Jove
Visualize
联系我们
JoVE
x logofacebook logolinkedin logoyoutube logo
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 JournalMethods CollectionsJoVE Encyclopedia of Experiments存档
教育
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab Manual教师资源中心教师网站
使用条款与条件
隐私政策
政策

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

User-Centered Evaluation of the Wearable Walker Lower Limb Exoskeleton; Preliminary Assessment Based on the Experience Protocol.

Sensors (Basel, Switzerland)·2024
Same author

Closed-Chain Inverse Dynamics for the Biomechanical Analysis of Manual Material Handling Tasks through a Deep Learning Assisted Wearable Sensor Network.

Sensors (Basel, Switzerland)·2023
Same author

Wearable Sensor Network for Biomechanical Overload Assessment in Manual Material Handling.

Sensors (Basel, Switzerland)·2020
Same author

Magnetic Levitation for Soft-Tethered Capsule Colonoscopy Actuated With a Single Permanent Magnet: A Dynamic Control Approach.

IEEE robotics and automation letters·2019

相关实验视频

Updated: Jan 11, 2026

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
06:37

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

Published on: December 15, 2023

5.2K

深度学习算法用于识别人工物料处理任务中的人类活动.

Giulia Bassani1,2, Carlo Alberto Avizzano1,2, Alessandro Filippeschi1,2

  • 1Institute of Mechanical Intelligence, Scuola Superiore Sant'Anna, 56124 Pisa, Italy.

Sensors (Basel, Switzerland)
|November 13, 2025
PubMed
概括

像BiLSTM和RCNN这样的深度学习模型显示出使用可穿戴传感器识别手动物料处理活动的前景. 这些在计算上较轻的算法提供了与复杂模型相匹配的性能,有助于人体工程学风险评估.

关键词:
自动编码器自动编码器卷积神经网络 (CNN) 是一种神经网络.人类活动识别 (HAR)手动物料处理 (MMH) 是指手动物料处理.经常性神经网络 (RNN)可穿戴传感器网络 (WSN)

更多相关视频

A Step-by-Step Implementation of DeepBehavior, Deep Learning Toolbox for Automated Behavior Analysis
05:41

A Step-by-Step Implementation of DeepBehavior, Deep Learning Toolbox for Automated Behavior Analysis

Published on: February 6, 2020

9.8K
Combining Eye-tracking Data with an Analysis of Video Content from Free-viewing a Video of a Walk in an Urban Park Environment
08:25

Combining Eye-tracking Data with an Analysis of Video Content from Free-viewing a Video of a Walk in an Urban Park Environment

Published on: May 7, 2019

9.5K

相关实验视频

Last Updated: Jan 11, 2026

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
06:37

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

Published on: December 15, 2023

5.2K
A Step-by-Step Implementation of DeepBehavior, Deep Learning Toolbox for Automated Behavior Analysis
05:41

A Step-by-Step Implementation of DeepBehavior, Deep Learning Toolbox for Automated Behavior Analysis

Published on: February 6, 2020

9.8K
Combining Eye-tracking Data with an Analysis of Video Content from Free-viewing a Video of a Walk in an Urban Park Environment
08:25

Combining Eye-tracking Data with an Analysis of Video Content from Free-viewing a Video of a Walk in an Urban Park Environment

Published on: May 7, 2019

9.5K

科学领域:

  • 职业健康和安全问题 职业健康和安全问题
  • 生物医学工程 生物医学工程
  • 人工智能的人工智能

背景情况:

  • 人类活动识别 (HAR) 在医疗保健中至关重要,但在手动物料处理 (MMH) 中未得到充分利用.
  • MMH的活动对工作人员的健康和安全产生重大影响.
  • 现有的HAR研究很少解决MMH的具体挑战.

研究的目的:

  • 开发和评估深度学习算法用于MMH中的HAR.
  • 为了比较不同HAR模型的性能和计算复杂性.
  • 确定适合实时人体工程学风险评估的高效算法.

主要方法:

  • 提出了四种深度学习算法:双向长期短期记忆 (BiLSTM),稀疏排斥自编码器 (Sp-DAE),循环Sp-DAE和循环卷积神经网络 (RCNN).
  • 利用来自14名受试者的可穿戴传感器数据来训练和测试模型.
  • 使用F1分数,70-30%分割和Leave-One-Subject-Out (LOSO) 验证,以及计算复杂度指标 (MAC,MA) 的性能比较.

主要成果:

  • BiLSTM获得了最高的分类性能 (95.7%在70-30%分割上,90.3%在LOSO上).
  • RCNN表现相似 (95.9%; 89.2%),随着受试者数量的增加,结果有所改善.
  • 虽然DeepConvLSTM的性能相对较高,但在计算上比BiLSTM和RCNN要复杂得多.

结论:

  • BiLSTM和RCNN为MMH提供了强大的分类准确性和计算效率平衡.
  • 这些较轻的模型适用于嵌入式系统和自动人体工程学风险评估.
  • 促进个性化风险评估,提高工业MMH实践中的安全性.