# Layer-wise Analysis for Quality of Multilingual Synthesized Speech

arXiv: https://arxiv.org/abs/2509.04830 · HTML: https://mlsanity.com/paper/2509.04830/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 5 Sep 2025.

## Built on by (most cited first, 2 of 2)

- [An Evaluation Framework for Text-to-Speech Voice Reconstruction](https://mlsanity.com/paper/2606.21343/) (2026)
- [CAL-MOS: Bridging Layers with Adapters for Robust MOS Prediction Across Speech Foundation Models](https://mlsanity.com/paper/2609.14956/) (2026)

## Abstract (arXiv)

While supervised quality predictors for synthesized speech have demonstrated strong correlations with human ratings, their requirement for in-domain labeled training data hinders their generalization ability to new domains. Unsupervised approaches based on pretrained self-supervised learning (SSL) based models and automatic speech recognition (ASR) models are a promising alternative; however, little is known about how these models encode information about speech quality. Towards the goal of better understanding how different aspects of speech quality are encoded in a multilingual setting, we present a layer-wise analysis of multilingual pretrained speech models based on reference modeling. We find that features extracted from early SSL layers show correlations with human ratings of synthesized speech, and later layers of ASR models can predict quality of non-neural systems as well as intelligibility. We also demonstrate the importance of using well-matched reference data.