Signal and image based analysis of human fear using FPGA

Abstract

Human emotion detection is a very important part to enhance the human machine interaction. Emotions can be classified as pure emotions and mixed emotions. Classification of emotions based on single domain feature set is not perfect enough. In this paper six emotions such as fear, anger, surprise, happy, neutral and sad are considered for analysis. Facial emotions and physical parameters are considered for the analysis. Human fear is analyzed predominantly out of all six emotions. The features for each emotion are implemented using FPGA. It is seen that the number of slices and look up table are varying according to the change of emotion. The slices and lookup table are taken as features for facial emotions and combined with physical parameter features to give better result. The combined features are classified by back propagation algorithm. Out of all emotions fear emotion has the more sensitivity and specificity of 97.36% and 91.67% respectively. The sensitivity and specificity for only physical parameters and facial images are 58.62%, 79.41%, 81.25%, 47.62% respectively. The fear emotion is best classified by taking combined feature set other than single feature set like human emotional faces or physical parameters.