EMOTION DETECTION USING SEPARABLE CONVOLUTIONS

Abstract

Emotion detection is a tremendously booming field that aims to identify and analyze human emotions using different manner such as text, speech, and facial features. This thesis provides a detailed point of the state of the art methods, applications, challenges in emotion detection. It begins with an introduction to emotions with their importance in human-computer interaction, followed by a review of the most embossed methods employed in emotion recognition. The thesis then goes into the applications of emotion detection in different field, such as mental health, education and marketing and discusses the ethical considerations and challenges faced by the field. Finally, it proposes future research ways to further advance the field of emotion detection and improve its impact in the ordinary life of human.