Socially assistive robots can engage in assistive human-robot interactions (HRI) by providing rehabilitation of cognitive, social, and physical abilities after a stroke, accident or diagnosis of a social, developmental or cognitive disorder. However, there are a number of research issues that need to be addressed in order to design such robots. In this paper, we address one main challenge in the development of intelligent socially assistive robots: A robot’s ability to identify human non-verbal communication during assistive interactions. In this paper, we present a unique non-contact automated sensory-based approach for identification and categorization of human upper body language in determining how accessible a person is to a robot during natural real-time HRI. This classification will allow a robot to effectively determine its own reactive task-driven behavior during assistive interactions. The types of interactions envisioned include providing reminders, health monitoring, and social and cognitive therapies. Preliminary experiments show the potential of integrating the proposed body language recognition and classification technique into socially assistive robotic systems partaking in HRI scenarios.

This content is only available via PDF.
You do not currently have access to this content.