List your company
ResearchFormerly WizWand

Fine-grained Classification on Cars

91.66Accuracyw/ LaViD
15.989635.634855.2874.9252Nov 25, 2022Jul 1, 2023Feb 4, 2024Sep 9, 2024Apr 15, 2025Nov 19, 2025Jun 25, 2026
Updated 4mo ago

Evaluation results

MethodLinks
w/ LaViDStudent=MNV2, Teacher=QN+R502026.0691.66
LSStudent=MNV2, Teacher=RN-502026.0691.27
w/ LaViDStudent=RN-18, Teacher=QN+R502026.0691.2
LSStudent=RN-18, Teacher=RN-502026.0690.98
MLKDStudent=MNV2, Teacher=RN-502026.0690.3
MLKDStudent=RN-18, Teacher=RN-502026.0690.23
w/ LaViDStudent=SNV2, Teacher=QN+R502026.0689.74
LSStudent=SNV2, Teacher=RN-502026.0689.68
MLKDStudent=SNV2, Teacher=RN-502026.0689.44
DKDStudent=RN-18, Teacher=RN-502026.0689.31
DKDStudent=MNV2, Teacher=RN-502026.0689.3
LaViDStudent=RN-18, Teacher=Qwen2026.0688.59
DKDStudent=SNV2, Teacher=RN-502026.0688.4
MaKDStudent=RN-18, Teacher=InternVL2026.0687.96
LaViDStudent=MNV2, Teacher=Qwen2026.0687.93
KDStudent=MNV2, Teacher=RN-502026.0687.33
KDStudent=RN-18, Teacher=RN-502026.0686.94
Ind StudentStudent=MNV2, Teacher=None2026.0686.93
CRDStudent=RN-18, Teacher=LLaVA2026.0686.86
MaKDStudent=MNV2, Teacher=InternVL2026.0686.83
FitNetStudent=MNV2, Teacher=LLaVA2026.0686.77
CLIPPre-training Dataset=WIT-400M, Pre-training Data Type=Real, Backbone=ViT-B/162023.0686.7
CRDStudent=MNV2, Teacher=LLaVA2026.0686.59
RKDStudent=MNV2, Teacher=RN-502026.0686.5
LaViDStudent=SNV2, Teacher=Qwen2026.0686.37
FitNetStudent=RN-18, Teacher=LLaVA2026.0685.9
RKDStudent=RN-18, Teacher=RN-502026.0685.78
Ind StudentStudent=RN-18, Teacher=None2026.0685.77
Ind StudentStudent=SNV2, Teacher=None2026.0685.67
MaKDStudent=SNV2, Teacher=InternVL2026.0685.47
KDStudent=SNV2, Teacher=RN-502026.0685.3
FitNetStudent=SNV2, Teacher=LLaVA2026.0685.04
RKDStudent=SNV2, Teacher=RN-502026.0685
CRDStudent=SNV2, Teacher=LLaVA2026.0684.83
StableRepPre-training Dataset=RedCaps, Pre-training Data Type=Syn, Pre-training Schedule=105 epochs, Backbone=ViT-B/162023.0683.9
StableRepPre-training Dataset=CC12M, Pre-training Data Type=Syn, Pre-training Schedule=105 epochs, Backbone=ViT-B/162023.0683.5
StableRepPre-training Dataset=CC12M, Pre-training Data Type=Syn, Pre-training Schedule=35 epochs, Backbone=ViT-B/162023.0680.9
StableRepPre-training Dataset=RedCaps, Pre-training Data Type=Syn, Pre-training Schedule=35 epochs, Backbone=ViT-B/162023.0678.8
CLIPPre-training Dataset=RedCaps, Pre-training Data Type=Syn, Pre-training Schedule=35 epochs, Backbone=ViT-B/162023.0677.3
GIF-SDBackbone=ResNet-50, Expansion Ratio=20x2022.1175.7
CLIPPre-training Dataset=RedCaps, Pre-training Data Type=Real, Pre-training Schedule=35 epochs, Backbone=ViT-B/162023.0674.9
CLIPBackbone=CLIP, Expansion Ratio=1x, Evaluation Protocol=Zero-shot2022.1155.8
SimCLRPre-training Dataset=RedCaps, Pre-training Data Type=Syn, Pre-training Schedule=35 epochs, Backbone=ViT-B/162023.0653.7
GIF-DALLEBackbone=ResNet-50, Expansion Ratio=20x2022.1153.1
SDBackbone=ResNet-50, Expansion Ratio=20x2022.1151.7
DALL-E2Backbone=ResNet-50, Expansion Ratio=20x2022.1148.3
GIF-MAEBackbone=ResNet-50, Expansion Ratio=20x2022.1144.5
RandAugmentBackbone=ResNet-50, Expansion Ratio=20x2022.1143.2
SimCLRPre-training Dataset=RedCaps, Pre-training Data Type=Real, Pre-training Schedule=35 epochs, Backbone=ViT-B/162023.0642.8
GridMaskBackbone=ResNet-50, Expansion Ratio=20x2022.1128.4
MAEBackbone=ResNet-50, Expansion Ratio=20x2022.1125.9
CutoutBackbone=ResNet-50, Expansion Ratio=20x2022.1125.8
OriginalBackbone=ResNet-50, Expansion Ratio=1x2022.1119.8
Distillation of CLIPBackbone=ResNet-50, Expansion Ratio=1x, Evaluation Protocol=Knowledge Distillation2022.1118.9