SA-SSL-MOS: Self-supervised Learning MOS Prediction with Spectral Augmentation for Generalized Multi-Rate Speech Assessment
This work addresses the problem of limited labeled data for multi-rate speech quality assessment, which is incremental by enhancing existing self-supervised learning methods with high-frequency features.
The paper tackled the challenge of speech quality assessment for multi-rate speech with varying sampling frequencies by proposing a spectrogram-augmented self-supervised learning method that incorporates high-frequency features up to 48 kHz, showing that leveraging this overlooked information is crucial for accurate assessment and that a two-step training scheme improves generalization when multi-rate data is limited.
Designing a speech quality assessment (SQA) system for estimating mean-opinion-score (MOS) of multi-rate speech with varying sampling frequency (16-48 kHz) is a challenging task. The challenge arises due to the limited availability of a MOS-labeled training dataset comprising multi-rate speech samples. While self-supervised learning (SSL) models have been widely adopted in SQA to boost performance, a key limitation is that they are pretrained on 16 kHz speech and therefore discard high-frequency information present in higher sampling rates. To address this issue, we propose a spectrogram-augmented SSL method that incorporates high-frequency features (up to 48 kHz sampling rate) through a parallel-branch architecture. We further introduce a two-step training scheme: the model is first pre-trained on a large 48 kHz dataset and then fine-tuned on a smaller multi-rate dataset. Experimental results show that leveraging high-frequency information overlooked by SSL features is crucial for accurate multi-rate SQA, and that the proposed two-step training substantially improves generalization when multi-rate data is limited.