Deep Voice 2: Multi-Speaker Neural Text-to-Speech

TLDR

The paper introduces Deep Voice 2, a multi‑speaker neural TTS system that augments a single‑model architecture with trainable speaker embeddings to generate diverse voices. The system employs a single‑model architecture with low‑dimensional speaker embeddings, a post‑processing neural vocoder, and is trained on two multi‑speaker TTS datasets to synthesize high‑quality speech. The model outperforms Deep Voice 1 and Tacotron, achieving significant audio quality gains and enabling a single system to learn hundreds of distinct voices from under 30 minutes of data per speaker while preserving speaker identity.

Abstract

We introduce a technique for augmenting neural text-to-speech (TTS) with lowdimensional trainable speaker embeddings to generate different voices from a single model. As a starting point, we show improvements over the two state-ofthe-art approaches for single-speaker neural TTS: Deep Voice 1 and Tacotron. We introduce Deep Voice 2, which is based on a similar pipeline with Deep Voice 1, but constructed with higher performance building blocks and demonstrates a significant audio quality improvement over Deep Voice 1. We improve Tacotron by introducing a post-processing neural vocoder, and demonstrate a significant audio quality improvement. We then demonstrate our technique for multi-speaker speech synthesis for both Deep Voice 2 and Tacotron on two multi-speaker TTS datasets. We show that a single neural TTS system can learn hundreds of unique voices from less than half an hour of data per speaker, while achieving high audio quality synthesis and preserving the speaker identities almost perfectly.