Speech Transmission Index from running speech : a neural network approach

Li, FF ORCID: https://orcid.org/0000-0001-9053-963X and Cox, TJ ORCID: https://orcid.org/0000-0002-4075-7564 2003, 'Speech Transmission Index from running speech : a neural network approach' , The Journal of the Acoustical Society of America (JASA), 113 (4) , pp. 1999-2008.

PDF - Published Version
Download (189kB) | Preview
Access Information: Copyright (1999) Acoustical Society of America. This article may be downloaded for personal use only. Any other use requires prior permission of the author and the Acoustical Society of America.


Speech Transmission Index (STI) is an important objective parameter concerning speech intelligibility for sound transmission channels. It is normally measured with specific test signals to ensure high accuracy and good repeatability. Measurement with running speech was previously proposed, but accuracy is compromised and hence applications limited. A new approach that uses artificial neural networks to accurately extract the STI from received running speech is developed in this paper. Neural networks are trained on a large set of transmitted speech examples with prior knowledge of the transmission channels' STIs. The networks perform complicated nonlinear function mappings and spectral feature memorization to enable accurate objective parameter extraction from transmitted speech. Validations via simulations demonstrate the feasibility of this new method on a one-net-one-speech extract basis. In this case, accuracy is comparable with normal measurement methods. This provides an alternative to standard measurement techniques, and it is intended that the neural network method can facilitate occupied room acoustic measurements.

Item Type: Article
Themes: Subjects / Themes > Q Science > QC Physics > QC221-246 Acoustics - Sound
Subjects / Themes > Q Science > QC Physics
Subjects outside of the University Themes
Schools: Schools > School of the Built Environment
Schools > School of Computing, Science and Engineering > Salford Innovation Research Centre
Journal or Publication Title: The Journal of the Acoustical Society of America (JASA)
Publisher: Acoustical Society of America
Refereed: Yes
ISSN: 0001-4966
Depositing User: H Kenna
Date Deposited: 11 Sep 2007 14:03
Last Modified: 16 Feb 2022 07:46
URI: https://usir.salford.ac.uk/id/eprint/438

Actions (login required)

Edit record (repository staff only) Edit record (repository staff only)


Downloads per month over past year