Evaluation of instrumental measures for the prediction of musical noise in enhanced noisy speech
M. R. J. Gerrits · VBN Forskningsportal (Aalborg Universitet) · 2014
Abstract. Instrumental measures have been primarily used in order to predict speech quality or speech intelligibility. The aim of the present work is to evaluate the performance of a broad range of established instrumental measures in terms of their ability to predict the amount of musical noise present in enhanced noisy speech signals. The considered instrumental measures were evaluated using musical noise quantity scores obtained from a specially designed listening-test which was performed by normal-hearing listeners. Of all considered instrumental measures, a mean squared distortion measure, the PESQ measure, and the STOI measure yielded the highest correlations. These results confirm the ability of instrumental measures to predict the amount of musical noise, but further evaluation shows limitations to their applicability as musical noise predictors. However, subsequently the results suggest that simultaneous optimization for the amount of musical noise and speech quality or speech intelligibly is not possible using a single instrumental measure. 1