List your company
ResearchFormerly WizWand

Image Classification on Pets (Top-1 accuracy)

95.4Top-1 AccuracyCapPa L/14
12.40833.95455.577.046Jun 1, 2023Dec 1, 2023Jun 2, 2024Dec 2, 2024Jun 4, 2025Dec 4, 2025Jun 6, 2026
Updated 4mo ago

Evaluation results

MethodLinks
CapPa L/14MAP head=true, Grain=Fine, Frozen representation=true, backbone=ViT-L/142023.0695.4
OSTBProtocol=Final Ensemble2026.0695.2
OriginalBackbone=DeiT-B, P. Budget=100, Comp. Ratio=–2024.1194.74
OursShots=16, Backbone=ViT-B/16, Seeds=32025.1294.6
FiPS+RigL FTBackbone=DeiT-B, P. Budget=40%, Comp. Ratio=(∼33%)2024.1194.52
FiPS+RigL FTBackbone=DeiT-B, P. Budget=50%, Comp. Ratio=(∼25%)2024.1194.41
CapPaMAP head=true, Grain=Fine, Frozen representation=true2023.0694.4
CLIP* L/14MAP head=true, Grain=Fine, Frozen representation=true, backbone=ViT-L/142023.0694.4
GFMBackbone=DeiT-B, P. Budget=50%, Comp. Ratio=(∼25%)2024.1194.22
OursShots=8, Backbone=ViT-B/16, Seeds=32025.1294.2
FiPS+RigL FTBackbone=DeiT-B, P. Budget=25%, Comp. Ratio=(∼45%)2024.1194.19
GFMBackbone=DeiT-B, P. Budget=40%, Comp. Ratio=(∼33%)2024.1193.95
OursShots=4, Backbone=ViT-B/16, Seeds=32025.1293.7
LlipData Size=2.5B, Vision Encoder=ViT-B/16, Evaluation Protocol=Zero-shot2024.1293.5
OursShots=1, Backbone=ViT-B/16, Seeds=32025.1292.9
Image-TextIntra-modal=✗, Classifier=Zero-Shot, Backbone=ViT-B/16-open2026.0392.8
OursShots=2, Backbone=ViT-B/16, Seeds=32025.1291.9
MetaCLIPData Size=2.5B, Vision Encoder=ViT-B/16, Evaluation Protocol=Zero-shot2024.1291.7
StableRepPre-training Data Source=Synthetic, Evaluation Protocol=5-way, 5-shot2023.0691.7
CLIPPre-training Data Source=Synthetic, Evaluation Protocol=5-way, 5-shot2023.0691.5
CapMAP head=true, Grain=Fine, Frozen representation=true2023.0691.5
CLIP* (8k)MAP head=true, Grain=Fine, Frozen representation=true2023.0691.2
CLIP* (16k)MAP head=true, Grain=Fine, Frozen representation=true2023.0691.2
CLIPPre-training Data Source=Real, Evaluation Protocol=5-way, 5-shot2023.0691.1
IsoCLIPIntra-modal=✓, Classifier=NCM, Backbone=ViT-B/16-open2026.0390.6
OpenCLIPData Size=2B, Vision Encoder=ViT-B/16, Evaluation Protocol=Zero-shot2024.1289.5
Image-ImageIntra-modal=✓, Classifier=NCM, Backbone=ViT-B/16-open2026.0388.9
CLIP-B/32Backbone=CLIP-B/32, Protocol=Zero-shot2026.0687.44
ITOZero-shot=true, Backbone=ViT-B/16, Pre-training Dataset=Laion100M, Epochs=302026.0383.6
CLIPZero-shot=true, Backbone=ViT-B/16, Pre-training Dataset=Laion100M, Epochs=302026.0382.7
SimCLRPre-training Data Source=Synthetic, Evaluation Protocol=5-way, 5-shot2023.0674.6
SimCLRPre-training Data Source=Real, Evaluation Protocol=5-way, 5-shot2023.0674.4
SLIPZero-shot=true, Backbone=ViT-B/16, Pre-training Dataset=Laion100M, Epochs=302026.0374.3
DreamLIPData Size=30M, Vision Encoder=ViT-B/16, Evaluation Protocol=Zero-shot2024.1264.1
FLAIRData Size=30M, Vision Encoder=ViT-B/16, Evaluation Protocol=Zero-shot2024.1255.6
PHyCLIPw/ boxes=true2025.1054.18
HyCoCLIPw/ boxes=true2025.1053.33
SigLIPData Size=30M, Vision Encoder=ViT-B/16, Evaluation Protocol=Zero-shot2024.1253.3
CLIPData Size=30M, Vision Encoder=ViT-B/16, Evaluation Protocol=Zero-shot2024.1249.3
CLIPw/ boxes=true2025.1046.19
CLIPw/ boxes=false2025.1045.89
CyCLIPProtocol=Linear probing2026.0544.2
DINORANKCLIPProtocol=Linear probing2026.0543.5
MERUw/ boxes=false2025.1043.36
MERUw/ boxes=true2025.1043.22
ALIPProtocol=Linear probing2026.0541.9
RANKCLIPProtocol=Linear probing2026.0540.6
CLIPProtocol=Linear probing2026.0540.4
FLAIRData Size=3M, Vision Encoder=ViT-B/16, Evaluation Protocol=Zero-shot2024.1234.2
MLLM-AData Size=3M, Vision Encoder=ViT-B/16, Evaluation Protocol=Zero-shot2024.1232.1
DreamLIPData Size=3M, Vision Encoder=ViT-B/16, Evaluation Protocol=Zero-shot2024.1227.4
CLIPData Size=3M, Vision Encoder=ViT-B/16, Evaluation Protocol=Zero-shot2024.1222.1
SigLIPData Size=3M, Vision Encoder=ViT-B/16, Evaluation Protocol=Zero-shot2024.1220.6
LaCLIPData Size=3M, Vision Encoder=ViT-B/16, Evaluation Protocol=Zero-shot2024.1215.6