Automated detection of impulsive movements in HCI

This paper introduces an algorithm for automatically measuring impulsivity. This can be used as a major expressive movement feature in the development of systems for realtime analysis of emotion expression from human full-body movement, a research area which has received increased attention in the affective computing community. In particular, our algorithm is developed in the framework of the EUH2020- ICT Project DANCE aiming at investigating techniques for sensory substitution in blind people, in order to enable perception of and participation in non-verbal, artistic whole-body experiences. The algorithm was tested by applying it to a reference archive of short dance performances. The archive includes a collection of both impulsive and fluid movements. Results show that our algorithm can reliably distinguish impulsive vs. sudden performances.