用于词级唇语阅读的协作式词汇子词和端到端建模

IF 8.4 1区 计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS
Hang Chen;Qing Wang;Jun Du;Gen-Shun Wan;Shi-Fu Xiong;Bao-Ci Yin;Jia Pan;Chin-Hui Lee
{"title":"用于词级唇语阅读的协作式词汇子词和端到端建模","authors":"Hang Chen;Qing Wang;Jun Du;Gen-Shun Wan;Shi-Fu Xiong;Bao-Ci Yin;Jia Pan;Chin-Hui Lee","doi":"10.1109/TMM.2024.3390148","DOIUrl":null,"url":null,"abstract":"We propose a viseme subword modeling (VSM) approach to improve the generalizability and interpretability capabilities of deep neural network based lip reading. A comprehensive analysis of preliminary experimental results reveals the complementary nature of the conventional end-to-end (E2E) and proposed VSM frameworks, especially concerning speaker head movements. To increase lip reading accuracy, we propose hybrid viseme subwords and end-to-end modeling (HVSEM), which exploits the strengths of both approaches through multitask learning. As an extension to HVSEM, we also propose collaborative viseme subword and end-to-end modeling (CVSEM), which further explores the synergy between the VSM and E2E frameworks by integrating a state-mapped temporal mask (SMTM) into joint modeling. Experimental evaluations using different model backbones on both the LRW and LRW-1000 datasets confirm the superior performance and generalizability of the proposed frameworks. Specifically, VSM outperforms the baseline E2E framework, while HVSEM outperforms VSM in a hybrid combination of VSM and E2E modeling. Building on HVSEM, CVSEM further achieves impressive accuracies on 90.75% and 58.89%, setting new benchmarks for both datasets.","PeriodicalId":13273,"journal":{"name":"IEEE Transactions on Multimedia","volume":"26 ","pages":"9358-9371"},"PeriodicalIF":8.4000,"publicationDate":"2024-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Collaborative Viseme Subword and End-to-End Modeling for Word-Level Lip Reading\",\"authors\":\"Hang Chen;Qing Wang;Jun Du;Gen-Shun Wan;Shi-Fu Xiong;Bao-Ci Yin;Jia Pan;Chin-Hui Lee\",\"doi\":\"10.1109/TMM.2024.3390148\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"We propose a viseme subword modeling (VSM) approach to improve the generalizability and interpretability capabilities of deep neural network based lip reading. A comprehensive analysis of preliminary experimental results reveals the complementary nature of the conventional end-to-end (E2E) and proposed VSM frameworks, especially concerning speaker head movements. To increase lip reading accuracy, we propose hybrid viseme subwords and end-to-end modeling (HVSEM), which exploits the strengths of both approaches through multitask learning. As an extension to HVSEM, we also propose collaborative viseme subword and end-to-end modeling (CVSEM), which further explores the synergy between the VSM and E2E frameworks by integrating a state-mapped temporal mask (SMTM) into joint modeling. Experimental evaluations using different model backbones on both the LRW and LRW-1000 datasets confirm the superior performance and generalizability of the proposed frameworks. Specifically, VSM outperforms the baseline E2E framework, while HVSEM outperforms VSM in a hybrid combination of VSM and E2E modeling. Building on HVSEM, CVSEM further achieves impressive accuracies on 90.75% and 58.89%, setting new benchmarks for both datasets.\",\"PeriodicalId\":13273,\"journal\":{\"name\":\"IEEE Transactions on Multimedia\",\"volume\":\"26 \",\"pages\":\"9358-9371\"},\"PeriodicalIF\":8.4000,\"publicationDate\":\"2024-04-17\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"IEEE Transactions on Multimedia\",\"FirstCategoryId\":\"94\",\"ListUrlMain\":\"https://ieeexplore.ieee.org/document/10504606/\",\"RegionNum\":1,\"RegionCategory\":\"计算机科学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"COMPUTER SCIENCE, INFORMATION SYSTEMS\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"IEEE Transactions on Multimedia","FirstCategoryId":"94","ListUrlMain":"https://ieeexplore.ieee.org/document/10504606/","RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"COMPUTER SCIENCE, INFORMATION SYSTEMS","Score":null,"Total":0}
引用次数: 0

摘要

我们提出了一种意象子词建模(VSM)方法,以提高基于深度神经网络的唇语阅读的泛化和可解释能力。对初步实验结果的综合分析表明,传统的端到端(E2E)框架和所提出的 VSM 框架具有互补性,尤其是在说话者头部运动方面。为了提高唇读的准确性,我们提出了混合视觉子词和端到端建模(HVSEM),通过多任务学习利用两种方法的优势。作为 HVSEM 的扩展,我们还提出了协作视觉子词和端到端建模(CVSEM),通过将状态映射时间掩码(SMTM)集成到联合建模中,进一步探索了 VSM 和 E2E 框架之间的协同作用。在 LRW 和 LRW-1000 数据集上使用不同的模型骨干进行的实验评估证实了所提出的框架具有卓越的性能和通用性。具体来说,VSM 的性能优于基准 E2E 框架,而在 VSM 和 E2E 建模的混合组合中,HVSEM 的性能优于 VSM。在 HVSEM 的基础上,CVSEM 进一步达到了令人印象深刻的 90.75% 和 58.89% 的准确率,为这两个数据集设定了新的基准。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
Collaborative Viseme Subword and End-to-End Modeling for Word-Level Lip Reading
We propose a viseme subword modeling (VSM) approach to improve the generalizability and interpretability capabilities of deep neural network based lip reading. A comprehensive analysis of preliminary experimental results reveals the complementary nature of the conventional end-to-end (E2E) and proposed VSM frameworks, especially concerning speaker head movements. To increase lip reading accuracy, we propose hybrid viseme subwords and end-to-end modeling (HVSEM), which exploits the strengths of both approaches through multitask learning. As an extension to HVSEM, we also propose collaborative viseme subword and end-to-end modeling (CVSEM), which further explores the synergy between the VSM and E2E frameworks by integrating a state-mapped temporal mask (SMTM) into joint modeling. Experimental evaluations using different model backbones on both the LRW and LRW-1000 datasets confirm the superior performance and generalizability of the proposed frameworks. Specifically, VSM outperforms the baseline E2E framework, while HVSEM outperforms VSM in a hybrid combination of VSM and E2E modeling. Building on HVSEM, CVSEM further achieves impressive accuracies on 90.75% and 58.89%, setting new benchmarks for both datasets.
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
IEEE Transactions on Multimedia
IEEE Transactions on Multimedia 工程技术-电信学
CiteScore
11.70
自引率
11.00%
发文量
576
审稿时长
5.5 months
期刊介绍: The IEEE Transactions on Multimedia delves into diverse aspects of multimedia technology and applications, covering circuits, networking, signal processing, systems, software, and systems integration. The scope aligns with the Fields of Interest of the sponsors, ensuring a comprehensive exploration of research in multimedia.
×
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
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
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 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信