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 Experiment Video

Updated: Nov 18, 2025

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

792

An Attention-Enhanced Multi-Scale and Dual Sign Language Recognition Network Based on a Graph Convolution Network.

Lu Meng1, Ronghui Li1

  • 1College of Information Science and Engineering, Northeastern University, Shenyang 110000, China.

Sensors (Basel, Switzerland)
|February 10, 2021
PubMed
Summary

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

Retrieval-augmented generation for medication safety: A case study using drug package inserts.

International journal of medical informatics·2026
Same author

Injectable Chito-oligosaccharide-hyaluronic acid hydrogels with Fe<sup>3+</sup>/AMP nano-enzyme promote ROS scavenging and immunomodulation for myocardial infarction repair.

Journal of advanced research·2026
Same author

Radiation induces an EGFR-dependent lactate-epigenetic program underlying synergy with EGFR inhibition.

Cellular oncology (Dordrecht, Netherlands)·2026
Same author

Clinical characteristics and treatment of low-dose methotrexate-related adverse reactions in rheumatoid arthritis: a retrospective study.

Naunyn-Schmiedeberg's archives of pharmacology·2026
Same author

Clinical characteristics, treatment, and outcomes of Oxaliplatin-induced immune thrombocytopenia.

Frontiers in immunology·2026
Same author

Exercise-regulated bone-brain axis: A novel perspective on age-related cognitive decline.

Neural regeneration research·2026

This study introduces a novel multi-scale and dual sign language recognition network (SLR-Net) using graph convolutional networks. The proposed SLR-Net achieves high accuracy in recognizing sign language from skeleton data, overcoming limitations of previous methods.

Area of Science:

  • Computer Vision
  • Artificial Intelligence
  • Human-Computer Interaction

Background:

  • Sign language is crucial for hearing-impaired communication.
  • Existing sign language recognition methods struggle with accuracy due to factors like occlusion and varied signing styles.

Purpose of the Study:

  • To develop an advanced sign language recognition system overcoming previous accuracy limitations.
  • To enhance understanding of sign language for broader communication.

Main Methods:

  • Proposed a multi-scale and dual sign language recognition network (SLR-Net) utilizing graph convolutional networks (GCN).
  • Extracted skeleton data from RGB videos for recognition.
  • Developed three key sub-modules: multi-scale attention network (MSA), multi-scale spatiotemporal attention network (MSSTA), and attention-enhanced temporal convolution network (ATCN).
Keywords:
GCNattention mechanismkeyframes extractionlarge-vocabularysign language recognition

More Related Videos

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

650

Related Experiment Videos

Last Updated: Nov 18, 2025

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

792
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

650
  • Introduced a keyframe extraction algorithm for improved efficiency.

    • Main Results:

    • Achieved 98.08% accuracy on the CSL-500 dataset (500-word vocabulary).
    • Reached 64.57% accuracy on the challenging DEVISIGN-L dataset (2000-word vocabulary).
    • Outperformed existing state-of-the-art sign language recognition methods.

    Conclusions:

    • The SLR-Net demonstrates superior performance in sign language recognition.
    • The proposed attention mechanisms and network architecture enhance robustness and accuracy.
    • The keyframe extraction algorithm offers a practical trade-off between efficiency and accuracy.