Foot position as indicator of spatial interest at public displays

Abstract

Motivated by and grounded in observations of foot patterns in a human-human dialogue, this study explores expressions of spatial interest through feet at public displays. We conducted an observation and recorded user foot orientation and position in a public information display environment leading to data about 84 interaction sessions. Our observations show that characteristic foot patterns can be matched with two user intentions: (A) Users who seek access to specific information, and (B) users who don't seek specific information. With the goal to detect intention through foot patterns, we classified characteristic foot patterns with a SVM pattern recognition algorithm, which resulted in a detection accuracy of 84.4%. This work can be valuable for researchers designing context-aware public displays.