用于端到端手势识别的可转移非侵入式模态融合变换器(NIMFT)

IF 3.7 3区 医学 Q2 ENGINEERING, BIOMEDICAL Journal of neural engineering Pub Date : 2024-04-09 DOI:10.1088/1741-2552/ad39a5
Tianxiang Xu, Kunkun Zhao, Yuxiang Hu, Liang Li, Wei Wang, Fulin Wang, Yuxuan Zhou, Jianqing Li
{"title":"用于端到端手势识别的可转移非侵入式模态融合变换器(NIMFT)","authors":"Tianxiang Xu, Kunkun Zhao, Yuxiang Hu, Liang Li, Wei Wang, Fulin Wang, Yuxuan Zhou, Jianqing Li","doi":"10.1088/1741-2552/ad39a5","DOIUrl":null,"url":null,"abstract":"<italic toggle=\"yes\">Objective.</italic> Recent studies have shown that integrating inertial measurement unit (IMU) signals with surface electromyographic (sEMG) can greatly improve hand gesture recognition (HGR) performance in applications such as prosthetic control and rehabilitation training. However, current deep learning models for multimodal HGR encounter difficulties in invasive modal fusion, complex feature extraction from heterogeneous signals, and limited inter-subject model generalization. To address these challenges, this study aims to develop an end-to-end and inter-subject transferable model that utilizes non-invasively fused sEMG and acceleration (ACC) data. <italic toggle=\"yes\">Approach.</italic> The proposed non-invasive modal fusion-transformer (NIMFT) model utilizes 1D-convolutional neural networks-based patch embedding for local information extraction and employs a multi-head cross-attention (MCA) mechanism to non-invasively integrate sEMG and ACC signals, stabilizing the variability induced by sEMG. The proposed architecture undergoes detailed ablation studies after hyperparameter tuning. Transfer learning is employed by fine-tuning a pre-trained model on new subject and a comparative analysis is performed between the fine-tuning and subject-specific model. Additionally, the performance of NIMFT is compared to state-of-the-art fusion models. <italic toggle=\"yes\">Main results.</italic> The NIMFT model achieved recognition accuracies of 93.91%, 91.02%, and 95.56% on the three action sets in the Ninapro DB2 dataset. The proposed embedding method and MCA outperformed the traditional invasive modal fusion transformer by 2.01% (embedding) and 1.23% (fusion), respectively. In comparison to subject-specific models, the fine-tuning model exhibited the highest average accuracy improvement of 2.26%, achieving a final accuracy of 96.13%. Moreover, the NIMFT model demonstrated superiority in terms of accuracy, recall, precision, and F1-score compared to the latest modal fusion models with similar model scale. <italic toggle=\"yes\">Significance.</italic> The NIMFT is a novel end-to-end HGR model, utilizes a non-invasive MCA mechanism to integrate long-range intermodal information effectively. Compared to recent modal fusion models, it demonstrates superior performance in inter-subject experiments and offers higher training efficiency and accuracy levels through transfer learning than subject-specific approaches.","PeriodicalId":16753,"journal":{"name":"Journal of neural engineering","volume":"108 1","pages":""},"PeriodicalIF":3.7000,"publicationDate":"2024-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Transferable non-invasive modal fusion-transformer (NIMFT) for end-to-end hand gesture recognition\",\"authors\":\"Tianxiang Xu, Kunkun Zhao, Yuxiang Hu, Liang Li, Wei Wang, Fulin Wang, Yuxuan Zhou, Jianqing Li\",\"doi\":\"10.1088/1741-2552/ad39a5\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<italic toggle=\\\"yes\\\">Objective.</italic> Recent studies have shown that integrating inertial measurement unit (IMU) signals with surface electromyographic (sEMG) can greatly improve hand gesture recognition (HGR) performance in applications such as prosthetic control and rehabilitation training. However, current deep learning models for multimodal HGR encounter difficulties in invasive modal fusion, complex feature extraction from heterogeneous signals, and limited inter-subject model generalization. To address these challenges, this study aims to develop an end-to-end and inter-subject transferable model that utilizes non-invasively fused sEMG and acceleration (ACC) data. <italic toggle=\\\"yes\\\">Approach.