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

DC-FusionGNN: A Dual-Channel Framework Integrating Global Self-Attention and Local Topology Learning for Identifying Key Resistance Genes Against <i>Fusarium graminearum</i> Infection in Maize.

Plants (Basel, Switzerland)·2026
Same author

Deep Learning-Based Identification of Pathogenicity Genes in <i>Phytophthora infestans</i> Using Time-Series Transcriptomics.

Plants (Basel, Switzerland)·2026
Same author

spaLLM: enhancing spatial domain analysis in multi-omics data through large language model integration.

Briefings in bioinformatics·2025
Same author

Codon Optimization Enables the Geneticin Resistance Gene to Be Applied Efficiently to the Genetic Manipulation of the Plant Pathogenic Fungus <i>Botrytis cinerea</i>.

Plants (Basel, Switzerland)·2024
Same author

A Relationship Prediction Method for <i>Magnaporthe oryzae</i>-Rice Multi-Omics Data Based on WGCNA and Graph Autoencoder.

Journal of fungi (Basel, Switzerland)·2023
Same author

<i>Brucella</i> activates the host RIDD pathway to subvert BLOS1-directed immune defense.

eLife·2022

Related Experiment Video

Updated: Oct 2, 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

1.8K

LHPE-nets: A lightweight 2D and 3D human pose estimation model with well-structural deep networks and multi-view pose

Hao Wang1,2, Ming-Hui Sun1,2, Hao Zhang1

  • 1College of Computer Science and Technology, Jilin University, Changchun, China.

Plos One
|February 23, 2022
PubMed
Summary

This study introduces LHPE-nets, a novel deep learning model for 3D human pose estimation that significantly reduces training time and improves accuracy. LHPE-nets offers faster training and enhanced joint estimation for efficient human pose analysis.

More Related Videos

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

9.5K
Integration of Animal Behavioral Assessment and Convolutional Neural Network to Study Wasabi-Alcohol Taste-Smell Interaction
06:19

Integration of Animal Behavioral Assessment and Convolutional Neural Network to Study Wasabi-Alcohol Taste-Smell Interaction

Published on: August 16, 2024

575

Related Experiment Videos

Last Updated: Oct 2, 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

1.8K
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

9.5K
Integration of Animal Behavioral Assessment and Convolutional Neural Network to Study Wasabi-Alcohol Taste-Smell Interaction
06:19

Integration of Animal Behavioral Assessment and Convolutional Neural Network to Study Wasabi-Alcohol Taste-Smell Interaction

Published on: August 16, 2024

575

Area of Science:

  • Computer Vision
  • Machine Learning
  • Artificial Intelligence

Background:

  • Multi-view fusion methods have advanced 3D human pose estimation, but the 2D pose estimation component incurs high training costs.
  • Existing deep learning networks for heat map generation in multi-view 2D pose estimation require substantial computational resources.

Purpose of the Study:

  • To investigate novel deep learning networks for efficient 2D pose estimation in cross-view 3D human pose estimation.
  • To develop a lightweight and faster-training deep learning model for 3D human pose estimation.

Main Methods:

  • Evaluated Mobilenetv2, Mobilenetv3, Efficientnetv2, and Resnet for 2D pose estimation.
  • Developed LHPE-nets, incorporating Low-Span and RDNS networks with optimized structures (evenly distributed channels, inverted residuals, external residual blocks, small-resolution sample processing).
  • Designed a static pose sample simplification method for efficient 3D pose data handling.

Main Results:

  • LHPE-nets trains 1-5 epochs faster than Resnet-34, demonstrating superior fast start-up and lightweight network characteristics.
  • Accuracy improved for approximately 60% of joints, with overall 3D human pose estimation exceeding other networks by over 7mm.
  • Detailed analysis confirmed LHPE-nets' efficiency in network size, training speed, and 2D/3D pose estimation performance.

Conclusions:

  • LHPE-nets offers a significant improvement in training efficiency and accuracy for 3D human pose estimation.
  • The model's lightweight design and faster training capabilities make it suitable for applications requiring rapid and accurate human pose analysis.
  • LHPE-nets achieves a high level of application performance compared to existing pose estimation models.