Emotion Recognition through Facial Expressions using CNN

Abstract

In psychology, forensics, and social media, facial expression detection is essential. An extremely high level of classification reliability is required in all of these disciplines. Human judgement is now employed in several domains, however people are not always precise in their decisions. As a result, there is a pressing need for an accurate and rapid method of determining human emotions. Recent breakthroughs in machine learning and pattern recognition have resulted in the development of a number of algorithms for recognising human emotion. The Convolutional Neural Network (CNN) is an example of one such algorithm. The CNN is capable of doing high-speed image processing while maintaining great dependability. In this work, a CNN is utilised to detect facial expressions based on their content. It has been demonstrated that, with sufficient training, this approach may achieve extremely high classification accuracy.