@inproceedings{kim-etal-2026-streamalign,
title = "{S}tream{A}lign: Streaming Text-Aligned Speech Tokenization",
author = "Kim, Kang-wook and
Park, Jinyoung and
Kim, Jinsoo and
Lee, Sehun and
Woo, Tony and
Kim, Gunhee",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.2270/",
pages = "48596--48615",
ISBN = "979-8-89176-462-0",
abstract = "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{--}text joint modeling. StreamAlign performs online speech{--}text alignment by combining character-level RNN-Transducer alignment with word-level ASR guidance, mitigating ASR{--}LLM vocabulary mismatch while preserving recognition accuracy. A proactive word boundary classifier anticipates word completion at chunk boundaries, reducing tokenization latency from 560ms to 270 ms. On LibriSpeech, StreamAlign achieves the lowest WER and highest UTMOS among evaluated tokenizers. Furthermore, StreamAlign-SLM, a spoken language model trained on StreamAlign units, demonstrates the strongest overall performance in terms of both acoustic and semantic consistency on the SALMon and spoken StoryCloze benchmarks."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="kim-etal-2026-streamalign">
<titleInfo>
<title>StreamAlign: Streaming Text-Aligned Speech Tokenization</title>
</titleInfo>
<name type="personal">
<namePart type="given">Kang-wook</namePart>
<namePart type="family">Kim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jinyoung</namePart>
<namePart type="family">Park</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jinsoo</namePart>
<namePart type="family">Kim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sehun</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tony</namePart>
<namePart type="family">Woo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Gunhee</namePart>
<namePart type="family">Kim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>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–text joint modeling. StreamAlign performs online speech–text alignment by combining character-level RNN-Transducer alignment with word-level ASR guidance, mitigating ASR–LLM vocabulary mismatch while preserving recognition accuracy. A proactive word boundary classifier anticipates word completion at chunk boundaries, reducing tokenization latency from 560ms to 270 ms. On LibriSpeech, StreamAlign achieves the lowest WER and highest UTMOS among evaluated tokenizers. Furthermore, StreamAlign-SLM, a spoken language model trained on StreamAlign units, demonstrates the strongest overall performance in terms of both acoustic and semantic consistency on the SALMon and spoken StoryCloze benchmarks.</abstract>
<identifier type="citekey">kim-etal-2026-streamalign</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.2270/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>48596</start>
<end>48615</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T StreamAlign: Streaming Text-Aligned Speech Tokenization
%A Kim, Kang-wook
%A Park, Jinyoung
%A Kim, Jinsoo
%A Lee, Sehun
%A Woo, Tony
%A Kim, Gunhee
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F kim-etal-2026-streamalign
%X 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–text joint modeling. StreamAlign performs online speech–text alignment by combining character-level RNN-Transducer alignment with word-level ASR guidance, mitigating ASR–LLM vocabulary mismatch while preserving recognition accuracy. A proactive word boundary classifier anticipates word completion at chunk boundaries, reducing tokenization latency from 560ms to 270 ms. On LibriSpeech, StreamAlign achieves the lowest WER and highest UTMOS among evaluated tokenizers. Furthermore, StreamAlign-SLM, a spoken language model trained on StreamAlign units, demonstrates the strongest overall performance in terms of both acoustic and semantic consistency on the SALMon and spoken StoryCloze benchmarks.
%U https://aclanthology.org/2026.findings-emnlp.2270/
%P 48596-48615
Markdown (Informal)
[StreamAlign: Streaming Text-Aligned Speech Tokenization](https://aclanthology.org/2026.findings-emnlp.2270/) (Kim et al., Findings 2026)
ACL
- Kang-wook Kim, Jinyoung Park, Jinsoo Kim, Sehun Lee, Tony Woo, and Gunhee Kim. 2026. StreamAlign: Streaming Text-Aligned Speech Tokenization. In Findings of the Association for Computational Linguistics: EMNLP 2026, pages 48596–48615, Budapest, Hungary. Association for Computational Linguistics.