SOURCE-LINKED INTELLIGENCE

StreamAlign: Streaming Text-Aligned Speech Tokenization

arXiv · AI, language, vision and robotics · article · Sep 9, 2026 · UTC

Text-aligned speech tokenization methods have emerged to better align speech tokens with LLM token spaces, enabling more effective utilization of pretrained LLMs. However, they rely on offline automatic speech recognition (ASR), leading to two key limitations: (i) the need for complete utterances before tokenization, precluding real-time streaming, and (ii) vocabulary mismatch between ASR and LLMs, which reduces acoustic granularity from the subword to the word level. We introduce StreamAlign, a text-aligned speech tokenization framework that enables streaming tokenization for real-time speech

Read original source ↗ Open in workspace

recordType

paper

region

Global