Title: VoXtream: Full-Stream Text-to-Speech with Extremely Low Latency Show Chinese title

Title: VoXtream：具有极低延迟的全流文本到语音

Abstract: We present VoXtream, a fully autoregressive, zero-shot streaming text-to-speech (TTS) system for real-time use that begins speaking from the first word. VoXtream directly maps incoming phonemes to audio tokens using a monotonic alignment scheme and a dynamic look-ahead that does not delay onset. Built around an incremental phoneme transformer, a temporal transformer predicting semantic and duration tokens, and a depth transformer producing acoustic tokens, VoXtream achieves, to our knowledge, the lowest initial delay among publicly available streaming TTS: 102 ms on GPU. Despite being trained on a mid-scale 9k-hour corpus, it matches or surpasses larger baselines on several metrics, while delivering competitive quality in both output- and full-streaming settings. Demo and code are available at https://herimor.github.io/voxtream. Show Chinese abstract

Abstract: 我们提出VoXtream，这是一个完全自回归的、零样本流式文本到语音（TTS）系统，适用于实时使用，可以从第一个词开始说话。 VoXtream使用单调对齐方案和动态前瞻直接将传入的音素映射到音频标记，而不会延迟开始。 围绕一个增量音素变压器、一个预测语义和持续时间标记的时间变压器以及一个生成声学标记的深度变压器构建，据我们所知，VoXtream在公开可用的流式TTS中实现了最低的初始延迟：在GPU上为102毫秒。 尽管在中等规模的9000小时语料库上进行训练，它在多个指标上达到了或超过了更大的基线，在输出流和全流设置中都提供了具有竞争力的质量。 演示和代码可在https://herimor.github.io/voxtream获取。