Deep Learning Based Early Recognition of Emergency Vehicles using On-Broad Microphones
車載マイクロフォンを用いた緊急車両の認識
- Delivery
- Available on this site
- Format
- Price
- Non-members (tax incl.)：￥1,100 Members (tax incl.)：￥880
- Publication code
- 20225004
- Paper/Info type
- Proceedings (Spring)
No.1-22
- Pages
- 1-6(Total 6 p)
- Date of publication
- May 2022
- Publisher
- JSAE
- Language
- Japanese
- Event
- 2022 JSAE Annual Congress (Spring)
Detailed Information
| Author(J) | 1) 高津 知里, 2) 米陀 佳祐, 3) 菅沼 直樹 |
|---|---|
| Author(E) | 1) Chisato Takatsu, 2) Keisuke Yoneda, 3) Naoki Suganuma |
| Affiliation(J) | 1) 金沢大学, 2) 金沢大学, 3) 金沢大学 |
| Affiliation(E) | 1) Kanazawa University, 2) Kanazawa University, 3) Kanazawa University |
| Abstract(J) | 自動運転車両を自律的に走行させるためには，緊急車両を車載センサにより認識し緊急車両の安全な走行の妨げとならない行動をとる必要があり，視界外に存在する緊急車両のサイレンを音で認識し，その接近を検知する必要がある．本研究では，事前にサイレンを使って作成した学習モデルを用いたサイレンの認識を目的とする． Translation |
| Abstract(E) | Early detection of emergency vehicles is very necessary for autonomous vehicles in order to enable smooth path planning without obstructing the emergency pass. This paper proposes a backbone framework to recognize emergency sirens using on-board microphones. As the recognition task is difficult in urban environments, we design a deep learning network to model spectrums and frequencies of various siren types from different directions and distances. The experiments illustrate promising results of the proposed model to be used in reality. |
