Quality and Usability

Speech Quality Engineering

Test methods for wideband speech quality engineering


Motivation & Project Description

  • Development of test methods for subjective evaluation of speech quality, e.g. multidimensional analyses, call-quality assessment, etc.
  • Development of enhanced parametric and signal-based models for the prediction of wideband speech quality:
    • Extension of the E-Model towards wideband speech, based on perceptual dimensions
    • Development of a signal-based model for estimating overall quality as well as perceptually relevant dimensions
  • Development of an overall engineering approach for quantifying and monitoring wideband speech quality in an end-to-end scenario (together with INET)

Expected Outcome:

  • Wideband E-model for planning and monitoring wideband speech networks
  • Evaluation and improvement of a signal-based model developed by T-Labs, France Télécom R&D, and Kiel University
  • ITU-T contributions and recommendations (ITU-T SG 12, Q.8/12 and Q.9/12)
  • Scientific publications
  • Project outcome directly applicable to other projects (e.g., HiQ Voice, Mobisense goes Video)

 

Time Frame: 01/2009-12/2010

T-labs Team Members: Nicolas Côté, Blazej Lewcio, Sebastian Möller,
Alexander Raake, Marcel Wältermann

Students: Katrin Bieker

Funding by: Deutsche Telekom Laboratories