@inproceedings{cagliero-etal-2025-detecting,
title = "Detecting and Mitigating Challenges in Zero-Shot Video Summarization with Video {LLM}s",
author = "Cagliero, Luca and
Vaiani, Lorenzo and
Pastor, Eliana and
Koudounas, Alkis and
Baralis, Elena and
Mazzia, Vittorio and
Pollastrini, Sandro and
Gueudre, Thomas and
Giollo, Manuel and
Amberti, Daniele and
Wu, Yue",
editor = "Che, Wanxiang and
Nabende, Joyce and
Shutova, Ekaterina and
Pilehvar, Mohammad Taher",
booktitle = "Findings of the Association for Computational Linguistics: ACL 2025",
month = jul,
year = "2025",
address = "Vienna, Austria",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-acl.16/",
doi = "10.18653/v1/2025.findings-acl.16",
pages = "286--301",
ISBN = "979-8-89176-256-5",
abstract = "Video summarization aims to generate a condensed textual version of an original video. Summaries may consist of either plain text or a shortlist of salient events, possibly including temporal or spatial references. Video Large Language Models (VLLMs) exhibit impressive zero-shot capabilities in video analysis. However, their performance varies significantly according to the LLM prompt, the characteristics of the video, and the properties of the training data and LLM architecture.In this work, we thoroughly evaluate the zero-shot summarization performance of four state-of-the-art open-source VLLMs specifically designed to address spatial and temporal reasoning. In light of the detected summarization issues, we propose different cost-effective mitigation strategies, based on Chain-of-Thought prompting, that involve the injection of knowledge extracted by external, lightweight models. To perform the VLLM evaluation, we design a new video summarization benchmark consisting of 100 videos with varying characteristics in terms of domain, duration, and spatio-temporal properties. Videos are manually annotated by three independent human experts with plain text, event-based, and spatio-temporal summaries. The experimental evaluation shows that VLLMs significantly benefit from prompting a list of recognized actions, whereas injecting automatically recognized objects and scene changes respectively improve spatially contextualized and event-based summaries in specific cases."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="cagliero-etal-2025-detecting">
<titleInfo>
<title>Detecting and Mitigating Challenges in Zero-Shot Video Summarization with Video LLMs</title>
</titleInfo>
<name type="personal">
<namePart type="given">Luca</namePart>
<namePart type="family">Cagliero</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lorenzo</namePart>
<namePart type="family">Vaiani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Eliana</namePart>
<namePart type="family">Pastor</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alkis</namePart>
<namePart type="family">Koudounas</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Elena</namePart>
<namePart type="family">Baralis</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Vittorio</namePart>
<namePart type="family">Mazzia</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sandro</namePart>
<namePart type="family">Pollastrini</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Thomas</namePart>
<namePart type="family">Gueudre</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Manuel</namePart>
<namePart type="family">Giollo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Daniele</namePart>
<namePart type="family">Amberti</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yue</namePart>
<namePart type="family">Wu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: ACL 2025</title>
</titleInfo>
<name type="personal">
<namePart type="given">Wanxiang</namePart>
<namePart type="family">Che</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Joyce</namePart>
<namePart type="family">Nabende</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ekaterina</namePart>
<namePart type="family">Shutova</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohammad</namePart>
<namePart type="given">Taher</namePart>
<namePart type="family">Pilehvar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Vienna, Austria</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-256-5</identifier>
</relatedItem>
