Real Time Sound Information Analysis for Grasping Disastrous Indoor Situations

• OKUBO Hikaru

  Procurement Agency, Kyushu Defense Bureau

• NAKASHIMA Tadayoshi

  Faculty of Engineering, Hokkaido University

• OKADA Shigeyuki

  Center for Natural Hazards Research, Hokkaido University

Bibliographic Information

Other Title

• 地震時室内状況把握のためのリアルタイム音情報解析
• ジシンジ シツナイ ジョウキョウ ハアク ノ タメ ノ リアルタイムオン ジョウホウ カイセキ

Abstract

In order to prevent injuries in earthquake, person at home is nessesary to take apporopriate actions, but it is difficult to make an accurate snap judgment during an earthquake. As one of the solution strategy, we have been developing an evacuation guidance system using five senses information. In the step of this study, we tried to propose a monitoring system capable of judging some events from indoor sound information at normal and earthquake. A recognition system function built in the system identifies the occurrence and location of events including living activities and indoor damage caused by earthquake through the analysis by deep learning.

Journal

• Journal of Social Safety Science

  Journal of Social Safety Science 37 (0), 1-9, 2020-11-01

  Institute of Social Safety Science

Keywords

• <i>Sound Information</i>
• <i>Deep Learning</i>
• <i>Living Activities</i>
• <i>Realtime Monitoring</i>
• <i>Event Location</i>
• <i>Event</i>

Details

• CRID: 1390568456356201984
• NII Article ID: 130007976951
• NII Book ID: AA11542656
• DOI: 10.11314/jisss.37.1
• ISSN: 21879842
• ISSN: 13452088
• NDL BIB ID: 031391786
• Web Site: http://id.ndl.go.jp/bib/031391786
• Web Site: https://ndlsearch.ndl.go.jp/books/R000000004-I031391786
• Text Lang: ja
• Data Source:
  • JaLC
  • NDL
  • CiNii Articles
  • KAKEN
• Abstract License Flag: Disallowed