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

Image Super-Resolution as a Defense Against Adversarial Attacks.

Aamir Mustafa, Salman H Khan, Munawar Hayat

    IEEE Transactions on Image Processing : a Publication of the IEEE Signal Processing Society
    |September 24, 2019
    PubMed
    Summary
    This summary is machine-generated.

    Related Concept Videos

    Super-resolution Fluorescence Microscopy01:37

    Super-resolution Fluorescence Microscopy

    12.2K
    Super-resolution fluorescence microscopy (SRFM) provides a better resolution than conventional fluorescence microscopy by reducing the point spread function (PSF). PSF is the light intensity distribution from a point that causes it to appear blurred. Due to PSF, each fluorescing point appears bigger than its actual size, and it is the PSF interference of nearby fluorophores that causes the blurred image. Various approaches to achieving higher resolution through SRFM have recently been...
    12.2K

    You might also read

    Related Articles

    Articles linked to this work by shared authors, journal, and citation graph.

    Sort by
    Same author

    Impact of adjuvant breast radiotherapy on the risk and the survival of second primary lung cancer: a large population-based study.

    Japanese journal of clinical oncology·2026
    Same author

    Recent Progress of Single-Ion Conducting Polymer Electrolytes for Rechargeable Mono- and Multivalent Cation-Based Metal Batteries.

    Angewandte Chemie (International ed. in English)·2026
    Same author

    Topology-Preserving Deep Hashing for Ultrafast Drone-Dominated Object Detection.

    IEEE transactions on neural networks and learning systems·2026
    Same author

    High-throughput screening of EGFR/Ca<sup>2+</sup> signaling modulators in cardiac hypertrophy using a tetrahedral DNA nanostructure-based hESC platform.

    Journal of pharmaceutical analysis·2026
    Same author

    Development and Application of a LAMP Assay for Detecting E198A-Type MBC-Resistant <i>Clarireedia monteithiana</i>.

    Plant disease·2026
    Same author

    Concept Drift and Long-Tailed Distribution in Fine-Grained Visual Categorization: Benchmark and Method.

    IEEE transactions on pattern analysis and machine intelligence·2026

    This study introduces an efficient image enhancement method to defend Convolutional Neural Networks (CNNs) against adversarial attacks. The approach restores adversarial images, enhancing security without retraining models.

    Area of Science:

    • Computer Vision
    • Machine Learning Security

    Background:

    • Convolutional Neural Networks (CNNs) excel in computer vision but are vulnerable to adversarial attacks.
    • These attacks use subtle noise to mislead CNNs, limiting their use in security-critical applications.

    Purpose of the Study:

    • Propose a computationally efficient image enhancement method for robust defense against adversarial perturbations.
    • Develop a defense that enhances image quality and maintains performance on clean images without altering classifiers.

    Main Methods:

    • Utilize deep image restoration networks to map adversarial samples back to the natural image manifold.
    • Employ a novel approach that does not require classifier modification or adversarial detection mechanisms.

    Main Results:

    Related Experiment Videos

    • Demonstrated effective mitigation of adversarial perturbations in gray-box settings.
    • Achieved robustness against various attack algorithms while preserving performance on clean images.

    Conclusions:

    • The proposed method offers a simple, training-free, and model-agnostic defense against adversarial attacks.
    • This approach enhances CNN security by restoring image integrity and classification accuracy.