A Study on Self‐Organizing Map of Indoor Activity Sound for Detecting Abnormal Situation in Daily Life

• Published in: IEEJ transactions on electrical and electronic engineering Vol. 16; no. 3; pp. 488 - 490
• Main Authors: Tanaka, Motoshi, Takata, Naru
• Format: Journal Article
• Language: English
• Published: Hoboken, USA John Wiley & Sons, Inc 01-03-2021; Wiley Subscription Services, Inc
• Subjects: Broadband; broadband frequency analysis; daily activity sound; detection of abnormal situation; self‐organizing map; Sound
• ISSN: 1931-4973; 1931-4981
• DOI: 10.1002/tee.23321

In order to develop a detection system of abnormal situations (such as accidents) for a person living alone, SOM (self‐organizing map) of only daily activity sound recorded with a broadband microphone, and detection of abnormal situation are investigated. The SOM was divided into three regions containing 98.9, 1.0, and 0.1% of the number of learned vectors, and transition of the vectors of simulated abnormality sound was observed. The result indicates the feasibility of use of the SOM for detecting abnormal situations. © 2021 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.