More for Less: Non-Intrusive Speech Quality Assessment with Limited Annotations

TL;DR

This paper introduces two multi-task learning models that leverage large unlabelled datasets to improve non-intrusive speech quality assessment with limited annotated data, outperforming existing baselines.

Contribution

It proposes novel multi-task models using deep clustering and degradation classification to enhance speech quality prediction with scarce annotations.

Findings

Deep clustering-based model outperforms degradation classifier model.

Models outperform baseline methods on TCD-VoIP dataset.

Multi-task learning with unlabelled data improves assessment accuracy.

Abstract

Non-intrusive speech quality assessment is a crucial operation in multimedia applications. The scarcity of annotated data and the lack of a reference signal represent some of the main challenges for designing efficient quality assessment metrics. In this paper, we propose two multi-task models to tackle the problems above. In the first model, we first learn a feature representation with a degradation classifier on a large dataset. Then we perform MOS prediction and degradation classification simultaneously on a small dataset annotated with MOS. In the second approach, the initial stage consists of learning features with a deep clustering-based unsupervised feature representation on the large dataset. Next, we perform MOS prediction and cluster label classification simultaneously on a small dataset. The results show that the deep clustering-based model outperforms the degradation classifier-based model and the 3 baselines (autoencoder features, P.563, and SRMRnorm) on TCD-VoIP. This paper indicates that multi-task learning combined with feature representations from unlabelled data is a promising approach to deal with the lack of large MOS annotated datasets.