Non-intrusive Speech Quality Assessment Using Neural Networks

By Anderson R. Avila, Hannes Gamper, Chandan K. A. Reddy, Ross Cutler, Ivan J. Tashev, Johannes Gehrke. 2019

114 citations recorded by the source corpus.

Estimating the perceived quality of an audio signal is critical for many multimedia and audio processing systems. Providers strive to offer optimal and reliable services in order to increase the user quality of experience (QoE). In this work, we present an investigation of the applicability of neural networks for non-intrusive audio quality assessment. We propose three neural network-based approaches for mean opinion score (MOS) estimation. We compare our results to three instrumental measures: the perceptual evaluation of speech quality (PESQ), the ITU-T Recommendation P.563, and the speech-to-reverberation energy ratio. Our evaluation uses a speech dataset contaminated with convo-lutive and additive noise, labeled using a crowd-based QoE evaluation, evaluated with Pearson correlation with MOS labels, and mean-squared-error of the estimated MOS. Our proposed approaches outperform the aforementioned instrumental measures, with a fully connected deep neural network using Mel-frequency features providing the best correlation (0.87) and the lowest mean squared error (0.15).

Subfields: Cognitive Neuroscience, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cognitive Neuroscience

In Signal Processing

Citations in the corpus

An Open source Implementation of ITU-T Recommendation P.808 with Validation (2020, background): “Therefore, they are not reliable replacements to listening opinion tests even though they are widely used by researchers due to their convenience [3].”
ViSQOL v3: An Open Source Production Ready Objective Speech and Audio Metric (2020, methodology): “Since then, deep neural network models have emerged and been applied to speech quality models [14], [15].”
Icassp 2022 Deep Noise Suppression Challenge (2022, background): “The objective measures of speech quality, such as PESQ [21], SDR, and POLQA [22] do not correlate well with subjective speech quality [23].”
Dnsmos: A Non-Intrusive Perceptual Objective Speech Quality Metric to Evaluate Noise Suppressors (2020, methodology): “Pearson Correlation Coefficient (PCC) or MSE between the predictions of the developed objective metric and the ground truth human ratings is commonly used to measure the accuracy of the model [7, 8].”
Utilizing Self-supervised Representations for MOS Prediction (2021, methodology): “Several machine-learning-based models [4, 5, 6, 7, 8] were thus proposed for automatic speech quality assessment.”
ICASSP 2021 Deep Noise Suppression Challenge (2020, background): “However, these metrics are shown to not correlate well with subjective speech quality in the presence of background noise [29].”
Intrusive and Non-Intrusive Perceptual Speech Quality Assessment Using a Convolutional Neural Network (2019, methodology): “The proposed CNN achieves the highest performance in terms of both RMSE and the correlation with the ground-truth MOS, slightly outperforming the previously proposed MLP-based architecture [12].”
ICASSP 2023 Deep Noise Suppression Challenge (2023, background): “Though, this was not a requirement for this challenge.”
ICASSP 2022 Acoustic Echo Cancellation Challenge (2022, background): “PESQ has also been shown to not have a high correlation to subjective speech quality in the presence of background noise [8].”
Dnsmos P.835: A Non-Intrusive Perceptual Objective Speech Quality Metric to Evaluate Noise Suppressors (2021, background): “Recently, Deep Neural Networks (DNNs) based approaches have been proposed to estimate the speech quality scores [13, 14, 12, 15, 16, 17, 18, 19, 20].”