提出可修订的生成框架，让非自回归语音合成更准确可控。
Mask, Sample, Revise: A Revisable CTMC Inference Stack for Guided Discrete Flow Matching Text-to-Speech
- 用三步推理栈实现低步数下的精准语音生成
- 在少步数下提升语音清晰度与稳定性
- 适合需要快速生成高质量语音的研究者
近期无对齐的非自回归（NAR）文本到语音（TTS）模型将语音合成建模为条件补全任务，无需显式持续时间预测器和外部对齐器。当语音以神经编解码器标记表示时，该问题变为离散形式，使离散流匹配（DFM）——一种用于离散生成的连续时间马尔可夫链（CTMC）框架——成为自然选择。然而，推理阶段的稳定低步数条件补全控制仍缺乏探索。本文提出‘掩码-采样-修订’（Mask, Sample, Revise），一个用于无对齐DFM-TTS的推理阶段CTMC堆栈。该堆栈结合无预测器引导以强化文本条件、与提示匹配的条件耦合以对齐概率路径，并引入SC-ReMask机制，通过约束调度的重新掩码，允许早期去掩码决策被修正。这些组件无需后期微调，且可在单一tau-leaping采样器中运行。受控消融实验表明，该堆栈在低数值函数评估（NFE）提示设置下提升了语音可懂度与鲁棒性，优于无引导与仅引导采样器，且使用显著更多步骤。
原文摘要 · Abstract (English)
Recent alignment-free non-autoregressive (NAR) text-to-speech (TTS) models formulate synthesis as a conditional infilling task, bypassing explicit duration predictors and external aligners. When speech is represented with neural codec tokens, the infilling problem becomes discrete, making Discrete Flow Matching (DFM), a Continuous-Time Markov Chain (CTMC) framework for discrete generation, a natural fit. However, inference-time control for stable low-step conditional infilling remains underexplored. We propose Mask, Sample, Revise, an inference-time CTMC stack for alignment-free DFM-TTS. The stack combines predictor-free guidance to strengthen text conditioning, prompt-matched conditional coupling to align the probability path with the acoustic prompt, and SC-ReMask, a schedule-constrained remasking mechanism that introduces token-to-mask transitions so early de-masking decisions can be revised. These components require no post-hoc fine-tuning and operate in a single tau-leaping sampler. Controlled ablations show that this stack improves intelligibility and robustness in the low-NFE prompted setting, outperforming unguided and guidance-only samplers with substantially more steps.
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。