SSR-Speech: Towards Stable, Safe and Robust Zero-shot Text-based Speech Editing and Synthesis

TL;DR

SSR-Speech is a novel neural codec model for stable, safe, and robust zero-shot speech editing and synthesis, incorporating watermarking and leveraging original speech segments for high-quality reconstruction.

Contribution

It introduces SSR-Speech, a Transformer-based model with classifier-free guidance and watermarking, achieving state-of-the-art results in speech editing and synthesis tasks.

Findings

State-of-the-art performance in RealEdit speech editing

Superior quality in LibriTTS text-to-speech synthesis

Robustness to background noise and multi-span editing

Abstract

In this paper, we introduce SSR-Speech, a neural codec autoregressive model designed for stable, safe, and robust zero-shot textbased speech editing and text-to-speech synthesis. SSR-Speech is built on a Transformer decoder and incorporates classifier-free guidance to enhance the stability of the generation process. A watermark Encodec is proposed to embed frame-level watermarks into the edited regions of the speech so that which parts were edited can be detected. In addition, the waveform reconstruction leverages the original unedited speech segments, providing superior recovery compared to the Encodec model. Our approach achieves state-of-the-art performance in the RealEdit speech editing task and the LibriTTS text-to-speech task, surpassing previous methods. Furthermore, SSR-Speech excels in multi-span speech editing and also demonstrates remarkable robustness to background sounds. The source code and demos are released.