Activity Recognition for Ambient Assisted Living: Potential and Challenges

Abstract:
Reliable Activity Recognition is a key requirement for Ambient Assisted Living. It spans application areas from health state monitoring to Smart Home functionality. Activity recognition enables AAL systems to provide automatic, proactive support. In order to achieve this, these systems need to be able to recognize and differentiate between a set of complex, possibly spatially and temporally dispersed activities of humans in their living environments. In this contribution, we provide a survey of the core objectives of activity recognition, as well as the fundamental algorithmic methods and their capabilities. In addition, we outline solutions for two current research issues: multi-sensor activity recognition and effective synthesis of activity models.