Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Life behaviors, educational attainment, and risk for endometriosis: a Mendelian randomization study.

Archives of medical science : AMS·2025
Same author

Correction: Survey on knowledge, attitudes and practices (KAP) of malaria prevention and control among Chinese expatriates in South Sudan.

Journal of health, population, and nutrition·2025
Same author

Study on the skin structure, hair follicle cycle, and GSDMA protein expression in Ganxi goats.

Frontiers in veterinary science·2025
Same author

Identification and verification of prognostic genes related to zinc homeostasis and zinc transport in breast cancer.

PeerJ·2025
Same author

Unveiling the Catalytic Pathway of Rh(II)/Silicalite-2 in Propene Carbonylation to Methyl Butyrate: A DFT Study.

Molecules (Basel, Switzerland)·2025
Same author

GluN2A-NMDA receptor inhibition disinhibits the prefrontal cortex, reduces forced swim immobility, and impairs sensorimotor gating.

Acta pharmacologica Sinica·2025

Related Experiment Video

Updated: Sep 17, 2025

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

4.1K

Improved convolutional neural network for precise exercise posture recognition and intelligent health indicator

He Chen1, Rongchang Fan2

  • 1Ministry of Sports, Jiangsu Health Vocational College, Nanjing, 211800, Jiangsu, China.

Scientific Reports
|July 2, 2025
PubMed
Summary

This study introduces a new AI framework for precise exercise posture recognition and health prediction using advanced neural networks. It offers automated exercise analysis and personalized health insights for fitness and healthcare applications.

Keywords:
Convolutional neural networksExercise monitoringFeature fusionHealth indicator predictionPosture recognitionSpatiotemporal attention

More Related Videos

A Human-machine-interface Integrating Low-cost Sensors with a Neuromuscular Electrical Stimulation System for Post-stroke Balance Rehabilitation
11:06

A Human-machine-interface Integrating Low-cost Sensors with a Neuromuscular Electrical Stimulation System for Post-stroke Balance Rehabilitation

Published on: April 12, 2016

10.5K
Computerized Dynamic Posturography for Postural Control Assessment in Patients with Intermittent Claudication
14:52

Computerized Dynamic Posturography for Postural Control Assessment in Patients with Intermittent Claudication

Published on: December 11, 2013

11.5K

Related Experiment Videos

Last Updated: Sep 17, 2025

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

4.1K
A Human-machine-interface Integrating Low-cost Sensors with a Neuromuscular Electrical Stimulation System for Post-stroke Balance Rehabilitation
11:06

A Human-machine-interface Integrating Low-cost Sensors with a Neuromuscular Electrical Stimulation System for Post-stroke Balance Rehabilitation

Published on: April 12, 2016

10.5K
Computerized Dynamic Posturography for Postural Control Assessment in Patients with Intermittent Claudication
14:52

Computerized Dynamic Posturography for Postural Control Assessment in Patients with Intermittent Claudication

Published on: December 11, 2013

11.5K

Area of Science:

  • Artificial Intelligence in Sports Medicine
  • Computer Vision for Human Pose Estimation
  • Biometric Data Analysis

Background:

  • Accurate exercise posture recognition is crucial for injury prevention and training effectiveness.
  • Predicting health indicators from physical activity requires robust and adaptable systems.
  • Current methods often lack real-time capabilities or comprehensive health metric analysis.

Purpose of the Study:

  • To develop an AI framework for accurate exercise posture recognition.
  • To enable reliable prediction of multiple health indicators from exercise data.
  • To create a computationally efficient system for real-time applications.

Main Methods:

  • Implemented a multi-scale feature fusion architecture with spatiotemporal attention for improved key point detection.
  • Developed a Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM) model for health indicator prediction.
  • Incorporated personalized parameter adaptation for health forecasting and conducted evaluations on diverse datasets.

Main Results:

  • Achieved superior posture recognition with 78.6% mAP and 91.5% PCK@0.5.
  • Real-time inference capabilities at 27.3 FPS were maintained.
  • Health indicator prediction accuracy ranged from 86.1% to 92.6% across various metrics.

Conclusions:

  • The novel framework demonstrates high accuracy and efficiency in exercise posture recognition and health prediction.
  • The system shows robustness across different exercises, conditions, and demographics.
  • Offers significant potential for automated fitness coaching, rehabilitation monitoring, and preventive healthcare.