</italic> The proposed non-invasive modal fusion-transformer (NIMFT) model utilizes 1D-convolutional neural networks-based patch embedding for local information extraction and employs a multi-head cross-attention (MCA) mechanism to non-invasively integrate sEMG and ACC signals, stabilizing the variability induced by sEMG. The proposed architecture undergoes detailed ablation studies after hyperparameter tuning. Transfer learning is employed by fine-tuning a pre-trained model on new subject and a comparative analysis is performed between the fine-tuning and subject-specific model. Additionally, the performance of NIMFT is compared to state-of-the-art fusion models. <italic toggle=\\\"yes\\\">Main results.</italic> The NIMFT model achieved recognition accuracies of 93.91%, 91.02%, and 95.56% on the three action sets in the Ninapro DB2 dataset. The proposed embedding method and MCA outperformed the traditional invasive modal fusion transformer by 2.01% (embedding) and 1.23% (fusion), respectively. In comparison to subject-specific models, the fine-tuning model exhibited the highest average accuracy improvement of 2.26%, achieving a final accuracy of 96.13%. Moreover, the NIMFT model demonstrated superiority in terms of accuracy, recall, precision, and F1-score compared to the latest modal fusion models with similar model scale. <italic toggle=\\\"yes\\\">Significance.</italic> The NIMFT is a novel end-to-end HGR model, utilizes a non-invasive MCA mechanism to integrate long-range intermodal information effectively. Compared to recent modal fusion models, it demonstrates superior performance in inter-subject experiments and offers higher training efficiency and accuracy levels through transfer learning than subject-specific approaches.\",\"PeriodicalId\":16753,\"journal\":{\"name\":\"Journal of neural engineering\",\"volume\":\"108 1\",\"pages\":\"\"},\"PeriodicalIF\":3.7000,\"publicationDate\":\"2024-04-09\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Journal of neural engineering\",\"FirstCategoryId\":\"5\",\"ListUrlMain\":\"https://doi.org/10.1088/1741-2552/ad39a5\",\"RegionNum\":3,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q2\",\"JCRName\":\"ENGINEERING, BIOMEDICAL\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of neural engineering","FirstCategoryId":"5","ListUrlMain":"https://doi.org/10.1088/1741-2552/ad39a5","RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"ENGINEERING, BIOMEDICAL","Score":null,"Total":0}
引用次数: 0

摘要

目的。最近的研究表明,在假肢控制和康复训练等应用中,将惯性测量单元(IMU)信号与表面肌电图(sEMG)整合在一起可以大大提高手势识别(HGR)性能。然而,目前用于多模态手势识别的深度学习模型在侵入式模态融合、异构信号的复杂特征提取和有限的主体间模型泛化方面遇到了困难。为了应对这些挑战,本研究旨在利用非侵入式融合的 sEMG 和加速度(ACC)数据,开发一种端到端和受试者间可转移的模型。方法。所提出的无创模态融合-转换器(NIMFT)模型利用基于一维卷积神经网络的补丁嵌入进行局部信息提取,并采用多头交叉注意(MCA)机制对 sEMG 和 ACC 信号进行无创融合,从而稳定 sEMG 引起的变异性。在超参数调整后,对所提出的架构进行了详细的消融研究。通过在新受试者身上微调预先训练好的模型,并在微调模型和特定受试者模型之间进行比较分析,采用了迁移学习方法。此外,还将 NIMFT 的性能与最先进的融合模型进行了比较。主要结果。在 Ninapro DB2 数据集中的三个动作集上,NIMFT 模型的识别准确率分别达到了 93.91%、91.02% 和 95.56%。提议的嵌入方法和 MCA 比传统的侵入式模态融合转换器分别高出 2.01%(嵌入)和 1.23%(融合)。与特定对象模型相比,微调模型的平均准确率提高了 2.26%,达到了 96.13% 的最终准确率。此外,与模型规模相似的最新模态融合模型相比,NIMFT 模型在准确度、召回率、精确度和 F1 分数方面都表现出了优势。意义重大。NIMFT 是一种新颖的端到端 HGR 模型,利用非侵入式 MCA 机制有效地整合了远距离多式联运信息。与最新的模态融合模型相比,它在受试者间实验中表现出更优越的性能,并通过迁移学习提供比特定受试者方法更高的训练效率和准确度。
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
Transferable non-invasive modal fusion-transformer (NIMFT) for end-to-end hand gesture recognition
