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

MFAN: Multi-Level Features Attention Network for Fake Certificate Image Detection.

Yu Sun1,2, Rongrong Ni1,2, Yao Zhao1,2

  • 1Institute of Information Science, Beijing Jiaotong University, Beijing 100044, China.

Entropy (Basel, Switzerland)
|January 21, 2022
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

Brucella-infected macrophage-derived microparticles enhance the host anti-Brucella acquired immune response.

Veterinary microbiology·2026
Same author

PDGFC facilitates enzalutamide resistance in prostate cancer through activation of the Rap1-MAPK pathway.

Journal of cancer research and clinical oncology·2025
Same author

Brucella-infected macrophage-derived Exosomes enhance the anti-Brucella acquired immune response.

International immunopharmacology·2025
Same author

The CTSZ-TRA2A-IL32 axis defines a targetable macrophage-dependent pathway in metastatic prostate cancer.

Journal of translational medicine·2025
Same author

LS-PRISM: A layer-selective pruning method via low-rank approximation and sparsification for efficient large language model compression.

Neural networks : the official journal of the International Neural Network Society·2025
Same author

Smart healthcare in mobile hospitals: enhancing response to Disease X.

BMC infectious diseases·2025

This study introduces a new forgery detection method for certificate images, the Multi-level Feature Attention Network (MFAN). MFAN effectively identifies diverse manipulations in fake certificates, improving image forensics.

Area of Science:

  • Digital Image Forensics
  • Computer Vision

Background:

  • Most image forensics methods focus on natural images.
  • Certificate images, crucial for rights and interests, are vulnerable to forgery.
  • Fake certificates exhibit varied tampered region scales and manipulation types.

Purpose of the Study:

  • To develop a novel forgery detection method for certificate images.
  • To address the challenges of variable tampered region scales and diverse manipulation types in fake certificates.
  • To enhance the application of image forensics technology to high-stakes documents.

Main Methods:

  • A Multi-level Feature Attention Network (MFAN) with an encoder-decoder structure was proposed.
  • The encoder utilizes Atrous Spatial Pyramid Pooling (ASPP) and low-level feature concatenation for multi-scale feature extraction.
Keywords:
certificate imagefeature recalibrationimage forensicsmulti-level features

Related Experiment Videos

  • Channel-wise recalibration of multi-level features suppresses irrelevant information and enhances tampered regions.

    • Main Results:

    • The proposed MFAN method demonstrated superior performance in detecting forgeries in certificate images.
    • Experimental results showed effectiveness against diverse manipulation traces and varying tampered region scales.
    • MFAN outperformed existing state-of-the-art image forensics methods.

    Conclusions:

    • The Multi-level Feature Attention Network (MFAN) is a robust and effective method for certificate image forgery detection.
    • The method successfully addresses key challenges in detecting sophisticated forgeries.
    • This work advances image forensics for critical document authentication.