| Presentation | 2023-08-04 Recognizing Human-Centered Contexts for In-Home Elderly Monitoring Using Vision-Based Edge AI Sinan Chen, Masahide Nakamura, Kiyoshi Yasuda, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | As the global population ages, including Japan, there is a significant trend toward transitioning from facility-based care to home-based care due to the shortage of caregiving facilities and personnel. While elderly individuals may find it easier to adapt to living at home, the daily caregiving and monitoring of their activities significantly burden family caregivers. Previous studies have proposed methods such as ``mind'' sensing for elderly people based on voice dialogue systems and quality assessment of elderly individuals' in-home activities using skeleton sensing technology. However, recognizing changes in elderly individuals' facial expressions, body postures, and behaviors (called context) is essential for monitoring elderly individuals at home and has yet to be fully realized. Therefore, this study examines a situation-centric context recognition method revolving around nonverbal features. Our key idea is to integrate multiple pre-trained models based on images in an edge environment, extract human-centric features, and characterize them as context. As an approach, we integrate locally executable image recognition technologies based on multiple pre-trained models to perform human-centric context recognition from live images. Following this method, we can expect to create a home monitoring system that is easy to implement in ordinary households using a general-purpose computer and a stationary USB camera. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Elderly monitoring / Image recognition / Edge AI / Human-centered / Context recognition |
| Paper # | LOIS2023-6 |
| Date of Issue | 2023-07-28 (LOIS) |
| Conference Information | |
| Committee | LOIS / IPSJ-DC |
|---|---|
| Conference Date | 2023/8/4(1days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Kyoto Tachibana University, Keisei-Kan, 1-G106 |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | Hiroyuki Toda(NTT) / Michiko Oba(Hitachi) |
| Vice Chair | Manabu Motegi(Takushoku Univ.) |
| Secretary | Manabu Motegi(Nagasaki Univ.) / (NTT) |
| Assistant | Makoto Takita(Univer. of Hyogo) |
| Paper Information | |
| Registration To | Technical Committee on Life Intelligence and Office Information Systems / Special Interest Group on Document Communication |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Recognizing Human-Centered Contexts for In-Home Elderly Monitoring Using Vision-Based Edge AI |
| Sub Title (in English) | |
| Keyword(1) | Elderly monitoring |
| Keyword(2) | Image recognition |
| Keyword(3) | Edge AI |
| Keyword(4) | Human-centered |
| Keyword(5) | Context recognition |
| 1st Author's Name | Sinan Chen |
| 1st Author's Affiliation | Kobe University(Kobe Univ.) |
| 2nd Author's Name | Masahide Nakamura |
| 2nd Author's Affiliation | Kobe University(Kobe Univ.) |
| 3rd Author's Name | Kiyoshi Yasuda |
| 3rd Author's Affiliation | Kobe University(Kobe Univ.) |
| Date | 2023-08-04 |
| Paper # | LOIS2023-6 |
| Volume (vol) | vol.123 |
| Number (no) | LOIS-150 |
| Page | pp.pp.18-22(LOIS), |
| #Pages | 5 |
| Date of Issue | 2023-07-28 (LOIS) |