This paper presents a new method to recognize posed and spontaneous expressions through modeling their spatial patterns. Gender and expression categories are employed as privileged information to further improve the recognition. The proposed approach includes three steps. First, geometric features about facial shape and Action Unit variations are extracted from the differences between apex and onset facial images to capture the spatial facial variation. Second, statistical hypothesis testings are conducted to explore the differences between posed and spontaneous expressions using the defined geometric features from three aspects: all samples, samples given the gender information, and samples given expression categories. Third, several Bayesian networks are built to capture posed and spontaneous spatial facial patterns respectively given gender and expression categories. The statistical analysis results on the USTC-NVIE and SPOS databases both demonstrate the effectiveness of the proposed geometric features. The recognition results on the USTC-NVIE database indicate that the privileged information of gender and expression can help model the spatial patterns caused by posed and spontaneous expressions. The recognition results on both databases outperform those of the state of the art.
In this paper, we introduce methods to differentiate posed expressions from spontaneous ones by capturing global spatial patterns embedded in posed and spontaneous expressions, and by incorporating gender and expression categories as privileged information during spatial pattern modeling. Specifically, we construct multiple Restricted Boltzmann Machines (RBMs) with continuous visible units to model spatial patterns from facial geometric features given expression-related factors, i.e. gender and expression categories. During testing, only facial geometric features are provided, and the samples are classified into posed or spontaneous expressions according to the RBM with the largest likelihood. Furthermore, we propose efficient inference algorithm by extending annealing importance sampling to RBM with continuous visible units for calculating partition function of RBMs. Experimental results on benchmark databases demonstrate the effectiveness of the proposed approach in modelling global spatial patterns as well as its superior posed and spontaneous expression distinction performance over existing approaches.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.