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

Diffusion models enable high-fidelity prediction of fuel cell impedance spectrum from short time-domain profiles.

Nature communications·2026
Same author

Robust reconstruction of equivalent source method based near-field acoustic holography using an alternative regularization parameter determination approach.

The Journal of the Acoustical Society of America·2019
Same author

Urantide improves the structure and function of right ventricle as determined by echocardiography in monocrotaline-induced pulmonary hypertension rat model.

Clinical rheumatology·2018
Same author

A Tunable Molten-Salt Route for Scalable Synthesis of Ultrathin Amorphous Carbon Nanosheets as High-Performance Anode Materials for Lithium-Ion Batteries.

ACS applied materials & interfaces·2018
Same author

Apatinib alone or combined with radiotherapy in metastatic prostate cancer: Results from a pilot, multicenter study.

Oncotarget·2018
Same author

Expression of PFKFB3 and Ki67 in lung adenocarcinomas and targeting PFKFB3 as a therapeutic strategy.

Molecular and cellular biochemistry·2018

Related Experiment Video

Updated: Jul 20, 2025

A Methodology for Capturing Joint Visual Attention Using Mobile Eye-Trackers
12:39

A Methodology for Capturing Joint Visual Attention Using Mobile Eye-Trackers

Published on: January 18, 2020

7.7K

Single-view multi-human pose estimation by attentive cross-dimension matching.

Wei Tian1, Zhong Gao1, Dayi Tan1

  • 1Institute of Intelligent Vehicles, School of Automotive Studies, Tongji University, Shanghai, China.

Frontiers in Neuroscience
|August 4, 2023
PubMed
Summary

This study introduces a novel k-block architecture for multi-person pose estimation, improving accuracy by using the entire heatmap and the SMPL model. The method enhances real-time human pose estimation in complex environments.

Keywords:
attentive learningcross-dimension matchingkeypoint predictionmulti-person pose estimationsingle-image pose estimation

More Related Videos

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
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

3.8K

Related Experiment Videos

Last Updated: Jul 20, 2025

A Methodology for Capturing Joint Visual Attention Using Mobile Eye-Trackers
12:39

A Methodology for Capturing Joint Visual Attention Using Mobile Eye-Trackers

Published on: January 18, 2020

7.7K
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
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

3.8K

Area of Science:

  • Computer Vision
  • Machine Learning
  • Human-Computer Interaction

Background:

  • Vision-based human pose estimation is crucial for AR, action recognition, and HCI.
  • Keypoint detection methods often waste information by focusing on small heatmap regions, limiting optimization.
  • Existing methods struggle with information utilization and model optimization.

Purpose of the Study:

  • To develop an advanced multi-person pose estimation architecture.
  • To improve keypoint estimation accuracy and efficiency.
  • To enable real-time pose estimation in complex scenarios.

Main Methods:

  • Designed a novel k-block architecture with a voting mechanism on the entire heatmap.
  • Integrated the SMPL 3D human body model for iterative pose correction.
  • Utilized information from the whole heatmap for simultaneous keypoint and uncertainty inference.

Main Results:

  • Achieved state-of-the-art performance on the 3DPW dataset.
  • Improved Mean Per Joint Position Error (MPJPE) by approximately 8 mm.
  • Improved Procrustes-aligned Mean Per Joint Position Error (PA-MPJPE) by approximately 5 mm.

Conclusions:

  • The proposed architecture effectively addresses information waste in pose estimation.
  • The integration of the SMPL model enhances pose accuracy by leveraging human body structure.
  • The real-time capability opens new avenues for multi-person pose estimation in complex, real-world applications.