Presentation 2010-06-25
A Study on Change Point Detection of Activities from Time Series Data of Wrist-worn Accelerometers
Toshinori MIYOSHI, Ken KAWAMOTO, Hiroyuki KURIYAMA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Accelerometers offer a practical method to monitor human activities continuously. We are trying to develop a system of supporting life management by exploiting sensor data measured for long-term periods. For this purpose, we need to recognize high-level activities such as deskworks, meals, commutings, which are important for discovering daily routines or lifestyles. These high-level activities are composed of various low-level activities and their context, therefore, we need to analyze sequences of time series data of low level activities for recognition of high-level activities. As a preprocessing operation of recognizing high-level activities, we propose a change point detection method of human activities to split out time series data of one low/high level activity. Our proposed method detects change points of activities by analyzing changes of statistical distributions of the posture of an arm.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) accelerometer / time series data / change point detection / activity recognition / wrist-worn
Paper # PRMU2010-53,HIP2010-42
Date of Issue

Conference Information
Committee PRMU
Conference Date 2010/6/17(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Pattern Recognition and Media Understanding (PRMU)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) A Study on Change Point Detection of Activities from Time Series Data of Wrist-worn Accelerometers
Sub Title (in English)
Keyword(1) accelerometer
Keyword(2) time series data
Keyword(3) change point detection
Keyword(4) activity recognition
Keyword(5) wrist-worn
1st Author's Name Toshinori MIYOSHI
1st Author's Affiliation Hitachi, Ltd., Central Research Laboratory()
2nd Author's Name Ken KAWAMOTO
2nd Author's Affiliation Hitachi, Ltd., Advanced Research Laboratory
3rd Author's Name Hiroyuki KURIYAMA
3rd Author's Affiliation Hitachi, Ltd., Advanced Research Laboratory
Date 2010-06-25
Paper # PRMU2010-53,HIP2010-42
Volume (vol) vol.110
Number (no) 97
Page pp.pp.-
#Pages 6
Date of Issue