AU-Guided Unsupervised Domain-Adaptive Facial Expression Recognition

Abstract

Domain diversities, including inconsistent annotation and varied image collection conditions, inevitably exist among different facial expression recognition (FER) datasets, posing an evident challenge for adapting FER models trained on one dataset to another one. Recent works mainly focus on domain-invariant deep feature learning with adversarial learning mechanisms, ignoring the sibling facial action unit (AU) detection task, which has obtained great progress. Considering that AUs objectively determine facial expressions, this paper proposes an AU-guided unsupervised domain-adaptive FER (AdaFER) framework to relieve the annotation bias between different FER datasets. In AdaFER, we first leverage an advanced model for AU detection on both a source and a target domain. Then, we compare the AU results to perform AU-guided annotating, i.e., target faces that own the same AUs as source faces would inherit the labels from the source domain. Meanwhile, to achieve domain-invariant compact features, we utilize an AU-guided triplet training, which randomly collects anchor–positive–negative triplets on both domains with AUs. We conduct extensive experiments on several popular benchmarks and show that AdaFER achieves state-of-the-art results on all these benchmarks.