Improving the Rate of Facial Emotion Recognition using CNN Based Attention Capsule Networks

Developing the emotional perception of computers is a technological trend. Emotion recognition composes cognitive systems with applicability in several areas. Facial expression is an effective way to recognize emotions, mainly because it is less intrusive in data collection when compared to other methods, and because of the ease of obtaining images of the face in view of the popularization of cameras. Through facial expressions it is possible to classify the group of basic emotions. Currently, convolution neural networks (CNN) have been the state of the art for image classification. Given this context, this work presents an approach to recognize emotions by facial expression using CNN based on attention capsule networks. On imaging datasets, the majority of conventional techniques work admirably. However, more on difficult data with more image variance and incomplete faces, they do not function along with they do in controlled conditions. It is found that, the attention capsule networks algorithm can achieve satisfactory classification result, and that it can serve as a useful empirical model for development on users’ emotion identification and graphical representation of emotions in intelligence learning.