Deep learning-based dimensional emotion recognition combining the attention mechanism and global second-order feature representations.

• One dimensional emotion recognition method is proposed to realize the dimensional emotion recognition by combing the attention mechanism and global second-order deep representations. • The residual attention network is used to extract emotionally relevant features to solve the negative influence of emotionally irrelevant features on the recognition performance. • The global second-order pooling network is integrated to learn the interdependence relationship between the long-distance features, increasing the nonlinear emotion representation learning ability of the trained model. • The experiments on the Affectnet dataset demonstrate the validity of the proposed model which is superior to some existing methods. For the emotion recognition tasks, the emotion-related features and the emotionally irrelevant ones are fully extracted by the deep neural network is particularly desirable. In this paper, a dimensional emotion recognition algorithm is proposed by combing the attention mechanism and global second-order feature representations. More specially, the residual attention network (RAN) is utilized to select the features related to the task, and then the global second-order pooling network is employed to model the correlation between these features to complete the modeling of the feature interdependence relationship, improving the deficiency of the RAN in this point so as to enhance the capacity of extracting the emotion-related features and the emotionally irrelevant ones. The experiments on the AffectNet dataset demonstrate the validity of the proposed model, which is comparable to and even superior to those from the support vector regression method, the convolutional neural network (CNN)-based method, and the recently-developed CNN-based variants. [Display omitted] [ABSTRACT FROM AUTHOR]