english | português | español

Human Activity Recognition using On-body Sensing

Ugulino, W.; Velloso, E.; Milidiu, R.; Fuks, H. Human Activity Recognition using On-body Sensing. Proceedings of III Symposium of the Brazilian Institute for Web Science Research (WebScience) , Volume 1. Rio de Janeiro: PUC-Rio, 2012.
Cited by 4 (Google Scholar)

Temas: Sistemas Embarcados, Internet das coisas, Hardware, Eletrônica, Prova de Conceito, Prototipação


Abstract: Human Activity Recognition (HAR) based on wearable sensors is gaining increasing attention by the pervasive computing research community, especially for development of context-aware systems. This paper presents our approach for HAR based on wearable accelerometers and supervised learning algorithms. We present the HARwear device, a wearable for HAR. Using HARwear we collected data and we developed a classifier for postures and movements. The classifier is able to identify 2 postures and 3 types of movements with an accuracy of 99.4%. Based on the lessons from experimental evaluation, we propose an improved version, the HARwear version 2, which is also presented on this paper.