Search papers, labs, and topics across Lattice.
This paper introduces UAU-Net, a framework for facial action unit (AU) detection that explicitly models uncertainty at both the representation and decision stages. At the representation stage, a conditional VAE (CV-AFE) learns probabilistic AU representations by estimating feature means and variances across spatio-temporal scales, conditioned on AU labels to capture inter-AU dependencies. At the decision stage, an Asymmetric Beta Evidential Neural Network (AB-ENN) parameterizes predictive uncertainty with Beta distributions and uses an asymmetric loss to address label imbalance. Experiments on BP4D and DISFA datasets demonstrate improved AU detection performance and robustness.
By explicitly modeling uncertainty in both feature representations and classification, UAU-Net achieves state-of-the-art facial action unit detection, highlighting the importance of uncertainty-awareness for robust performance.
Facial action unit (AU) detection remains challenging because it involves heterogeneous, AU-specific uncertainties arising at both the representation and decision stages. Recent methods have improved discriminative feature learning, but they often treat the AU representations as deterministic, overlooking uncertainty caused by visual noise, subject-dependent appearance variations, and ambiguous inter-AU relationships, all of which can substantially degrade robustness. Meanwhile, conventional point-estimation classifiers often provide poorly calibrated confidence, producing overconfident predictions, especially under the severe label imbalance typical of AU datasets. We propose UAU-Net, an Uncertainty-aware AU detection framework that explicitly models uncertainty at both stages. At the representation stage, we introduce CV-AFE, a conditional VAE (CVAE)-based AU feature extraction module that learns probabilistic AU representations by jointly estimating feature means and variances across multiple spatio-temporal scales; conditioning on AU labels further enables CV-AFE to capture uncertainty associated with inter-AU dependencies. At the decision stage, we design AB-ENN, an Asymmetric Beta Evidential Neural Network for multi-label AU detection, which parameterizes predictive uncertainty with Beta distributions and mitigates overconfidence via an asymmetric loss tailored to highly imbalanced binary labels. Extensive experiments on BP4D and DISFA show that UAU-Net achieves strong AU detection performance, and further analyses indicate that modeling uncertainty in both representation learning and evidential prediction improves robustness and reliability.
