Emotion Detection in Multimedia Data Using Convolution Neural Network

Abstract

Emotions is the way of understanding the human action mainly those conveyed with facial expression. We understand only one third of other people feeling through their words and tone of voice and rest two third comes from the facial expression. Emotion detection is carried out using deep learning approaches. Surveying social media content through text, speech, and images and video should be essential in various uses. Emotion recognition can be applied to images, videos, text and written content. In this paper we have applied emotion recognition to video as it carries human face in motion and gives more information on human feelings. In this paper deep learning techniques are used to recognise human emotions in stored video. We have considered the Convolutional Neural Networks (CNN) to forecast the different emotions namely anger, surprise, happiness and neutral portrayed in a stored video. Convolution Neural Network applied to emotion detection has provided promising results. Emotion recognition technology can be applied to activities like market research and can be beneficial for security purpose utilizing CCTVs to catch the behavior of some unspecified person and forecast either he/she is suspicious or not.