Variation and Synthetic Speech

TL;DR

This paper presents a neural network-based speech synthesis system that models linguistic and speaker-specific variation to produce more natural synthetic speech.

Contribution

It introduces a neural network architecture with a postlexical module trained on phonetic data to capture variation and adapt to individual speakers.

Findings

The system effectively models linguistic variation.

It improves naturalness in synthetic speech.

The architecture allows speaker-specific adaptation.

Abstract

We describe the approach to linguistic variation taken by the Motorola speech synthesizer. A pan-dialectal pronunciation dictionary is described, which serves as the training data for a neural network based letter-to-sound converter. Subsequent to dictionary retrieval or letter-to-sound generation, pronunciations are submitted a neural network based postlexical module. The postlexical module has been trained on aligned dictionary pronunciations and hand-labeled narrow phonetic transcriptions. This architecture permits the learning of individual postlexical variation, and can be retrained for each speaker whose voice is being modeled for synthesis. Learning variation in this way can result in greater naturalness for the synthetic speech that is produced by the system.