List your company
ResearchFormerly WizWand

Gender Classification on UTKFace Cross-dataset

1.6DoBCLIP + NL
1.54361.92432.3052.6857Aug 17, 2022
Updated 8mo ago

Evaluation results

MethodLinks
CLIP + NLBackbone=CLIP, Method=Normalizing Likelihood2022.081.696.4794.59
CLIP + LPBackbone=CLIP, Protocol=Linear Probing2022.081.8696.5892.91
Baseline2022.081.9694.6792.6
CLIP + MT + NL + EDLBackbone=CLIP, Multi-task learning=true, Method=Normalizing Likelihood, Evidential Deep Learning=true2022.082.0496.5292.85
NL (No LR)Method=Normalizing Likelihood, Learning Rate Ratio=None2022.082.2694.7691.55
MT + NL + EDLMulti-task learning=true, Method=Normalizing Likelihood, Evidential Deep Learning=true2022.082.2794.591.53
MT + NLMulti-task learning=true, Method=Normalizing Likelihood2022.082.3194.4790.91
NL (LR=2)Method=Normalizing Likelihood, Learning Rate Ratio=22022.082.5594.5490.11
MTMulti-task learning=true2022.083.0194.2588.89