Performance Evaluation of an Ambient Noise Clustering Method for Objective Speech Intelligibility Estimate

Abstract

We investigated the clustering of non-stationary noise used in the subjective and objective assessment of speech intelligibility. The feature vector used in the clustering comprises 15-dimensional features used typically in MIR and clustering was done using the x-means method. We then conducted tests to validate the clustering results using the Japanese Diagnostic Rhyme Test. As a result, with the JEIDA-NOISE database, the noise can be classified into three clusters, and significant difference between speech intelligibility of each cluster was seen. Finally, we tested the objective speech intelligibility assessment for each cluster using the fwSNRseg and the logistic function. As a result, the performance of objective assessment improved by about 0.01 compared to the case without clustering.