GReFEL: Geometry-Aware Reliable Facial Expression Learning under Bias and Imbalanced Data Distribution

TL;DR

GReFEL introduces a geometry-aware, transformer-based approach to improve facial expression recognition by addressing bias and data imbalance, leading to more reliable and accurate predictions across diverse demographics.

Contribution

The paper proposes a novel geometry-aware, anchor-based module within a Vision Transformer framework to mitigate bias and imbalance in facial expression learning.

Findings

Outperforms state-of-the-art methods on multiple datasets.

Effectively reduces bias and mislabeling in facial expression predictions.

Enhances reliability and accuracy across diverse demographic groups.

Abstract

Reliable facial expression learning (FEL) involves the effective learning of distinctive facial expression characteristics for more reliable, unbiased and accurate predictions in real-life settings. However, current systems struggle with FEL tasks because of the variance in people's facial expressions due to their unique facial structures, movements, tones, and demographics. Biased and imbalanced datasets compound this challenge, leading to wrong and biased prediction labels. To tackle these, we introduce GReFEL, leveraging Vision Transformers and a facial geometry-aware anchor-based reliability balancing module to combat imbalanced data distributions, bias, and uncertainty in facial expression learning. Integrating local and global data with anchors that learn different facial data points and structural features, our approach adjusts biased and mislabeled emotions caused by intra-class disparity, inter-class similarity, and scale sensitivity, resulting in comprehensive, accurate, and reliable facial expression predictions. Our model outperforms current state-of-the-art methodologies, as demonstrated by extensive experiments on various datasets.