视频理解中注意机制的不同分类

IF 6.8 1区计算机科学 0 COMPUTER SCIENCE, INFORMATION SYSTEMS

Information Sciences Pub Date : 2025-08-26 DOI:10.1016/j.ins.2025.122609

Shilin Chen , Xingwang Wang , Yafeng Sun , Kun Yang , Xiaohui Wei

{"title":"视频理解中注意机制的不同分类","authors":"Shilin Chen , Xingwang Wang , Yafeng Sun , Kun Yang , Xiaohui Wei","doi":"10.1016/j.ins.2025.122609","DOIUrl":null,"url":null,"abstract":"<div><div>The exponential growth of video data necessitates advanced attention mechanisms to address computational efficiency and recognition accuracy challenges. This paper proposes a novel taxonomy categorizing attention mechanisms into feature-related, structure-related, and query-related classes, analyzing their roles in optimizing video understanding models. We analyze the challenges and assessment metrics addressed by different models incorporating attention mechanisms. Through a comprehensive analysis of representative models across different application scenarios, we provide a unified framework for selecting attention mechanisms and identifying future research directions. The proposed taxonomy bridges the gap between recent research and practical implementation, offering guidance for optimizing video understanding systems.</div></div>","PeriodicalId":51063,"journal":{"name":"Information Sciences","volume":"721 ","pages":"Article 122609"},"PeriodicalIF":6.8000,"publicationDate":"2025-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"A distinct classification of attention mechanisms in video understanding\",\"authors\":\"Shilin Chen , Xingwang Wang , Yafeng Sun , Kun Yang , Xiaohui Wei\",\"doi\":\"10.1016/j.ins.2025.122609\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><div>The exponential growth of video data necessitates advanced attention mechanisms to address computational efficiency and recognition accuracy challenges. This paper proposes a novel taxonomy categorizing attention mechanisms into feature-related, structure-related, and query-related classes, analyzing their roles in optimizing video understanding models. We analyze the challenges and assessment metrics addressed by different models incorporating attention mechanisms. Through a comprehensive analysis of representative models across different application scenarios, we provide a unified framework for selecting attention mechanisms and identifying future research directions. The proposed taxonomy bridges the gap between recent research and practical implementation, offering guidance for optimizing video understanding systems.</div></div>\",\"PeriodicalId\":51063,\"journal\":{\"name\":\"Information Sciences\",\"volume\":\"721 \",\"pages\":\"Article 122609\"},\"PeriodicalIF\":6.8000,\"publicationDate\":\"2025-08-26\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Information Sciences\",\"FirstCategoryId\":\"94\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S002002552500742X\",\"RegionNum\":1,\"RegionCategory\":\"计算机科学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"0\",\"JCRName\":\"COMPUTER SCIENCE, INFORMATION SYSTEMS\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Information Sciences","FirstCategoryId":"94","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S002002552500742X","RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"0","JCRName":"COMPUTER SCIENCE, INFORMATION SYSTEMS","Score":null,"Total":0}

引用次数: 0

摘要

视频数据的指数级增长需要先进的注意力机制来解决计算效率和识别精度的挑战。本文提出了一种新的分类方法，将注意力机制分为特征相关类、结构相关类和查询相关类，并分析了它们在优化视频理解模型中的作用。我们分析了包含注意力机制的不同模型所解决的挑战和评估指标。通过对不同应用场景代表性模型的综合分析，为选择注意机制和确定未来研究方向提供了统一的框架。提出的分类法弥合了最近的研究和实际实施之间的差距，为优化视频理解系统提供了指导。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

查看原文本刊更多论文

A distinct classification of attention mechanisms in video understanding

The exponential growth of video data necessitates advanced attention mechanisms to address computational efficiency and recognition accuracy challenges. This paper proposes a novel taxonomy categorizing attention mechanisms into feature-related, structure-related, and query-related classes, analyzing their roles in optimizing video understanding models. We analyze the challenges and assessment metrics addressed by different models incorporating attention mechanisms. Through a comprehensive analysis of representative models across different application scenarios, we provide a unified framework for selecting attention mechanisms and identifying future research directions. The proposed taxonomy bridges the gap between recent research and practical implementation, offering guidance for optimizing video understanding systems.

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

Information Sciences 工程技术-计算机：信息系统

CiteScore

14.00

自引率

17.30%

发文量

1322

审稿时长

10.4 months

期刊介绍： Informatics and Computer Science Intelligent Systems Applications is an esteemed international journal that focuses on publishing original and creative research findings in the field of information sciences. We also feature a limited number of timely tutorial and surveying contributions. Our journal aims to cater to a diverse audience, including researchers, developers, managers, strategic planners, graduate students, and anyone interested in staying up-to-date with cutting-edge research in information science, knowledge engineering, and intelligent systems. While readers are expected to share a common interest in information science, they come from varying backgrounds such as engineering, mathematics, statistics, physics, computer science, cell biology, molecular biology, management science, cognitive science, neurobiology, behavioral sciences, and biochemistry.