| Presentation | 2017-11-10 Prediction method of tumble using machine learning of footsteps. Takehiro Mori, Hiroyuki Nishi, Yoshimasa Kimura, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | Along with the progress of super aged society, lonely elderly people are increasing. In addition, accidents in homes and out of the home of elderly people also increased, and accidents caused by falls accounted for about 22% of in-home accidents. In order to improve this, the authors have predicted a fall using the walking sound of elderly people and are investigating a walking sound identification method to prevent falls beforehand. In this research, we focus on the fact that elderly people tend to walk on the shank as a preliminary stage of falling. By using the fact that the acoustic features of walking sounds are different greatly between normal walking and sleeping, by using a neural network to distinguish them, when a surrogate is detected, a warning or a caution alert is given to the elderly It prevents falls by giving. On the premise of a flooring floor, when wearing slippers, we recorded ordinary walking sounds by bare feet and walking sounds of abrasion for about 5 minutes, and a discrimination experiment was conducted with half of them as learning and the rest as evaluation. As a result of the evaluation experiment using the dimension number of the intermediate layer as a parameter, 95.5% of the highest performance was obtained when the number of dimensions is 20. As a result, we were able to detect the survival state with high accuracy, give a warning to the elderly, and obtain a prospect of preventing falls. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Elderly people / Observation / footsteps / Neural network / tumble |
| Paper # | ISEC2017-61,SITE2017-43,LOIS2017-38 |
| Date of Issue | 2017-11-02 (ISEC, SITE, LOIS) |
| Conference Information | |
| Committee | LOIS / ISEC / SITE |
|---|---|
| Conference Date | 2017/11/9(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | Hiroyuki Nishi(Sojo Univ.) / Kazuto Ogawa(NHK) / Hitoshi Okada(NII) |
| Vice Chair | Tomohiro Yamada(NTT) / Atsushi Fujioka(Kanagawa Univ.) / Shiho Moriai(NICT) / Tetsuya Morizumi(Kanagawa Univ.) / Masaru Ogawa(Kobe Gakuin Univ.) |
| Secretary | Tomohiro Yamada(Nagasaki Univ.) / Atsushi Fujioka(NTT) / Shiho Moriai(Tohoku Univ.) / Tetsuya Morizumi(Tokai Univ.) / Masaru Ogawa(Gifu Shotoku Gakuen Univ.) |
| Assistant | Motoi Okamoto(Research Organization of Information and Systems) / Keita Emura(NICT) / Yuichi Komano(TOSHIBA) / Yuuji Suga(IIJ) / Akiyoshi Kabeya(Chiba Univ.) / Hisanori Kato(KDDI) |
| Paper Information | |
| Registration To | Technical Committee on Life Intelligence and Office Information Systems / Technical Committee on Information Security / Technical Committee on Social Implications of Technology and Information Ethics |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Prediction method of tumble using machine learning of footsteps. |
| Sub Title (in English) | |
| Keyword(1) | Elderly people |
| Keyword(2) | Observation |
| Keyword(3) | footsteps |
| Keyword(4) | Neural network |
| Keyword(5) | tumble |
| 1st Author's Name | Takehiro Mori |
| 1st Author's Affiliation | Sojo University(Sojo Univ.) |
| 2nd Author's Name | Hiroyuki Nishi |
| 2nd Author's Affiliation | Sojo University(Sojo Univ.) |
| 3rd Author's Name | Yoshimasa Kimura |
| 3rd Author's Affiliation | Sojo University(Sojo Univ.) |
| Date | 2017-11-10 |
| Paper # | ISEC2017-61,SITE2017-43,LOIS2017-38 |
| Volume (vol) | vol.117 |
| Number (no) | ISEC-285,SITE-286,LOIS-287 |
| Page | pp.pp.79-82(ISEC), pp.79-82(SITE), pp.79-82(LOIS), |
| #Pages | 4 |
| Date of Issue | 2017-11-02 (ISEC, SITE, LOIS) |