Long-Context Speech Synthesis with Context-Aware Memory

TL;DR

This paper introduces a Context-Aware Memory (CAM) mechanism for long-context speech synthesis, improving coherence and naturalness in paragraph-level speech by dynamically integrating long-term memory and local context.

Contribution

The paper presents a novel CAM-based TTS model that effectively captures long-term context and local details, enhancing paragraph-level speech synthesis quality.

Findings

Outperforms baseline methods in prosody expressiveness

Improves coherence and style consistency in long speech

Reduces context inference cost

Abstract

In long-text speech synthesis, current approaches typically convert text to speech at the sentence-level and concatenate the results to form pseudo-paragraph-level speech. These methods overlook the contextual coherence of paragraphs, leading to reduced naturalness and inconsistencies in style and timbre across the long-form speech. To address these issues, we propose a Context-Aware Memory (CAM)-based long-context Text-to-Speech (TTS) model. The CAM block integrates and retrieves both long-term memory and local context details, enabling dynamic memory updates and transfers within long paragraphs to guide sentence-level speech synthesis. Furthermore, the prefix mask enhances the in-context learning ability by enabling bidirectional attention on prefix tokens while maintaining unidirectional generation. Experimental results demonstrate that the proposed method outperforms baseline and state-of-the-art long-context methods in terms of prosody expressiveness, coherence and context inference cost across paragraph-level speech.