Automated human behavioral analysis framework using facial feature extraction and machine learning
Demiyan Smirnov, Sean Banger, Sara R. Davis, Rajani Muraleedharan, Ravi P. Ramachandran · 2013
Emotional intelligence is essential in understanding and predicting human behavior. Although human emotion is best captured using non-intrusive methods, due to factors such as system complexity, computation time and decision response time, the reality of automated behavioral analysis is hindered. In this paper, we propose a framework capable of recognizing emotions of an individual to identify any suspicious behavior. Our research shows 91.1% of emotion classification accuracy for cooperative individuals using facial feature extraction and machine learning techniques, thus outperforming existing state-of-the-art approaches.