Search research articles

关于 JoVE

概览领导团队博客 JoVE 帮助中心

作者

出版流程编辑委员会范围与政策同行评审常见问题投稿

图书馆员

用户评价订阅访问资源图书馆顾问委员会常见问题

研究

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments 存档

教育

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual 教师资源中心教师网站

使用条款与条件

Search research articles

相关实验视频

Updated: Jan 17, 2026

Design and Analysis for Fall Detection System Simplification

Design and Analysis for Fall Detection System Simplification

Published on: April 6, 2020

实时活动和跌倒检测使用基于变压器的深度学习模型用于老年护理应用程序.

Raja Omman Zafar¹, Farhan Zafar²

¹Dalarna University-Campus Borlange, Borlänge, Sweden roz@du.se.

BMJ health & care informatics

|September 18, 2025

概括

此摘要是机器生成的。

相关概念视频

您也可能阅读

相关文章

通过共同作者、期刊和引用图与本文相关的文章。

排序

Same author

Ex Vivo Hydrodynamic Assessment of Expanded Polytetrafluoroethylene Trileaflet Pulmonary Valves.

ASAIO journal (American Society for Artificial Internal Organs : 1992)·2025

Same author

Stabilizing the Unstable: An Analysis of The Society of Thoracic Surgeons Congenital Heart Surgery Database of Children Supported With Extracorporeal Membrane Oxygenation Before Cardiac Surgery.

The Annals of thoracic surgery·2025

Same author

Machine Learning Driven Synthesis of Cobalt Oxide Entrapped Heteroatom-Doped Graphitic Carbon Nitride for Enhanced Oxygen Evolution Reaction.

PloS one·2025

Same author

Machine-Learning-Assisted Synthesis of Bimetallic Metal-Organic Frameworks for the Optimized Oxygen Evolution Reaction.

ACS applied materials & interfaces·2025

Same author

Clinical Outcomes With Normothermic Pulsatile Organ Perfusion in Heart Transplantation: A Report From the OCS Heart Perfusion Registry.

Circulation·2025

Same author

Letter in reference to "Association of Procurement Technique with Organ Yield and Cost Following Donation After Circulatory Death".

American journal of transplantation : official journal of the American Society of Transplantation and the American Society of Transplant Surgeons·2025

一个新的基于变压器的深度学习模型在实时活动识别和跌倒检测中实现了超过98%的准确性. 这种先进的系统的性能优于传统模型,为老年护理和防摔应用提供了更高的可靠性.

科学领域:

人工智能的人工智能
生物医学工程生物医学工程
计算机科学计算机科学

背景情况:

现有的活动识别和摔倒检测方法在准确性和实时性能方面扎.
穿戴式传感器数据分析对于监控日常生活活动和检测跌倒至关重要.

研究的目的:

开发基于变压器的深度学习模型,用于准确和实时的活动识别和落检测.
解决当前系统在准确性和实时适用性方面的局限性.

主要方法:

使用具有自我注意机制的变压器编码器通过滑动窗口分割来处理可穿戴传感器数据 (加速计,陀螺仪,定向).
在广泛的MobiAct数据集上训练和评估模型,包括来自66名参与者的超过1400万条记录,涉及16项活动,包括各种落类型.

主要成果:

达到了超过98%的准确性,具有极好的精度和复杂的降落类别的回忆,如向前倾斜和向侧倾斜.
与卷积神经网络-长期短期记忆 (CNN-LSTM) 和时间卷积网络相比,在分类指标和训练稳定性方面表现出卓越的性能.

结论:

变压器模型有效地捕捉复杂的时间依赖性,减轻错误分类和错误阳性在活动识别和跌落检测.
开发的基于变压器的系统提供了高效的实时部署和可靠的解决方案,用于老年人护理和预防跌倒.

关键词:

人工智能的人工智能是人工智能.BMJ 医疗信息学医疗信息学数据科学数据科学数据科学深度学习 (Deep Learning) 是一种深度学习.提供医疗保健服务.

更多相关视频

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

相关实验视频

Last Updated: Jan 17, 2026

Design and Analysis for Fall Detection System Simplification

Design and Analysis for Fall Detection System Simplification

Published on: April 6, 2020

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023