Fuzzy logic based activity life cycle tracking and recognition

Abstract

Human activity has a life cycle that could be tracked to provide crucial information that increases activity recognition performance. Life cycle tracking also enables powerful and more versatile programming models over the sentience of human activities. This paper introduces a new activity recognition approach that is based on fuzzy logic and activity life cycle tracking. We conduct preliminary performance validation and show how activity life cycle tracking offers higher accuracy when compared to the traditional approach of no-life cycle activity recognition. © 2013 Springer-Verlag Berlin Heidelberg.