Psychoacoustic Challenges Of Speech Enhancement On VoIP Platforms

TL;DR

This paper investigates the psychoacoustic effects of proprietary denoising algorithms on VoIP speech quality, using innovative econometric analysis and psychoacoustic metrics across popular platforms like Google Meets and Zoom.

Contribution

It introduces a novel application of Blinder-Oaxaca decomposition to analyze acoustic perturbations in VoIP, and provides comprehensive benchmarking of speech enhancement models in this context.

Findings

Proprietary denoising impacts speech quality and intelligibility.

Psychoacoustic metrics reveal perceptual distortions caused by VoIP processing.

Benchmarking shows variability in enhancement model performance.

Abstract

Within the ambit of VoIP (Voice over Internet Protocol) telecommunications, the complexities introduced by acoustic transformations merit rigorous analysis. This research, rooted in the exploration of proprietary sender-side denoising effects, meticulously evaluates platforms such as Google Meets and Zoom. The study draws upon the Deep Noise Suppression (DNS) 2020 dataset, ensuring a structured examination tailored to various denoising settings and receiver interfaces. A methodological novelty is introduced via Blinder-Oaxaca decomposition, traditionally an econometric tool, repurposed herein to analyze acoustic-phonetic perturbations within VoIP systems. To further ground the implications of these transformations, psychoacoustic metrics, specifically PESQ and STOI, were used to explain of perceptual quality and intelligibility. Cumulatively, the insights garnered underscore the intricate landscape of VoIP-influenced acoustic dynamics. In addition to the primary findings, a multitude of metrics are reported, extending the research purview. Moreover, out-of-domain benchmarking for both time and time-frequency domain speech enhancement models is included, thereby enhancing the depth and applicability of this inquiry.