@inproceedings{zhang-etal-2022-movieun,
title = "{M}ovie{UN}: A Dataset for Movie Understanding and Narrating",
author = "Zhang, Qi and
Yue, Zihao and
Hu, Anwen and
Wang, Ziheng and
Jin, Qin",
editor = "Goldberg, Yoav and
Kozareva, Zornitsa and
Zhang, Yue",
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2022",
month = dec,
year = "2022",
address = "Abu Dhabi, United Arab Emirates",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2022.findings-emnlp.135",
doi = "10.18653/v1/2022.findings-emnlp.135",
pages = "1873--1885",
abstract = "Automatic movie narration generation and narration grounding are very important to provide a true movie experience for the blind and visually impaired. To tell the movie story well, it is necessary to mention plot-related details (such as character names) and keep the narrations in a plot coherent. Taking these two points into consideration, we construct a Chinese large-scale video benchmark from 101 movies for Movie Understanding and Narrating (MovieUN) to support the Movie Clip Narrating (MCN) task and Temporal Narration Grounding (TNG) task. We split movies in MovieUN into movie clips according to plots, and pair them with corresponding narrations provided by the movie narrators. Ultimately, the TNG task involves 3,253 long video clips totaling 179 hours. The MCN task contains 33,060 video clips totaling 105 hours. We benchmark state-of-the-art video captioning models and temporal grounding models in MCN and TNG tasks, respectively. Furthermore, to accurately comprehend plots of different characters, we propose methods to incorporate portraits of actors as external knowledge in both tasks. The experiment results demonstrate the effectiveness of our proposed methods. The dataset and codes are released at \url{https://github.com/yuezih/MovieUN}.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhang-etal-2022-movieun">
<titleInfo>
<title>MovieUN: A Dataset for Movie Understanding and Narrating</title>
</titleInfo>
<name type="personal">
<namePart type="given">Qi</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zihao</namePart>
<namePart type="family">Yue</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Anwen</namePart>
<namePart type="family">Hu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ziheng</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Qin</namePart>
<namePart type="family">Jin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2022-12</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2022</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yoav</namePart>
<namePart type="family">Goldberg</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zornitsa</namePart>
<namePart type="family">Kozareva</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yue</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Abu Dhabi, United Arab Emirates</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Automatic movie narration generation and narration grounding are very important to provide a true movie experience for the blind and visually impaired. To tell the movie story well, it is necessary to mention plot-related details (such as character names) and keep the narrations in a plot coherent. Taking these two points into consideration, we construct a Chinese large-scale video benchmark from 101 movies for Movie Understanding and Narrating (MovieUN) to support the Movie Clip Narrating (MCN) task and Temporal Narration Grounding (TNG) task. We split movies in MovieUN into movie clips according to plots, and pair them with corresponding narrations provided by the movie narrators. Ultimately, the TNG task involves 3,253 long video clips totaling 179 hours. The MCN task contains 33,060 video clips totaling 105 hours. We benchmark state-of-the-art video captioning models and temporal grounding models in MCN and TNG tasks, respectively. Furthermore, to accurately comprehend plots of different characters, we propose methods to incorporate portraits of actors as external knowledge in both tasks. The experiment results demonstrate the effectiveness of our proposed methods. The dataset and codes are released at https://github.com/yuezih/MovieUN.</abstract>
<identifier type="citekey">zhang-etal-2022-movieun</identifier>
<identifier type="doi">10.18653/v1/2022.findings-emnlp.135</identifier>
<location>
<url>https://aclanthology.org/2022.findings-emnlp.135</url>
</location>
<part>
<date>2022-12</date>
<extent unit="page">
<start>1873</start>
<end>1885</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T MovieUN: A Dataset for Movie Understanding and Narrating
%A Zhang, Qi
%A Yue, Zihao
%A Hu, Anwen
%A Wang, Ziheng
%A Jin, Qin
%Y Goldberg, Yoav
%Y Kozareva, Zornitsa
%Y Zhang, Yue
%S Findings of the Association for Computational Linguistics: EMNLP 2022
%D 2022
%8 December
%I Association for Computational Linguistics
%C Abu Dhabi, United Arab Emirates
%F zhang-etal-2022-movieun
%X Automatic movie narration generation and narration grounding are very important to provide a true movie experience for the blind and visually impaired. To tell the movie story well, it is necessary to mention plot-related details (such as character names) and keep the narrations in a plot coherent. Taking these two points into consideration, we construct a Chinese large-scale video benchmark from 101 movies for Movie Understanding and Narrating (MovieUN) to support the Movie Clip Narrating (MCN) task and Temporal Narration Grounding (TNG) task. We split movies in MovieUN into movie clips according to plots, and pair them with corresponding narrations provided by the movie narrators. Ultimately, the TNG task involves 3,253 long video clips totaling 179 hours. The MCN task contains 33,060 video clips totaling 105 hours. We benchmark state-of-the-art video captioning models and temporal grounding models in MCN and TNG tasks, respectively. Furthermore, to accurately comprehend plots of different characters, we propose methods to incorporate portraits of actors as external knowledge in both tasks. The experiment results demonstrate the effectiveness of our proposed methods. The dataset and codes are released at https://github.com/yuezih/MovieUN.
%R 10.18653/v1/2022.findings-emnlp.135
%U https://aclanthology.org/2022.findings-emnlp.135
%U https://doi.org/10.18653/v1/2022.findings-emnlp.135
%P 1873-1885
Markdown (Informal)
[MovieUN: A Dataset for Movie Understanding and Narrating](https://aclanthology.org/2022.findings-emnlp.135) (Zhang et al., Findings 2022)
ACL
- Qi Zhang, Zihao Yue, Anwen Hu, Ziheng Wang, and Qin Jin. 2022. MovieUN: A Dataset for Movie Understanding and Narrating. In Findings of the Association for Computational Linguistics: EMNLP 2022, pages 1873–1885, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics.