# Stage-wise Adaptive Label Distribution for Facial Age Estimation

## TL;DR

This paper introduces a novel stage-wise adaptive label distribution learning method for facial age estimation, effectively modeling varying ambiguity across age stages to improve accuracy and robustness.

## Contribution

The paper proposes SA-LDL, a new algorithm that captures stage-wise label ambiguity patterns using adaptive variance modeling and weighted loss, advancing age estimation techniques.

## Key findings

- Achieves MAE of 1.74 on MORPH-II dataset.
- Achieves MAE of 2.15 on FG-NET dataset.
- Demonstrates competitive performance through extensive experiments.

## Abstract

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.