@inproceedings{dongxian-etal-2026-mstree,
title = "{MST}ree: Multi-Scale Evidence Tree for Training-Free Long-Form Video Question Answering",
author = "Dongxian, Chen and
Liu, Mingshuai and
Peng, Yijie and
Liu, Hengrui and
Li, Xin",
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.1949/",
pages = "44321--44336",
ISBN = "979-8-89176-459-0",
abstract = "Long-form video question answering hinges on locating brief, sparse evidence buried in minutes of dense action. Flattening the entire video into a single caption stream forces the LLM to sift through thousands of tokens, where decisive cues are easily drowned out by surrounding context. We present MSTree, a training-free framework that organizes captions into a Multi-Scale Evidence Tree. The tree is built by a coordinated dual-modal segmentation method that combines visual scene transitions and textual coherence under temporal contiguity, producing segments that are both visually and semantically consistent. Inference follows a mine-then-reason workflow: the LLM first scans coarse segment summaries to locate the most relevant event, then reasons over its fine-grained captions. When the first-round prediction falls below a confidence threshold, Contrastive Targeted Reflection (CTR) activates. The model articulates a discriminative gap that specifies what evidence is still missing to separate the top candidate answers, and an answer-anchoring constraint requires any newly retrieved evidence to both support the revised prediction and contradict the original. On EgoSchema, NExT-QA, and Video-MME, MSTree reaches 68.4{\%}, 77.5{\%}, and 55.7{\%} accuracy, outperforming all training-free baselines."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="dongxian-etal-2026-mstree">
<titleInfo>
<title>MSTree: Multi-Scale Evidence Tree for Training-Free Long-Form Video Question Answering</title>
</titleInfo>
<name type="personal">
<namePart type="given">Chen</namePart>
<namePart type="family">Dongxian</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mingshuai</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yijie</namePart>
<namePart type="family">Peng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hengrui</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xin</namePart>
<namePart type="family">Li</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>Long-form video question answering hinges on locating brief, sparse evidence buried in minutes of dense action. Flattening the entire video into a single caption stream forces the LLM to sift through thousands of tokens, where decisive cues are easily drowned out by surrounding context. We present MSTree, a training-free framework that organizes captions into a Multi-Scale Evidence Tree. The tree is built by a coordinated dual-modal segmentation method that combines visual scene transitions and textual coherence under temporal contiguity, producing segments that are both visually and semantically consistent. Inference follows a mine-then-reason workflow: the LLM first scans coarse segment summaries to locate the most relevant event, then reasons over its fine-grained captions. When the first-round prediction falls below a confidence threshold, Contrastive Targeted Reflection (CTR) activates. The model articulates a discriminative gap that specifies what evidence is still missing to separate the top candidate answers, and an answer-anchoring constraint requires any newly retrieved evidence to both support the revised prediction and contradict the original. On EgoSchema, NExT-QA, and Video-MME, MSTree reaches 68.4%, 77.5%, and 55.7% accuracy, outperforming all training-free baselines.</abstract>
<identifier type="citekey">dongxian-etal-2026-mstree</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.1949/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>44321</start>
<end>44336</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T MSTree: Multi-Scale Evidence Tree for Training-Free Long-Form Video Question Answering
%A Dongxian, Chen
%A Liu, Mingshuai
%A Peng, Yijie
%A Liu, Hengrui
%A Li, Xin
%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 dongxian-etal-2026-mstree
%X Long-form video question answering hinges on locating brief, sparse evidence buried in minutes of dense action. Flattening the entire video into a single caption stream forces the LLM to sift through thousands of tokens, where decisive cues are easily drowned out by surrounding context. We present MSTree, a training-free framework that organizes captions into a Multi-Scale Evidence Tree. The tree is built by a coordinated dual-modal segmentation method that combines visual scene transitions and textual coherence under temporal contiguity, producing segments that are both visually and semantically consistent. Inference follows a mine-then-reason workflow: the LLM first scans coarse segment summaries to locate the most relevant event, then reasons over its fine-grained captions. When the first-round prediction falls below a confidence threshold, Contrastive Targeted Reflection (CTR) activates. The model articulates a discriminative gap that specifies what evidence is still missing to separate the top candidate answers, and an answer-anchoring constraint requires any newly retrieved evidence to both support the revised prediction and contradict the original. On EgoSchema, NExT-QA, and Video-MME, MSTree reaches 68.4%, 77.5%, and 55.7% accuracy, outperforming all training-free baselines.
%U https://aclanthology.org/2026.emnlp-main.1949/
%P 44321-44336
Markdown (Informal)
[MSTree: Multi-Scale Evidence Tree for Training-Free Long-Form Video Question Answering](https://aclanthology.org/2026.emnlp-main.1949/) (Dongxian et al., EMNLP 2026)
ACL