Objective. Recent studies have shown that integrating inertial measurement unit (IMU) signals with surface electromyographic (sEMG) can greatly improve hand gesture recognition (HGR) performance in applications such as prosthetic control and rehabilitation training. However, current deep learning models for multimodal HGR encounter difficulties in invasive modal fusion, complex feature extraction from heterogeneous signals, and limited inter-subject model generalization. To address these challenges, this study aims to develop an end-to-end and inter-subject transferable model that utilizes non-invasively fused sEMG and acceleration (ACC) data. Approach. The proposed non-invasive modal fusion-transformer (NIMFT) model utilizes 1D-convolutional neural networks-based patch embedding for local information extraction and employs a multi-head cross-attention (MCA) mechanism to non-invasively integrate sEMG and ACC signals, stabilizing the variability induced by sEMG. The proposed architecture undergoes detailed ablation studies after hyperparameter tuning. Transfer learning is employed by fine-tuning a pre-trained model on new subject and a comparative analysis is performed between the fine-tuning and subject-specific model. Additionally, the performance of NIMFT is compared to state-of-the-art fusion models. Main results. The NIMFT model achieved recognition accuracies of 93.91%, 91.02%, and 95.56% on the three action sets in the Ninapro DB2 dataset. The proposed embedding method and MCA outperformed the traditional invasive modal fusion transformer by 2.01% (embedding) and 1.23% (fusion), respectively. In comparison to subject-specific models, the fine-tuning model exhibited the highest average accuracy improvement of 2.26%, achieving a final accuracy of 96.13%. Moreover, the NIMFT model demonstrated superiority in terms of accuracy, recall, precision, and F1-score compared to the latest modal fusion models with similar model scale. Significance. The NIMFT is a novel end-to-end HGR model, utilizes a non-invasive MCA mechanism to integrate long-range intermodal information effectively. Compared to recent modal fusion models, it demonstrates superior performance in inter-subject experiments and offers higher training efficiency and accuracy levels through transfer learning than subject-specific approaches.
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
Journal of neural engineering
Journal of neural engineering 工程技术-工程:生物医学
CiteScore
7.80
自引率
12.50%
发文量
319
审稿时长
4.2 months
期刊介绍: The goal of Journal of Neural Engineering (JNE) is to act as a forum for the interdisciplinary field of neural engineering where neuroscientists, neurobiologists and engineers can publish their work in one periodical that bridges the gap between neuroscience and engineering. The journal publishes articles in the field of neural engineering at the molecular, cellular and systems levels. The scope of the journal encompasses experimental, computational, theoretical, clinical and applied aspects of: Innovative neurotechnology; Brain-machine (computer) interface; Neural interfacing; Bioelectronic medicines; Neuromodulation; Neural prostheses; Neural control; Neuro-rehabilitation; Neurorobotics; Optical neural engineering; Neural circuits: artificial & biological; Neuromorphic engineering; Neural tissue regeneration; Neural signal processing; Theoretical and computational neuroscience; Systems neuroscience; Translational neuroscience; Neuroimaging.
期刊最新文献
Building consensus on clinical outcome assessments for BCI devices. A summary of the 10th BCI society meeting 2023 workshop. o-CLEAN: a novel multi-stage algorithm for the ocular artifacts' correction from EEG data in out-of-the-lab applications. PDMS/CNT electrodes with bioamplifier for practical in-the-ear and conventional biosignal recordings. DOCTer: a novel EEG-based diagnosis framework for disorders of consciousness. I see artifacts: ICA-based EEG artifact removal does not improve deep network decoding across three BCI tasks.
×
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