Presentation 2003/5/22
Recognition of Sound of Moving Vehicles Using HMM Composition
Yoshitaka HIRAMATSU, Jien KATO, Toyohide WATANABE,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) With a view of estimating traffic density, this paper proposes a hidden Markov model that recognizes the sound of moving cars on a single lane as an alternative means to image processing, in case vehicle detection through images is not available (such as at night). This model is supposed applicable to recognizing the sound mixed by three or less neighboring moving vehicles. It is generated using HMM composition method: firstly, an HMM with four states is produced for recognizing one moving car, then three states of the HMM are chosen and their transition probabilities are composed to generate the model which is able to recognize the mixed sound from moving cars up to three. Some experiments applying this model to practical traffic data have been performed to verify the model effectiveness.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) HMM / vehicle sound / vehicle detection
Paper # PRMU2003-7
Date of Issue

Conference Information
Committee PRMU
Conference Date 2003/5/22(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Pattern Recognition and Media Understanding (PRMU)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Recognition of Sound of Moving Vehicles Using HMM Composition
Sub Title (in English)
Keyword(1) HMM
Keyword(2) vehicle sound
Keyword(3) vehicle detection
1st Author's Name Yoshitaka HIRAMATSU
1st Author's Affiliation Graduate School of Engineering, Nagoya University()
2nd Author's Name Jien KATO
2nd Author's Affiliation Graduate School of Information Science, Nagoya University
3rd Author's Name Toyohide WATANABE
3rd Author's Affiliation Graduate School of Information Science, Nagoya University
Date 2003/5/22
Paper # PRMU2003-7
Volume (vol) vol.103
Number (no) 95
Page pp.pp.-
#Pages 6
Date of Issue