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 Videos

An FPGA-based IP for recognizing violence against children.

Muhammad Alhammami1, Samir Marwan Hammami2

  • 1Higher Institute for Applied Sciences and Technology, Damascus, Syria.

Methodsx
|August 25, 2021
PubMed
Summary

This paper introduces a novel FPGA-based IP for detecting child abuse using only skeleton data, preserving privacy. The system achieves high accuracy and speed for real-world deployment in schools.

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

Vision-based system model for detecting violence against children.

MethodsX·2020
Same author

Violent actions against children.

Data in brief·2017
See all related articles

Area of Science:

  • Computer Science
  • Artificial Intelligence
  • Hardware Engineering

Background:

  • Child abuse detection is crucial for safeguarding vulnerable populations.
  • Existing methods often raise privacy concerns due to image data usage.
  • There is a need for privacy-preserving, efficient detection systems.

Purpose of the Study:

  • To develop a novel FPGA-based Intellectual Property (IP) for recognizing common violent actions against children (VACR IP).
  • To ensure privacy by utilizing only skeleton joint data, avoiding the use of actual images.
  • To create a deployable system for educational and therapeutic settings.

Main Methods:

  • Implementation of a custom FPGA-based IP core.
  • Utilization of skeleton joint data as the sole input for action recognition.
Keywords:
Action recognitionChildren abuse detectionFPGA-based hardware IPFog-based systemMachine learning

Related Experiment Videos

  • Hardware acceleration for real-time processing and low latency.

    • Main Results:

    • Achieved a high detection rate of 97.72% for violent actions against children.
    • Demonstrated a processing speed of 761 Frames Per Second (FPS).
    • Exhibited a low latency of 2.79 milliseconds with a 50MHz system clock.

    Conclusions:

    • The developed FPGA-based VACR IP is the first of its kind to detect child abuse using only skeleton data, ensuring privacy.
    • The IP can accurately identify the presence and type of violent actions.
    • The system is suitable for integration into complete solutions for deployment in schools, aiding psychologists and counselors.