Anxiety-based affective communication for implicit human–machine interaction

An implicit human–machine interaction framework that is sensitive to human anxiety is presented. The overall goal is to achieve detection and recognition of anxiety based on physiological signals. This involves building an anxiety-recognition system capable of interpreting the information contained in physiological processes to predict the probable anxiety state. Since anxiety plays an important role in various human–machine interaction tasks and can be related to task performance, the presented anxiety-recognition methods can be potentially applied to the design of advanced machines and engineering systems capable of intelligent decision-making while interacting with humans. Regression tree and fuzzy logic methodologies have been investigated for the above task. This paper presents the results of applying these two methods and discusses their comparative merits. Three human participant experiments were designed and trials were conducted with five participants. The experimental results demonstrate the feasibility of the proposed anxiety-recognition methods. To the best of our knowledge, our work is the first consolidated effort at fusing multiple physiological indices for robust, real-time detection of anxiety using pattern recognition methods such as fuzzy logic and regression trees.