List your company
ResearchFormerly WizWand

Image Classification on ISIC

84.93AccuracyInsert LN
16.986834.625952.26569.9041Mar 2, 2023Aug 30, 2023Feb 28, 2024Aug 27, 2024Feb 25, 2025Aug 25, 2025Feb 23, 2026
Updated 4mo ago

Evaluation results

MethodLinks
Insert LNBackbone=ResNet50, Training Scenario=FT2025.0484.93
T = 1Backbone=ResNet50, Training Scenario=FT2025.0484.72
Insert LNBackbone=SwinT, Training Scenario=FT2025.0484.51
Ours (T*)Backbone=ResNet50, Training Scenario=FT2025.0484.39
Ours (T*csgcn)Backbone=ResNet50, Training Scenario=FT2025.0484.38
T = 1Backbone=SwinT, Training Scenario=FT2025.0484.36
Ours (T*cn)Backbone=ResNet50, Training Scenario=FT2025.0483.88
Ours (T*csg)Backbone=SwinT, Training Scenario=FT2025.0483.64
T = sqrt(M)Backbone=SwinT, Training Scenario=FT2025.0483.46
T = sqrt(M)Backbone=ResNet50, Training Scenario=FT2025.0482.75
Ours (T*csgcn)Backbone=SwinT, Training Scenario=FT2025.0481.32
Ours (T*)Backbone=SwinT, Training Scenario=FT2025.0480.88
Ours (T*cn)Backbone=SwinT, Training Scenario=FT2025.0480.83
Ours (T*csg)Backbone=ResNet50, Training Scenario=FT2025.0480.58
T = 1Backbone=ResNet50, Training Scenario=init2025.0476.16
Insert LNBackbone=ResNet50, Training Scenario=init2025.0475.17
T = sqrt(M)Backbone=ResNet50, Training Scenario=init2025.0474.07
Ours (T*csg)Backbone=ResNet50, Training Scenario=init2025.0472.16
Ours (T*cn)Backbone=ResNet50, Training Scenario=init2025.0470.03
Ours (T*csgcn)Backbone=ResNet50, Training Scenario=init2025.0469.32
Ours (T*)Backbone=ResNet50, Training Scenario=init2025.0466.57
CBBudget (%)=1002024.1165.7
BADGEBudget (%)=1002024.1165.59
PCBBudget (%)=1002024.1164.54
CBBudget (%)=752024.1164.41
CBBudget (%)=502024.1164.01
BADGEBudget (%)=752024.1163.89
CBBudget (%)=252024.1163.38
EntropyBudget (%)=1002024.1162.25
BADGEBudget (%)=502024.1162.1
PCBBudget (%)=752024.1162.1
EntropyBudget (%)=752024.1161.65
EntropyBudget (%)=502024.1161.35
CoreSetBudget (%)=1002024.1160.91
CoreSetBudget (%)=752024.1160.81
CoreSetBudget (%)=502024.1159.21
BADGEBudget (%)=252024.1158.37
EntropyBudget (%)=252024.1157.87
PCBBudget (%)=502024.1157.52
CoreSetBudget (%)=252024.1153.69
PCBBudget (%)=252024.1151.49
SupervisedBackbone=ResNet-50, Pre-trained=ImageNet, Shot=5-shot, Way=5-way, Number of tasks=6002023.0343.34
BYOL + PsCoBackbone=ResNet-50, Pre-trained=ImageNet, Shot=5-shot, Way=5-way, Number of tasks=6002023.0342.9
MoCo v2 + PsCoBackbone=ResNet-50, Pre-trained=ImageNet, Shot=5-shot, Way=5-way, Number of tasks=6002023.0341.64
MoCo v2Backbone=ResNet-50, Pre-trained=ImageNet, Shot=5-shot, Way=5-way, Number of tasks=6002023.0338.26
BYOLBackbone=ResNet-50, Pre-trained=ImageNet, Shot=5-shot, Way=5-way, Number of tasks=6002023.0337.27
GEOMMeta-Album Size=Extended, Training Dataset=Meta-Album2025.0731.7
GEOMMeta-Album Size=Mini, Training Dataset=Meta-Album2025.0731.38
ChimeraLoRAEvaluation Protocol=Few-shot (Augmented), Shot Count=4 Real + 500 Synthetic2026.0229.2
GEOM-INTraining Dataset=ImageNet-1k2025.0727.9
GEOMMeta-Album Size=Micro, Training Dataset=Meta-Album2025.0725.69
LoFTEvaluation Protocol=Few-shot (Augmented), Shot Count=4 Real + 500 Synthetic2026.0225.6
IsSynthEvaluation Protocol=Few-shot (Augmented), Shot Count=4 Real + 500 Synthetic2026.0223.8
CLIPEvaluation Protocol=Zero-shot, Shot Count=02026.0221.1
DataDreamEvaluation Protocol=Few-shot (Augmented), Shot Count=4 Real + 500 Synthetic2026.0220.7
CLIPEvaluation Protocol=Few-shot (Real-only), Shot Count=42026.0219.6