@inproceedings{hyun-bin-etal-2026-spatio,
title = "Spatio-Temporal Audio Language Modeling for Dynamic Sound Sources",
author = "Hyun-Bin, Oh and
Shimada, Kazuki and
Takida, Yuhta and
Sung-Bin, Kim and
Uesaka, Toshimitsu and
Shibuya, Takashi and
Lee, Kyeongyoon and
Oh, Tae-Hyun and
Mitsufuji, Yuki",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.2043/",
pages = "46568--46590",
ISBN = "979-8-89176-459-0",
abstract = "Sound events have semantic identities, locations, and trajectories, yet existing audio-language models typically encode clips as global event content. Conversely, sound event localization models track source directions over time but offer limited semantic coverage for language reasoning. To bridge this gap, we introduce ST-AudioQA, a spatio-temporal audio QA dataset and benchmark built from first-order ambisonic (FOA) renderings of static and moving sound sources. Each scene provides source identity, activity, direction, distance, and motion metadata, enabling dense trajectory supervision and QA about what is sounding, where it is, how it moves, and how sources relate. We further introduce ST-Audio Encoder, which jointly learns event semantics and time-resolved source trajectories, and ST-AudioLM, which exposes these audio tokens to an LLM for spatio-temporal audio QA. Experiments show that this representation improves the balance between semantic preservation and localization and yields stronger reasoning than static spatial and localization-oriented baselines. Beyond the controlled benchmark, ST-AudioLM shows zero-shot transfer to real recordings, with further gains after light adaptation. Project page: https://st-audiolm.github.io/."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="hyun-bin-etal-2026-spatio">
<titleInfo>
<title>Spatio-Temporal Audio Language Modeling for Dynamic Sound Sources</title>
</titleInfo>
<name type="personal">
<namePart type="given">Oh</namePart>
<namePart type="family">Hyun-Bin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kazuki</namePart>
<namePart type="family">Shimada</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yuhta</namePart>
<namePart type="family">Takida</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kim</namePart>
<namePart type="family">Sung-Bin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Toshimitsu</namePart>
<namePart type="family">Uesaka</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Takashi</namePart>
<namePart type="family">Shibuya</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kyeongyoon</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tae-Hyun</namePart>
<namePart type="family">Oh</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yuki</namePart>
<namePart type="family">Mitsufuji</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>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</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">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>Sound events have semantic identities, locations, and trajectories, yet existing audio-language models typically encode clips as global event content. Conversely, sound event localization models track source directions over time but offer limited semantic coverage for language reasoning. To bridge this gap, we introduce ST-AudioQA, a spatio-temporal audio QA dataset and benchmark built from first-order ambisonic (FOA) renderings of static and moving sound sources. Each scene provides source identity, activity, direction, distance, and motion metadata, enabling dense trajectory supervision and QA about what is sounding, where it is, how it moves, and how sources relate. We further introduce ST-Audio Encoder, which jointly learns event semantics and time-resolved source trajectories, and ST-AudioLM, which exposes these audio tokens to an LLM for spatio-temporal audio QA. Experiments show that this representation improves the balance between semantic preservation and localization and yields stronger reasoning than static spatial and localization-oriented baselines. Beyond the controlled benchmark, ST-AudioLM shows zero-shot transfer to real recordings, with further gains after light adaptation. Project page: https://st-audiolm.github.io/.</abstract>
<identifier type="citekey">hyun-bin-etal-2026-spatio</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.2043/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>46568</start>
<end>46590</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Spatio-Temporal Audio Language Modeling for Dynamic Sound Sources
%A Hyun-Bin, Oh
%A Shimada, Kazuki
%A Takida, Yuhta
%A Sung-Bin, Kim
%A Uesaka, Toshimitsu
%A Shibuya, Takashi
%A Lee, Kyeongyoon
%A Oh, Tae-Hyun
%A Mitsufuji, Yuki
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F hyun-bin-etal-2026-spatio
%X Sound events have semantic identities, locations, and trajectories, yet existing audio-language models typically encode clips as global event content. Conversely, sound event localization models track source directions over time but offer limited semantic coverage for language reasoning. To bridge this gap, we introduce ST-AudioQA, a spatio-temporal audio QA dataset and benchmark built from first-order ambisonic (FOA) renderings of static and moving sound sources. Each scene provides source identity, activity, direction, distance, and motion metadata, enabling dense trajectory supervision and QA about what is sounding, where it is, how it moves, and how sources relate. We further introduce ST-Audio Encoder, which jointly learns event semantics and time-resolved source trajectories, and ST-AudioLM, which exposes these audio tokens to an LLM for spatio-temporal audio QA. Experiments show that this representation improves the balance between semantic preservation and localization and yields stronger reasoning than static spatial and localization-oriented baselines. Beyond the controlled benchmark, ST-AudioLM shows zero-shot transfer to real recordings, with further gains after light adaptation. Project page: https://st-audiolm.github.io/.
%U https://aclanthology.org/2026.emnlp-main.2043/
%P 46568-46590
Markdown (Informal)
[Spatio-Temporal Audio Language Modeling for Dynamic Sound Sources](https://aclanthology.org/2026.emnlp-main.2043/) (Hyun-Bin et al., EMNLP 2026)
ACL
- Oh Hyun-Bin, Kazuki Shimada, Yuhta Takida, Kim Sung-Bin, Toshimitsu Uesaka, Takashi Shibuya, Kyeongyoon Lee, Tae-Hyun Oh, and Yuki Mitsufuji. 2026. Spatio-Temporal Audio Language Modeling for Dynamic Sound Sources. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 46568–46590, Budapest, Hungary. Association for Computational Linguistics.