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

The Cutting Edge: Interventional Strategies for Pain Management.

Physical medicine and rehabilitation clinics of North America·2026
Same author

Artificial Intelligence Across the Radiology Workflow: A Nine-Stage Narrative Review.

Diagnostics (Basel, Switzerland)·2026
Same author

An explainable multi-stage framework for brain tumor classification using hybrid feature fusion and EfficientNetB5 model.

Scientific reports·2026
Same author

A gated task-attentive multi-task network for unified retinal image analysis.

Scientific reports·2026
Same author

Characteristics and Outcomes of Patients With Malignancies Prior to Colorectal Cancer: A Propensity Score Matched Analysis.

Journal of gastrointestinal cancer·2026
Same author

Rodent Social Behavior Recognition Using a Global Context-Aware Vision Transformer Network.

AI (Basel, Switzerland)·2026

相关实验视频

Updated: Jan 9, 2026

Human Brown Adipose Tissue Depots Automatically Segmented by Positron Emission Tomography/Computed Tomography and Registered Magnetic Resonance Images
09:21

Human Brown Adipose Tissue Depots Automatically Segmented by Positron Emission Tomography/Computed Tomography and Registered Magnetic Resonance Images

Published on: February 18, 2015

12.6K

EfficientPoseSegNet:一个弱监督的,以注意力为导向的框架,用于人类姿势估计,解剖细分和隐藏物体检测,用于背散毫米波安全查.

Muhammad Zaheer Sajid1, Muhammad Fareed Hamid2, Imran Qureshi3

  • 1Department of Electrical Engineering and Computer Science, University of Missouri, Columbia, 65211, USA. ms2wt@missouri.edu.

Scientific reports
|December 5, 2025
PubMed
概括

本研究介绍了EfficientPoseSegNet,这是一种用于加强机场安全的深度学习模型. 它准确地检测隐藏的物体,并分析毫米波图像中的人类姿势和身体部位.

关键词:
身体部位的细分 身体部位的细分人类姿势估计估计关键点估计 关键点估计对象检测,异常检测检测.姿势改进,二维姿势校正

更多相关视频

Author Spotlight: Development of an Automated Camera-Based System for Real-Time Blast Overpressure Monitoring and TBI Risk Assessment in Military Training
06:20

Author Spotlight: Development of an Automated Camera-Based System for Real-Time Blast Overpressure Monitoring and TBI Risk Assessment in Military Training

Published on: December 6, 2024

3.2K
Bioelectric Analyses of an Osseointegrated Intelligent Implant Design System for Amputees
14:31

Bioelectric Analyses of an Osseointegrated Intelligent Implant Design System for Amputees

Published on: July 15, 2009

14.5K

相关实验视频

Last Updated: Jan 9, 2026

Human Brown Adipose Tissue Depots Automatically Segmented by Positron Emission Tomography/Computed Tomography and Registered Magnetic Resonance Images
09:21

Human Brown Adipose Tissue Depots Automatically Segmented by Positron Emission Tomography/Computed Tomography and Registered Magnetic Resonance Images

Published on: February 18, 2015

12.6K
Author Spotlight: Development of an Automated Camera-Based System for Real-Time Blast Overpressure Monitoring and TBI Risk Assessment in Military Training
06:20

Author Spotlight: Development of an Automated Camera-Based System for Real-Time Blast Overpressure Monitoring and TBI Risk Assessment in Military Training

Published on: December 6, 2024

3.2K
Bioelectric Analyses of an Osseointegrated Intelligent Implant Design System for Amputees
14:31

Bioelectric Analyses of an Osseointegrated Intelligent Implant Design System for Amputees

Published on: July 15, 2009

14.5K

科学领域:

  • 计算机视觉和机器学习
  • 运输科学与物流 运输科学与物流
  • 安全和监控技术安全和监控技术

背景情况:

  • 在毫米波图像中自动检测隐藏的物体,解剖关键点和人体部位对于机场安全至关重要.
  • 由于图像质量差,注释有限以及隐私问题,现有系统面临挑战.
  • 对实时安全查的高效,强大和符合隐私的解决方案的需求.

研究的目的:

  • 开发EfficientPoseSegNet,这是一个混合深度学习框架,用于在安全查图像中高效地使用注释.
  • 为了实现准确的隐藏物体检测,人类姿势估计和身体部位细分.
  • 提高自动化安全查系统的可靠性和可扩展性.

主要方法:

  • 利用并行EfficientNet和DenseNet的骨干来从低分辨率扫描中进行多尺度的特征提取.
  • 整合了一个卷积块注意模块 (CBAM),以专注于关键的解剖区域并减少噪音.
  • 使用软-argmax的空间热图来提取关键点,将身体细分为17个区域,以及在弱监督下进行强有力的训练的随机重量平均 (SWA).

主要成果:

  • 在关键点检测 (99.79%),姿势估计 (99%) 和身体细分 (97% IoU) 中实现了高精度.
  • 在隐藏物体检测方面表现强,平均异常检测AUC为0.94.4.
  • 在TSA乘客查数据集上,EfficientPoseSegNet显示了测试损失和平均绝对误差的显著改善.

结论:

  • EfficientPoseSegNet提供了一个可扩展的,符合隐私的解决方案,用于实时的人体姿势估计,身体部位细分和隐藏物体检测.
  • 混合深度学习框架有效地解决了毫米波图像分析方面的挑战,以提高运输安全.
  • 这种方法对快速发展的环境中安全查技术的进步做出了重大贡献.