Salvato in:
Dettagli Bibliografici
Autori principali: Wu, Bo, Ai, Zhiqi, Jiang, Jun, Zhu, Congcong, Xu, Shugong
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2509.00450
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866914015063572480
author Wu, Bo
Ai, Zhiqi
Jiang, Jun
Zhu, Congcong
Xu, Shugong
author_facet Wu, Bo
Ai, Zhiqi
Jiang, Jun
Zhu, Congcong
Xu, Shugong
contents Label ambiguity poses a significant challenge in age estimation tasks. Most existing methods address this issue by modeling correlations between adjacent age groups through label distribution learning. However, they often overlook the varying degrees of ambiguity present across different age stages. In this paper, we propose a Stage-wise Adaptive Label Distribution Learning (SA-LDL) algorithm, which leverages the observation -- revealed through our analysis of embedding similarities between an anchor and all other ages -- that label ambiguity exhibits clear stage-wise patterns. By jointly employing stage-wise adaptive variance modeling and weighted loss function, SA-LDL effectively captures the complex and structured nature of label ambiguity, leading to more accurate and robust age estimation. Extensive experiments demonstrate that SA-LDL achieves competitive performance, with MAE of 1.74 and 2.15 on the MORPH-II and FG-NET datasets.
format Preprint
id arxiv_https___arxiv_org_abs_2509_00450
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Stage-wise Adaptive Label Distribution for Facial Age Estimation
Wu, Bo
Ai, Zhiqi
Jiang, Jun
Zhu, Congcong
Xu, Shugong
Computer Vision and Pattern Recognition
Label ambiguity poses a significant challenge in age estimation tasks. Most existing methods address this issue by modeling correlations between adjacent age groups through label distribution learning. However, they often overlook the varying degrees of ambiguity present across different age stages. In this paper, we propose a Stage-wise Adaptive Label Distribution Learning (SA-LDL) algorithm, which leverages the observation -- revealed through our analysis of embedding similarities between an anchor and all other ages -- that label ambiguity exhibits clear stage-wise patterns. By jointly employing stage-wise adaptive variance modeling and weighted loss function, SA-LDL effectively captures the complex and structured nature of label ambiguity, leading to more accurate and robust age estimation. Extensive experiments demonstrate that SA-LDL achieves competitive performance, with MAE of 1.74 and 2.15 on the MORPH-II and FG-NET datasets.
title Stage-wise Adaptive Label Distribution for Facial Age Estimation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2509.00450