无需训练数据，推理时生成连贯长语音
MagpieTTS-LF: Inference-Time Long-Form Speech Generation Without Training on Long-Form data
- 推理阶段引入软注意力先验与状态化算法
- 长文本生成在可懂度、语调连贯性上显著提升
- 适合需要高质量长语音的场景如有声书
神经文本转语音系统在短句生成上表现优异，但在长语音生成中存在语调漂移、说话人不一致和句子边界伪影问题。现有方法或压缩序列、增加上下文长度，或简单拼接独立合成的片段。本文提出一种推理时方法 MagpieTTS-LF，使 MagpieTTS 在不重新训练模型的前提下生成连贯长语音。方法包含三项关键创新：(1) 软注意力先验，引导单调对齐同时保留过去和未来上下文；(2) 状态化推理算法，跨句子块保持上下文，确保语调连续性；(3) 历史感知文本编码，利用历史文本进行话语级语调规划。在长文本上的实验表明，相比其他基线，该方法在长距离可懂度、语调连贯性、说话人一致性及边界自然度方面均有显著提升。
原文摘要 · Abstract (English)
Neural Text-to-Speech (TTS) systems achieve remarkable quality on short utterances but long-form speech generation shows prosodic drift, speaker inconsistencies and sentence boundary artifacts. Existing approaches either compress sequences, increase context length or naively concatenate independently synthesized chunks. We present an inference-time approach called MagpieTTS-LF that enables MagpieTTS to produce coherent long-form speech without model retraining. Our method introduces three key innovations: (1) soft attention priors to guide monotonic alignment while preserving past and future context; (2) a stateful inference algorithm that maintains context across sentence chunks, ensuring prosodic continuity; (3) history-aware text encoding that uses past text for discourse-level prosodic planning. Experiments on long texts show significant improvements in long-range intelligibility, prosodic coherence, speaker consistency, and boundary naturalness compared to other baselines.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。