<abstract>Video summarization aims to generate a condensed textual version of an original video. Summaries may consist of either plain text or a shortlist of salient events, possibly including temporal or spatial references. Video Large Language Models (VLLMs) exhibit impressive zero-shot capabilities in video analysis. However, their performance varies significantly according to the LLM prompt, the characteristics of the video, and the properties of the training data and LLM architecture.In this work, we thoroughly evaluate the zero-shot summarization performance of four state-of-the-art open-source VLLMs specifically designed to address spatial and temporal reasoning. In light of the detected summarization issues, we propose different cost-effective mitigation strategies, based on Chain-of-Thought prompting, that involve the injection of knowledge extracted by external, lightweight models. To perform the VLLM evaluation, we design a new video summarization benchmark consisting of 100 videos with varying characteristics in terms of domain, duration, and spatio-temporal properties. Videos are manually annotated by three independent human experts with plain text, event-based, and spatio-temporal summaries. The experimental evaluation shows that VLLMs significantly benefit from prompting a list of recognized actions, whereas injecting automatically recognized objects and scene changes respectively improve spatially contextualized and event-based summaries in specific cases.</abstract>
<identifier type="citekey">cagliero-etal-2025-detecting</identifier>
<identifier type="doi">10.18653/v1/2025.findings-acl.16</identifier>
<location>
<url>https://aclanthology.org/2025.findings-acl.16/</url>
</location>
<part>
<date>2025-07</date>
<extent unit="page">
<start>286</start>
<end>301</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Detecting and Mitigating Challenges in Zero-Shot Video Summarization with Video LLMs
%A Cagliero, Luca
%A Vaiani, Lorenzo
%A Pastor, Eliana
%A Koudounas, Alkis
%A Baralis, Elena
%A Mazzia, Vittorio
%A Pollastrini, Sandro
%A Gueudre, Thomas
%A Giollo, Manuel
%A Amberti, Daniele
%A Wu, Yue
%Y Che, Wanxiang
%Y Nabende, Joyce
%Y Shutova, Ekaterina
%Y Pilehvar, Mohammad Taher
%S Findings of the Association for Computational Linguistics: ACL 2025
%D 2025
%8 July
%I Association for Computational Linguistics
%C Vienna, Austria
%@ 979-8-89176-256-5
%F cagliero-etal-2025-detecting
%X Video summarization aims to generate a condensed textual version of an original video. Summaries may consist of either plain text or a shortlist of salient events, possibly including temporal or spatial references. Video Large Language Models (VLLMs) exhibit impressive zero-shot capabilities in video analysis. However, their performance varies significantly according to the LLM prompt, the characteristics of the video, and the properties of the training data and LLM architecture.In this work, we thoroughly evaluate the zero-shot summarization performance of four state-of-the-art open-source VLLMs specifically designed to address spatial and temporal reasoning. In light of the detected summarization issues, we propose different cost-effective mitigation strategies, based on Chain-of-Thought prompting, that involve the injection of knowledge extracted by external, lightweight models. To perform the VLLM evaluation, we design a new video summarization benchmark consisting of 100 videos with varying characteristics in terms of domain, duration, and spatio-temporal properties. Videos are manually annotated by three independent human experts with plain text, event-based, and spatio-temporal summaries. The experimental evaluation shows that VLLMs significantly benefit from prompting a list of recognized actions, whereas injecting automatically recognized objects and scene changes respectively improve spatially contextualized and event-based summaries in specific cases.
%R 10.18653/v1/2025.findings-acl.16
%U https://aclanthology.org/2025.findings-acl.16/
%U https://doi.org/10.18653/v1/2025.findings-acl.16
%P 286-301
Markdown (Informal)
[Detecting and Mitigating Challenges in Zero-Shot Video Summarization with Video LLMs](https://aclanthology.org/2025.findings-acl.16/) (Cagliero et al., Findings 2025)
ACL
- Luca Cagliero, Lorenzo Vaiani, Eliana Pastor, Alkis Koudounas, Elena Baralis, Vittorio Mazzia, Sandro Pollastrini, Thomas Gueudre, Manuel Giollo, Daniele Amberti, and Yue Wu. 2025. Detecting and Mitigating Challenges in Zero-Shot Video Summarization with Video LLMs. In Findings of the Association for Computational Linguistics: ACL 2025, pages 286–301, Vienna, Austria. Association for Computational Linguistics.