Human Emotion Recognition from Physiological Biosignals
Trivedi, Priyank G.
The goal of this research is to develop and evaluate methods of classifying human emotions using physiological bio-signals. As computing technology has been steadily growing and becoming more ubiquitous, it is essential that intelligent computer systems can determine the affective state of human subjects and adjust accordingly. Moreover, the assessment of a subject’s emotional state in real-time can facilitate advanced therapeutic intervention tools, such as Virtual Reality Exposure Therapy (VRET), which can complement the traditional approaches. Although physiological responses in humans have been shown to correlate with their affective state, accurately determining someone's emotion from bio-signals remains a challenge. In this work, we have developed and experimented with a set of machine learning tools to optimize the task of emotion classification using physiological bio-signals. In a series of experiments, using a publicly available dataset as well as a dataset we collected during sessions of virtual reality exposure therapy with veterans suffering from social anxiety, we evaluate the utility of automatic emotion recognition for improved human-computer interaction as well as its use as an objective metric of emotional response monitoring during therapeutic interventions.
Trivedi, P. G. (2018). <i>Human emotion recognition from physiological biosignals</i> (Unpublished thesis). Texas State University, San Marcos, Texas.