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

Iterative Dynamic Generic Learning for Face Recognition From a Contaminated Single-Sample Per Person.

Meng Pang, Yiu-Ming Cheung, Qiquan Shi

    IEEE Transactions on Neural Networks and Learning Systems
    |April 21, 2020
    PubMed
    Summary
    This summary is machine-generated.

    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

    PI-H2T: Enhancing Long-Tailed Visual Recognition with Permutation-Invariant and Head-to-Tail Feature Fusion.

    IEEE transactions on pattern analysis and machine intelligence·2026
    Same author

    Development of a risk stratification tool for rapidly progressive diabetic retinopathy in type 2 diabetes.

    Frontiers in endocrinology·2026
    Same author

    STG-DB net for athlete pose estimation and pose rational optimization in complex sports scenes.

    Scientific reports·2026
    Same author

    Stage-specific disruption of erythropoiesis leads to anemia in newly diagnosed multiple myeloma patients.

    Frontiers in cell and developmental biology·2026
    Same author

    Implementing trust in non-small cell lung cancer diagnosis with a conformalized uncertainty-aware AI framework.

    Nature biomedical engineering·2026
    Same author

    Structured bigel-based foams with tailored textural properties: a bigel-based strategy for cream replacement.

    Current research in food science·2026

    This study introduces Iterative Dynamic Generic Learning (IDGL) to improve face recognition accuracy when enrolment data is contaminated by variations like poor lighting or disguises. IDGL enhances prototype representation for robust single-sample face recognition (SSPP FR) systems.

    Area of Science:

    • Computer Vision
    • Machine Learning
    • Biometrics

    Background:

    • Single-sample per person face recognition (SSPP FR) faces challenges with contaminated enrolment databases (SSPP-ce FR).
    • Nuisance variations (poor lighting, disguises) degrade performance of generic learning methods using the prototype plus variation (P+V) model.
    • Existing methods fail due to inaccurate prototypes and inadequate variation dictionaries from contaminated data.

    Purpose of the Study:

    • To address the limitations of SSPP-ce FR by proposing a novel Iterative Dynamic Generic Learning (IDGL) method.
    • To enhance the accuracy of face recognition in the presence of significant variations in enrolment data.

    Main Methods:

    • IDGL employs a dynamic label feedback network utilizing both labeled enrolment and unlabeled query data.

    Related Experiment Videos

  • It uses a semisupervised low-rank representation (SSLRR) framework to recover accurate prototypes from contaminated samples.
  • A representative variation dictionary is learned by extracting sample-specific corruptions, iteratively refining prototypes and improving label estimation.

    • Main Results:

    • The proposed IDGL method significantly improves accuracy in SSPP-ce FR tasks.
    • Iterative refinement of prototypes and label estimation leads to enhanced recognition performance.
    • Experiments on benchmark datasets demonstrate IDGL's superiority over existing state-of-the-art methods.

    Conclusions:

    • IDGL offers a robust solution for single-sample face recognition with contaminated enrolment databases.
    • The dynamic learning approach effectively handles nuisance variations, improving overall system reliability.
    • This method advances the field of biometrics by enabling more accurate face recognition under challenging real-world conditions.