A Public Domain Dataset for Human Activity Recognition in Free-Living Conditions

• DOI: 10.1109/smartworld-uic-atc-scalcom-iop-sci.2019.00071
• Publication Date: 2020-04-10T03:03:00Z
• Authors: Federico Cruciani; Chen Sun; Shuai Zhang; Chris Nugent; Chunping Li; Shaoxu Song; Cheng Cheng; Ian Cleland; Paul Mccullagh
• Keywords: Activity Recognition; Extrasensory perception

In Human Activity Recognition (HAR), supervised Machine Learning methods are predominantly used, making availability of datasets a major issue for research in the field. particular, majority available collected under controlled conditions. Consequently, models trained similar circumstances, generally exhibit significant decrease recognition accuracy when they moved to final deployment wild, within unconstrained settings. This paper presents new dataset HAR, free-living and 10 subjects were recruited purpose data collection. Data was recorded over 6 week period using smartphone app, wristband activity monitor. During first last observation, participants also wore an ActivPAL" logger. The have been partially self-labeled by participants, means mobile app provided can be used evaluate HAR algorithm real-world Together with description dataset, this work some preliminary results, obtained cross-validating model on publicly Extrasensory testing its performance our newly dataset. Results highlighted high cross-subject variability subjects, balanced varying between 53.33% 90.01%, average 71.73%.