Uncertain Label Correction via Auxiliary Action Unit Graphs for Facial Expression Recognition

TL;DR

This paper introduces ULC-AG, a novel method that corrects uncertain labels in facial expression datasets by leveraging auxiliary action unit graphs and semantic re-labeling, improving recognition accuracy.

Contribution

The paper proposes a new approach combining graph convolutional networks and a re-labeling strategy to correct uncertain labels in FER datasets, enhancing model performance.

Findings

Achieves 89.31% accuracy on RAF-DB

Achieves 61.57% accuracy on AffectNet

Outperforms baseline and state-of-the-art methods

Abstract

High-quality annotated images are significant to deep facial expression recognition (FER) methods. However, uncertain labels, mostly existing in large-scale public datasets, often mislead the training process. In this paper, we achieve uncertain label correction of facial expressions using auxiliary action unit (AU) graphs, called ULC-AG. Specifically, a weighted regularization module is introduced to highlight valid samples and suppress category imbalance in every batch. Based on the latent dependency between emotions and AUs, an auxiliary branch using graph convolutional layers is added to extract the semantic information from graph topologies. Finally, a re-labeling strategy corrects the ambiguous annotations by comparing their feature similarities with semantic templates. Experiments show that our ULC-AG achieves 89.31% and 61.57% accuracy on RAF-DB and AffectNet datasets, respectively, outperforming the baseline and state-of-the-art methods.