Dynamic Probabilistic Graph Convolution for Facial Action Unit Intensity Estimation

Deep learning methods have been widely applied to automatic facial action unit (AU) intensity estimation and achieved state-of-the-art performance. These methods, however, are mostly appearance-based and fail to exploit the underlying structural information among the AUs. In this paper, we propose a novel dynamic probabilistic graph convolution (DPG) model to simultaneously exploit AU appearances, AU dynamics, and their semantic structural dependencies for AU intensity estimation. First, we propose to use Bayesian Network to capture the inherent dependencies among the AUs. Second, we introduce probabilistic graph convolution that allows to perform graph convolution on the distribution of Bayesian Network structure to extract AU structural features. Finally, we introduce a dynamic deep model based on LSTM to simultaneously combine AU appearance features, AU dynamic features, and AU structural features for improved AU intensity estimation. In experiments, our method achieves comparable and even better performance with state-of-the-art methods on two benchmark facial AU intensity estimation databases, i.e., FERA 2015 and DISFA.