.54
ResNet-110 + SEBackbone=ResNet-110, Params(M)=1.302026.0676.35
ResNet-110 + MRLA-LBackbone=ResNet-110, Params(M)=1.242026.0675.81
ResNet-56 + KCLABackbone=ResNet-56, Params(M)=0.652026.0675.45
ResNet-110Backbone=ResNet-110, Params(M)=1.172026.0675.34
ResNet-56 + Layer AttBackbone=ResNet-56, Params(M)=0.642026.0675.31
ResNet-56 + DLA-LBackbone=ResNet-56, Params(M)=0.822026.0675.18
ResNet-56 + MRLA-BBackbone=ResNet-56, Params(M)=0.642026.0675.12
ResNet-56 + MRLA-LBackbone=ResNet-56, Params(M)=0.652026.0674.74
ResNet-56 + DIANetBackbone=ResNet-56, Params(M)=0.832026.0674.69
ResNet-56 + SEBackbone=ResNet-56, Params(M)=0.682026.0674.58
ResNet-56Backbone=ResNet-56, Params(M)=0.612026.0673.27
ResNet-20 + KCLABackbone=ResNet-20, Params(M)=0.262026.0672.25
ResNet-20 + Layer AttBackbone=ResNet-20, Params(M)=0.252026.0672.09
ResNet-20 + MRLA-BBackbone=ResNet-20, Params(M)=0.252026.0672.05
ResNet-20 + DLA-LBackbone=ResNet-20, Params(M)=0.432026.0671.84
ResNet-20 + MRLA-LBackbone=ResNet-20, Params(M)=0.252026.0671.36
ResNet-20 + SEBackbone=ResNet-20, Params(M)=0.272026.0670.6
ResNet-20 + DIANetBackbone=ResNet-20, Params(M)=0.462026.0670.01
ResNet-20Backbone=ResNet-20, Params(M)=0.242026.0669.22