Prosodic Prominence and Boundaries in Sequence-to-Sequence Speech Synthesis

TL;DR

This paper investigates whether augmenting text input with automatically extracted prosodic labels improves the naturalness and accuracy of prosody in sequence-to-sequence speech synthesis, addressing limitations in reproducing local prosodic features.

Contribution

The study introduces a wavelet-based method to extract prosodic labels and demonstrates their effectiveness in enhancing prosodic accuracy in speech synthesis.

Findings

Prosodic labels improve f0 contour accuracy.

Energy contour fidelity is enhanced with labels.

Objective metrics show significant improvement.

Abstract

Recent advances in deep learning methods have elevated synthetic speech quality to human level, and the field is now moving towards addressing prosodic variation in synthetic speech.Despite successes in this effort, the state-of-the-art systems fall short of faithfully reproducing local prosodic events that give rise to, e.g., word-level emphasis and phrasal structure. This type of prosodic variation often reflects long-distance semantic relationships that are not accessible for end-to-end systems with a single sentence as their synthesis domain. One of the possible solutions might be conditioning the synthesized speech by explicit prosodic labels, potentially generated using longer portions of text. In this work we evaluate whether augmenting the textual input with such prosodic labels capturing word-level prominence and phrasal boundary strength can result in more accurate realization of sentence prosody. We use an automatic wavelet-based technique to extract such labels from speech material, and use them as an input to a tacotron-like synthesis system alongside textual information. The results of objective evaluation of synthesized speech show that using the prosodic labels significantly improves the output in terms of faithfulness of f0 and energy contours, in comparison with state-of-the-art implementations.