Zero-Shot Streaming Text to Speech Synthesis with Transducer and Auto-Regressive Modeling

TL;DR

This paper introduces SMLLE, a streaming text-to-speech framework that generates high-quality speech in real-time without relying on lookahead, reducing latency while maintaining naturalness.

Contribution

The paper presents a novel streaming TTS system combining a Transducer and autoregressive model with a Delete < Bos > Mechanism for minimal delay and high-quality speech synthesis.

Findings

Outperforms existing streaming TTS methods in quality

Achieves comparable performance to sentence-level TTS

Operates with minimal delay in real-time synthesis

Abstract

Zero-shot streaming text-to-speech is an important research topic in human-computer interaction. Existing methods primarily use a lookahead mechanism, relying on future text to achieve natural streaming speech synthesis, which introduces high processing latency. To address this issue, we propose SMLLE, a streaming framework for generating high-quality speech frame-by-frame. SMLLE employs a Transducer to convert text into semantic tokens in real time while simultaneously obtaining duration alignment information. The combined outputs are then fed into a fully autoregressive (AR) streaming model to reconstruct mel-spectrograms. To further stabilize the generation process, we design a Delete < Bos > Mechanism that allows the AR model to access future text introducing as minimal delay as possible. Experimental results suggest that the SMLLE outperforms current streaming TTS methods and achieves comparable performance over sentence-level TTS systems. Samples are available on shy-98.github.io/SMLLE_demo_page/.