Adaptive Graph-Based Feature Normalization for Facial Expression Recognition

TL;DR

This paper introduces an Adaptive Graph-based Feature Normalization method for facial expression recognition that leverages expression associations to improve robustness against data uncertainties and mislabeled data.

Contribution

It proposes a novel graph-based normalization technique with a Poisson graph generator and coordinate descent optimization to enhance FER accuracy under noisy conditions.

Findings

Achieves 91.84% and 91.11% accuracy on FERPlus and RAF-DB datasets.

Outperforms state-of-the-art methods, especially with increased mislabeled data.

Surpasses existing methods by 3.38% and 4.52% when 20% of data is mislabeled.

Abstract

Facial Expression Recognition (FER) suffers from data uncertainties caused by ambiguous facial images and annotators' subjectiveness, resulting in excursive semantic and feature covariate shifting problem. Existing works usually correct mislabeled data by estimating noise distribution, or guide network training with knowledge learned from clean data, neglecting the associative relations of expressions. In this work, we propose an Adaptive Graph-based Feature Normalization (AGFN) method to protect FER models from data uncertainties by normalizing feature distributions with the association of expressions. Specifically, we propose a Poisson graph generator to adaptively construct topological graphs for samples in each mini-batches via a sampling process, and correspondingly design a coordinate descent strategy to optimize proposed network. Our method outperforms state-of-the-art works with accuracies of 91.84% and 91.11% on the benchmark datasets FERPlus and RAF-DB, respectively, and when the percentage of mislabeled data increases (e.g., to 20%), our network surpasses existing works significantly by 3.38% and 4.52%.