TABLE 2.

Comparison of age estimates on the FG-NET dataset.

Method	MAE
Mean-Variance Loss (Pan et al., 2018)	4.10
DRFs (Dagher and Barbara, 2021)	3.85
M-LSDML (Liu et al., 2017)	3.74
DLDFL (Shen et al., 2019)	3.71
DFR (Shen et al., 2019)	3.47
DAG-VGG16 (Taheri and Toygar, 2019)	3.08
DAG-GoogleNet (Panis et al., 2016)	3.05
AGEn (Tan et al., 2017)	2.96
C3AE (Zhang et al., 2019)	2.95
ADPF (Wang et al., 2022)	2.86
Swin Transformer (Liu Z. et al., 2021)	2.69
Anet+Gnet+Rnet (Deng et al., 2021)	2.59
BridgeNet (Hou et al., 2016)	2.56
ABC+Swin Transformer (Ours)	2.52

Bold values represent the best data.