An End-to-End Deep Model With Discriminative Facial Features for Facial Expression Recognition

Due to the complex challenges of environment and emotion expressions, most facial expression recognition systems cannot achieve a high rate. More discriminative features can describe expressions more accurately, so feature extraction is key technology for recognition. In this article, an effective end-to-end deep model proposed improve accuracy face Considering importance data pre-processing (very few studies have focused on process), first, enhancement method locate range target enhance image contrast. Next, obtain further features, hybrid representation proposed, in which four typical are combined. After that, designed train test samples optimal parameters with less computation cost. Ablation study results show that help accuracy. Finally, comprehensively evaluate performance model, series experiments conducted three benchmark datasets. The rate achieved 94.5%, 98.6%, 97.2% FER2013, AR dataset, CK+ respectively.