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

Adapting developmental science for a world of diverse families.

The Behavioral and brain sciences·2026
Same author

Re Du Ning exerts anti-influenza activity by counteracting viral NS1-mediated suppression of host type I interferon signaling.

Journal of ethnopharmacology·2026
Same author

PriMAT: Robust multi-animal tracking of primates in the wild.

PloS one·2026
Same author

Kinematic modulation across the ontogenetic transition to active social engagement.

Early human development·2026
Same author

Maternal-infant immune signatures in infants at risk for SARS-CoV-2-associated neurodevelopmental disorders.

Communications biology·2026
Same author

Communicative Development Inventories (CDIs) in etiologically diverse developmental conditions: A systematic review.

Research in developmental disabilities·2026

相关实验视频

Updated: May 20, 2025

Quantified Assessment of Infant's Gross Motor Abilities Using a Multisensor Wearable
09:24

Quantified Assessment of Infant's Gross Motor Abilities Using a Multisensor Wearable

Published on: May 17, 2024

1.3K

基于2D图像的方法进行婴儿姿势估计的比较.

Lennart Jahn1,2, Sarah Flügge3, Dajie Zhang4,5

  • 1Child and Adolescent Psychiatry and Psychotherapy, University Medical Center Göttingen; German Center for Child and Adolescent Health (DZKJ), Leibniz Science Campus Göttingen, Von-Siebold-Str. 5, Göttingen, Germany. lennart.jahn@phys.uni-goettingen.de.

Scientific reports
|April 9, 2025
PubMed
概括

像ViTPose这样的通用姿势估计模型在婴儿一般运动评估 (GMA) 中表现最好. 在婴儿数据上重新训练模型可以提高准确性,强调需要仔细选择姿势估计器,以获得可靠的自动化GMA.

关键词:
深度神经网络是一种深度神经网络.全身姿势估计估计这是GMA的GMA.婴儿运动分析

更多相关视频

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

1.4K
Gaze in Action: Head-mounted Eye Tracking of Children's Dynamic Visual Attention During Naturalistic Behavior
07:09

Gaze in Action: Head-mounted Eye Tracking of Children's Dynamic Visual Attention During Naturalistic Behavior

Published on: November 14, 2018

10.5K

相关实验视频

Last Updated: May 20, 2025

Quantified Assessment of Infant's Gross Motor Abilities Using a Multisensor Wearable
09:24

Quantified Assessment of Infant's Gross Motor Abilities Using a Multisensor Wearable

Published on: May 17, 2024

1.3K
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

1.4K
Gaze in Action: Head-mounted Eye Tracking of Children's Dynamic Visual Attention During Naturalistic Behavior
07:09

Gaze in Action: Head-mounted Eye Tracking of Children's Dynamic Visual Attention During Naturalistic Behavior

Published on: November 14, 2018

10.5K

科学领域:

  • 生物医学工程 生物医学工程
  • 发育儿科 发育儿科
  • 计算机视觉 计算机视觉

背景情况:

  • 自动通用运动评估 (GMA) 依赖于视频的准确婴儿姿势估计.
  • 现有的姿势估计器通常是通用的 (训练成年人) 或专门用于婴儿,性能各不相同.

研究的目的:

  • 为了比较通用和专用婴儿姿势估计器对自动化GMA的性能.
  • 评估视角 (对角与上下) 对婴儿姿势估计准确度的影响.

主要方法:

  • 从75个婴儿录音 (4-16周) 中利用了4500个注释的视频.
  • 计算错误和正确关键点的百分比 (PCK) 来评估姿势估计的准确性.
  • 将通用模型 (例如,ViTPose) 和婴儿特定模型的性能进行比较,并评估对角线和上下相机视图.

主要成果:

  • 在成年人身上训练的通用ViTPose模型在婴儿数据集上表现出卓越的表现.
  • 与通用产品相比,使用专门的婴儿姿势估计器没有观察到显著的改善.
  • 在婴儿数据上重新训练一个通用模型显著改善了姿势估计的准确性.
  • 顶向下摄像头视图比传统的对角视图产生了明显更好的姿势估计准确度.

结论:

  • 一般的姿势估计器,特别是当在目标数据上重新训练时,对于GMA中的婴儿姿势估计是有效的.
  • 婴儿特定姿势估计器在不同婴儿数据集中显示有限的概括能力.
  • 建议在未来的自动GMA录制设置中使用上下视角来提高准确性.