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

Genome-Wide Association Analysis of Gibberellin Sensitivity for Panicle Exsertion Length in Rice and Candidate Gene Identification.

Plants (Basel, Switzerland)·2026
Same author

Integrative and interpretable machine learning framework for early non-invasive detection of clinically significant liver fibrosis.

Frontiers in medicine·2026
Same author

Causal associations of circulating C-reactive protein with sarcopenia-related traits: A 2-sample Mendelian randomization study.

Medicine·2026
Same author

Voronoi-based analysis linking microscopic void evolution to macroscopic swelling in supercritical CO<sub>2</sub>-saturated polymers.

Physical chemistry chemical physics : PCCP·2026
Same author

GPR124 Alleviates Blood-Brain Barrier Disruption by Enhancing Microvascular Endothelial Function after Traumatic Brain Injury.

Advanced science (Weinheim, Baden-Wurttemberg, Germany)·2026
Same author

Evaluation of the TiaoShenZhiAi Regimen for Patients With Ovarian Cancer Experiencing a Psychoneurological Symptom Cluster: Protocol for a Multicenter, Double-Blind, Randomized Controlled Trial.

JMIR research protocols·2026

相关实验视频

Updated: Jul 26, 2025

Estimation of Contact Regions Between Hands and Objects During Human Multi-Digit Grasping
09:41

Estimation of Contact Regions Between Hands and Objects During Human Multi-Digit Grasping

Published on: April 21, 2023

1.6K

从单一图像和稀疏的IMU中重建3D人类姿势和形状.

Xianhua Liao1, Jiayan Zhuang2, Ze Liu3

  • 1School of Information Science and Engineering, Ningbo University, Ningbo, China.

PeerJ. Computer science
|June 22, 2023
PubMed
概括

这项研究通过将单个图像与稀疏的惯性测量单位 (IMU) 合并来增强3D人类姿势估计. 这种新的方法提高了准确性,减少了人类运动分析中的错误.

关键词:
3D人体姿势和形状一个单一的图像与稀疏的惯性测量单位.双流特征提取网络 双流特征提取网络具有残余模块的模型注意网络.回归是一种回归.

更多相关视频

3D Ultrasound Imaging: Fast and Cost-effective Morphometry of Musculoskeletal Tissue
08:52

3D Ultrasound Imaging: Fast and Cost-effective Morphometry of Musculoskeletal Tissue

Published on: November 27, 2017

23.3K
Author Spotlight: Advancements in 3D Optical Imaging for Comprehensive Body Composition Assessment in Modern Research
06:48

Author Spotlight: Advancements in 3D Optical Imaging for Comprehensive Body Composition Assessment in Modern Research

Published on: June 7, 2024

1.3K

相关实验视频

Last Updated: Jul 26, 2025

Estimation of Contact Regions Between Hands and Objects During Human Multi-Digit Grasping
09:41

Estimation of Contact Regions Between Hands and Objects During Human Multi-Digit Grasping

Published on: April 21, 2023

1.6K
3D Ultrasound Imaging: Fast and Cost-effective Morphometry of Musculoskeletal Tissue
08:52

3D Ultrasound Imaging: Fast and Cost-effective Morphometry of Musculoskeletal Tissue

Published on: November 27, 2017

23.3K
Author Spotlight: Advancements in 3D Optical Imaging for Comprehensive Body Composition Assessment in Modern Research
06:48

Author Spotlight: Advancements in 3D Optical Imaging for Comprehensive Body Composition Assessment in Modern Research

Published on: June 7, 2024

1.3K

科学领域:

  • 计算机视觉 计算机视觉
  • 人类运动分析分析
  • 生物医学工程 生物医学工程

背景情况:

  • 基于模型的3D姿势估计对于人类运动分析至关重要.
  • 基于视觉和基于惯性的方法具有诸如遮蔽和漂移等局限性.
  • 现有的混合方法面临的挑战是复杂的设备和缓慢的融合.

研究的目的:

  • 为了提高3D人类姿势估计准确度.
  • 开发一种将单个图像与稀疏惯性测量单位 (IMU) 融合的方法.
  • 克服现有的基于视觉,惯性和混合方法的局限性.

主要方法:

  • 使用双流特征提取网络.
  • 一个具有残余模块的模型注意网络融合了图像和IMU数据.
  • 通过直接回归策略获得3D姿势和形状参数.

主要成果:

  • 在总捕获数据集上,每顶点误差 (PVE) 减少了9.4mm.
  • 在Human3.6M数据集上,平均每关节位置误差 (MPJPE) 减少了7.8毫米.
  • 证明了稀疏的IMU数据和图像的有效融合,以提高姿势准确性.

结论:

  • 拟议的方法有效地融合了稀疏的IMU数据和单个图像.
  • 在3D人类姿势估计准确度方面取得了显著的改进.
  • 这种方法为无标记的人类运动捕捉提供了更强大,更准确的解决方案.