用于视频问题解答的多对象事件图表示学习

Yanan Wang, Shuichiro Haruta, Donghuo Zeng, Julio Vizcarra, Mori Kurokawa
{"title":"用于视频问题解答的多对象事件图表示学习","authors":"Yanan Wang, Shuichiro Haruta, Donghuo Zeng, Julio Vizcarra, Mori Kurokawa","doi":"arxiv-2409.07747","DOIUrl":null,"url":null,"abstract":"Video question answering (VideoQA) is a task to predict the correct answer to\nquestions posed about a given video. The system must comprehend spatial and\ntemporal relationships among objects extracted from videos to perform causal\nand temporal reasoning. While prior works have focused on modeling individual\nobject movements using transformer-based methods, they falter when capturing\ncomplex scenarios involving multiple objects (e.g., \"a boy is throwing a ball\nin a hoop\"). We propose a contrastive language event graph representation\nlearning method called CLanG to address this limitation. Aiming to capture\nevent representations associated with multiple objects, our method employs a\nmulti-layer GNN-cluster module for adversarial graph representation learning,\nenabling contrastive learning between the question text and its relevant\nmulti-object event graph. Our method outperforms a strong baseline, achieving\nup to 2.2% higher accuracy on two challenging VideoQA datasets, NExT-QA and\nTGIF-QA-R. In particular, it is 2.8% better than baselines in handling causal\nand temporal questions, highlighting its strength in reasoning multiple\nobject-based events.","PeriodicalId":501030,"journal":{"name":"arXiv - CS - Computation and Language","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2024-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Multi-object event graph representation learning for Video Question Answering\",\"authors\":\"Yanan Wang, Shuichiro Haruta, Donghuo Zeng, Julio Vizcarra, Mori Kurokawa\",\"doi\":\"arxiv-2409.07747\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Video question answering (VideoQA) is a task to predict the correct answer to\\nquestions posed about a given video. The system must comprehend spatial and\\ntemporal relationships among objects extracted from videos to perform causal\\nand temporal reasoning. While prior works have focused on modeling individual\\nobject movements using transformer-based methods, they falter when capturing\\ncomplex scenarios involving multiple objects (e.g., \\\"a boy is throwing a ball\\nin a hoop\\\"). We propose a contrastive language event graph representation\\nlearning method called CLanG to address this limitation. Aiming to capture\\nevent representations associated with multiple objects, our method employs a\\nmulti-layer GNN-cluster module for adversarial graph representation learning,\\nenabling contrastive learning between the question text and its relevant\\nmulti-object event graph. Our method outperforms a strong baseline, achieving\\nup to 2.2% higher accuracy on two challenging VideoQA datasets, NExT-QA and\\nTGIF-QA-R. In particular, it is 2.8% better than baselines in handling causal\\nand temporal questions, highlighting its strength in reasoning multiple\\nobject-based events.\",\"PeriodicalId\":501030,\"journal\":{\"name\":\"arXiv - CS - Computation and Language\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-09-12\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - CS - Computation and Language\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2409.07747\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - CS - Computation and Language","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2409.07747","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

摘要

视频问题解答(VideoQA)是一项预测给定视频问题正确答案的任务。系统必须理解从视频中提取的物体之间的空间和时间关系,以执行因果和时间推理。之前的研究主要是使用基于变换器的方法对单个物体的运动进行建模,但在捕捉涉及多个物体的复杂场景时(如 "一个男孩正在把球扔进一个篮圈"),这些方法就显得力不从心了。针对这一局限,我们提出了一种名为 CLanG 的对比语言事件图表征学习方法。为了捕捉与多个对象相关的事件表征,我们的方法采用了多层 GNN 簇模块进行对抗图表征学习,实现了问题文本与其相关的多对象事件图之间的对比学习。我们的方法优于强大的基线,在两个具有挑战性的视频质量保证数据集 NExT-QA 和 TGIF-QA-R 上的准确率最高提高了 2.2%。特别是在处理因果和时间问题上,它比基线方法高出 2.8%,突出了它在推理多对象事件方面的优势。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
Multi-object event graph representation learning for Video Question Answering
Video question answering (VideoQA) is a task to predict the correct answer to questions posed about a given video. The system must comprehend spatial and temporal relationships among objects extracted from videos to perform causal and temporal reasoning. While prior works have focused on modeling individual object movements using transformer-based methods, they falter when capturing complex scenarios involving multiple objects (e.g., "a boy is throwing a ball in a hoop"). We propose a contrastive language event graph representation learning method called CLanG to address this limitation. Aiming to capture event representations associated with multiple objects, our method employs a multi-layer GNN-cluster module for adversarial graph representation learning, enabling contrastive learning between the question text and its relevant multi-object event graph. Our method outperforms a strong baseline, achieving up to 2.2% higher accuracy on two challenging VideoQA datasets, NExT-QA and TGIF-QA-R. In particular, it is 2.8% better than baselines in handling causal and temporal questions, highlighting its strength in reasoning multiple object-based events.
求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
自引率
0.00%
发文量
0
期刊最新文献
LLMs + Persona-Plug = Personalized LLMs MEOW: MEMOry Supervised LLM Unlearning Via Inverted Facts Extract-and-Abstract: Unifying Extractive and Abstractive Summarization within Single Encoder-Decoder Framework Development and bilingual evaluation of Japanese medical large language model within reasonably low computational resources Human-like Affective Cognition in Foundation Models
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1