Title: In-vehicle application of common speech intelligibility metrics

Authors: Nikolina Samardzic; Colin Novak

Addresses: Department of Mechanical, Automotive and Materials Engineering, University of Windsor, 401 Sunset Ave., Windsor, Ontario N9B 3P4, Canada. ' Department of Mechanical, Automotive and Materials Engineering, University of Windsor, 401 Sunset Ave., Windsor, Ontario N9B 3P4, Canada

Abstract: The purpose of this investigation is to quantify the loss of speech intelligibility related to communication between passengers inside a vehicle at different vehicle operating conditions and road surfaces using common objective speech intelligibility metrics. The goal was to identify the most appropriate metric, if a single one exists, for use in automotive applications. The objective metrics include the articulation index (AI), the speech intelligibility index (SII) and the speech transmission index (STI). The SII method, utilising user-defined, measured, speech signal was found to be the best out of the three metrics for quantifying in-vehicle speech intelligibility. Since the effect of reverberation on the loss of speech intelligibility was negligible, this method resulted in a close correlation with the more measurement-intensive STI method. This potentially provides a reduction in measurement effort while preserving the accuracy of the results.

Keywords: in-vehicle communication; articulation index; speech intelligibility index; speech transmission index; passenger communication; speech signals; vehicle noise.

DOI: 10.1504/IJVNV.2011.043193

International Journal of Vehicle Noise and Vibration, 2011 Vol.7 No.4, pp.328 - 346

Published online: 14 Oct 2014 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article