Objective estimation of perceived speech quality .II. Evaluation of the measuring normalizing block technique

Abstract

Part I of this paper describes a new approach to the objective estimation of perceived speech quality. This new approach uses a simple but effective perceptual transformation and a distance measure that consists of a hierarchy of measuring normalizing blocks. Each measuring normalizing block integrates two perceptually transformed signals over some time or frequency interval to determine the average difference across that interval. This difference is then normalized out of one signal, and is further processed to generate one or more measurements. In Part II, the resulting estimates of perceived speech quality are correlated with the results of nine subjective listening tests. Together, these tests include 219 4-kHz bandwidth speech codecs, transmission systems, and reference conditions, with bit rates ranging from 2.4 to 64 kb/s. When compared with six other estimators, significant improvements are seen in many cases, particularly at lower bit rates, and when bit errors or frame erasures are present. These hierarchical structures of measuring normalizing blocks, or other structures of measuring normalizing blocks may also address open issues in perceived audio quality estimation, layered speech or audio coding, automatic speech or speaker recognition, audio signal enhancement, and other areas.

Voice quality evaluation in wireless packet communication systems: a tutorial and performance results for RoHC

“…These results are confirmed by Wu and Pols [7], who estimated a correlation of 0.926 for the LPC cepstral distance measure with the mean opinion score. This correlation performance has been further verified for waveform preserving codecs and for the MNRU, which is one of the most common reference conditions for subjective and objective voice quality assessments, as part of the recent study by Voran [2]. Because of its widely verified correlation performance to subjective hearing tests, we use the results of the fundamental study [6] to predict the mean opinion score from the cepstral distance, as detailed shortly.…”

“…Various complex metrics have been developed and refined over the last decade. These include the Bark spectral distance, the measuring normalizing blocks (MNB) technique [2], and the PESQ measure [3], which was recently standardized by ITU-T as recommendation P.862. echo, and peak clipping: [7] Elementary objective voice quality metrics rely on low-complexity signal processing techniques to predict the subjective voice quality.…”

An Analysis of the Impact of Playout Delay Adjustments introduced by VoIP Jitter Buffers on Listening Speech Quality

“…Theyidentify the audible distortions based on the perceptual domain representation of twosignals incorporating human auditory models. Several intrusive models have been developed overrecent years, likeP erceptual Speech Quality Measure (PSQM) [18], Measuring Normalizing System (MNB) [ 19,20], Perceptual Analysis Measurement System (PAMS) [ 21], Perceptual Evaluation of Speech Quality (PESQ) [22,23] and Perceptual Objective Listening Quality Assessment (POLQA) [ 24,25]. Among the models mentioned above, PSQM, PESQand most recently,POLQA have been standardised by the ITU-T as Recommendations P. 861 [26], P. 862 [27] and P. 863 [28] respectively.M oreover, MNB is described in Appendix II of ITU-T Rec.…”

Hybrid Signal-and-Link-Parametric Speech Quality Measurement for VoIP Communications

“…1(a)] are "comparison-based" and depend on some form of distance metric between the input (clean) and output (degraded) speech signals to estimate subjective quality. Representative algorithms include perceptual speech quality measure (PSQM) [22], measuring normalizing block (MNB) [23], [24], and statistical data mining quality assessment [25]. ITU-T Recommendation P.862, also known as perceptual evaluation of speech quality (PESQ), represents the current state-of-art double-ended algorithm for traditional telephony applications [26].…”