@inproceedings{wang-etal-2026-age,
title = "{AGE}-Sampler: Adaptive-granularity Graph-enhanced Event Sampling for Long-Video Understanding",
author = "Wang, Yanzi and
Dai, Tao and
Zha, Yaohua and
Xia, Shu-Tao",
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.2383/",
pages = "54141--54155",
ISBN = "979-8-89176-459-0",
abstract = "Informative keyframe selection is critical for applying large vision-language models (LVLMs) to long-form video understanding under limited frame budgets. However, existing methods heavily rely on query-frame similarity, which can yield coarse boundaries and miss latent contextual events that are answer-relevant but not explicitly mentioned in the query. To address this issue, we propose AGE-Sampler, a training-free Adaptive-granularity Graph-enhanced Event sampling framework for long-video understanding. AGE-Sampler jointly exploits low-level visual transition cues and query-relevance signals to construct query-adaptive events, thus preserving fine-grained structures in relevant regions while merging low-relevance regions into compact intervals. To recover answer-bearing contextual evidence, we further develop a fine-grained event graph and propagate importance from high-confidence query-relevant anchors via Personalized PageRank, followed by event-level budget allocation and representative keyframe selection. Extensive experiments with different LVLM backbones show that our AGE-Sampler consistently improves long-video understanding performance, achieving average absolute gains of +6.2{\%} on LongVideoBench, +8.8{\%} on MLVU, and +3.8{\%} on Video-MME, while remaining lightweight and plug-and-play."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="wang-etal-2026-age">
<titleInfo>
<title>AGE-Sampler: Adaptive-granularity Graph-enhanced Event Sampling for Long-Video Understanding</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yanzi</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tao</namePart>
<namePart type="family">Dai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yaohua</namePart>
<namePart type="family">Zha</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shu-Tao</namePart>
<namePart type="family">Xia</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>Informative keyframe selection is critical for applying large vision-language models (LVLMs) to long-form video understanding under limited frame budgets. However, existing methods heavily rely on query-frame similarity, which can yield coarse boundaries and miss latent contextual events that are answer-relevant but not explicitly mentioned in the query. To address this issue, we propose AGE-Sampler, a training-free Adaptive-granularity Graph-enhanced Event sampling framework for long-video understanding. AGE-Sampler jointly exploits low-level visual transition cues and query-relevance signals to construct query-adaptive events, thus preserving fine-grained structures in relevant regions while merging low-relevance regions into compact intervals. To recover answer-bearing contextual evidence, we further develop a fine-grained event graph and propagate importance from high-confidence query-relevant anchors via Personalized PageRank, followed by event-level budget allocation and representative keyframe selection. Extensive experiments with different LVLM backbones show that our AGE-Sampler consistently improves long-video understanding performance, achieving average absolute gains of +6.2% on LongVideoBench, +8.8% on MLVU, and +3.8% on Video-MME, while remaining lightweight and plug-and-play.</abstract>
<identifier type="citekey">wang-etal-2026-age</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.2383/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>54141</start>
<end>54155</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T AGE-Sampler: Adaptive-granularity Graph-enhanced Event Sampling for Long-Video Understanding
%A Wang, Yanzi
%A Dai, Tao
%A Zha, Yaohua
%A Xia, Shu-Tao
%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 wang-etal-2026-age
%X Informative keyframe selection is critical for applying large vision-language models (LVLMs) to long-form video understanding under limited frame budgets. However, existing methods heavily rely on query-frame similarity, which can yield coarse boundaries and miss latent contextual events that are answer-relevant but not explicitly mentioned in the query. To address this issue, we propose AGE-Sampler, a training-free Adaptive-granularity Graph-enhanced Event sampling framework for long-video understanding. AGE-Sampler jointly exploits low-level visual transition cues and query-relevance signals to construct query-adaptive events, thus preserving fine-grained structures in relevant regions while merging low-relevance regions into compact intervals. To recover answer-bearing contextual evidence, we further develop a fine-grained event graph and propagate importance from high-confidence query-relevant anchors via Personalized PageRank, followed by event-level budget allocation and representative keyframe selection. Extensive experiments with different LVLM backbones show that our AGE-Sampler consistently improves long-video understanding performance, achieving average absolute gains of +6.2% on LongVideoBench, +8.8% on MLVU, and +3.8% on Video-MME, while remaining lightweight and plug-and-play.
%U https://aclanthology.org/2026.emnlp-main.2383/
%P 54141-54155
Markdown (Informal)
[AGE-Sampler: Adaptive-granularity Graph-enhanced Event Sampling for Long-Video Understanding](https://aclanthology.org/2026.emnlp-main.2383/) (Wang et al., EMNLP 2026)
ACL