Presentation | 2020-05-29 Characterizing Quality of In-home Physical Activities Using Bone-based Human Sensing Sinan Chen, Sachio Saiki, Masahide Nakamura, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | In recent years, with the conversion to home care in the aging society, it is significant for extending healthy life span on how to continue physical activities at home. Especially, to comprehend signs of disuse atrophy and dementia of the elderly, the physical activity amount at home is an important index.The goal of this study is to measure the physical activity amount of residents in a simple, non-invasive manner, and to connect it with the qualitative evaluation of physical activities at home. More specifically, we capture the in-home real-time video with a fixed-point camera, extract the bone data of a resident using the bone sensing techniques, and accumulate solely the coordinate of feature points as the time-series data. We examine a method to characterize the physical activity amount, which calculates the amount of change from the coordinate data on time series, such as the posture and position of residents. It does not require the wearable devices in the proposed method. The non-invasive physical activity sensing can be achieved since the original in-home video is not accumulated. We first used the machine learning model PoseNet for the pose estimation, which detected the coordinates of the body boundary area and 17 feature points in real-time. We then analyzed the accumulated data, to checked the possible with the quantitation of the basic bone movements, posture changes, and position movements. Such as sitting, standing, and walking.Furthermore, we also discussed the association with standard activity scales (ex., METs), and looked ahead to the specific application of the proposed method. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Home care / Physical activities / Video / PoseNet / Bone sensing |
Paper # | SC2020-1 |
Date of Issue | 2020-05-22 (SC) |
Conference Information | |
Committee | SC |
---|---|
Conference Date | 2020/5/29(1days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | Online/Univ of Aizu |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | AI Application for Service Computing Environment and Other Issues |
Chair | Masahide Nakamura(Kobe Univ.) |
Vice Chair | Shinji Kikuchi(NIMS) / Yoji Yamato(NTT) |
Secretary | Shinji Kikuchi(Tokyo Univ. of Tech.) / Yoji Yamato(Fujitsu Lab.) |
Assistant |
Paper Information | |
Registration To | Technical Committee on Service Computing |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Characterizing Quality of In-home Physical Activities Using Bone-based Human Sensing |
Sub Title (in English) | |
Keyword(1) | Home care |
Keyword(2) | Physical activities |
Keyword(3) | Video |
Keyword(4) | PoseNet |
Keyword(5) | Bone sensing |
1st Author's Name | Sinan Chen |
1st Author's Affiliation | Kobe University(Kobe Univ.) |
2nd Author's Name | Sachio Saiki |
2nd Author's Affiliation | Kobe University(Kobe Univ.) |
3rd Author's Name | Masahide Nakamura |
3rd Author's Affiliation | Kobe University(Kobe Univ.) |
Date | 2020-05-29 |
Paper # | SC2020-1 |
Volume (vol) | vol.120 |
Number (no) | SC-49 |
Page | pp.pp.1-6(SC), |
#Pages | 6 |
Date of Issue | 2020-05-22 (SC) |