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

A multi-class framework for face mask compliance detection using lightweight deep learning models.

Balraj E1, Manikandan P2, Sambath M2

  • 1School of Computer Science and Engineering, Vellore Institute of Technology, Chennai, India. balraj.e@vit.ac.in.

Scientific Reports
|May 19, 2026
PubMed
Summary

Related Concept Videos

Force Classification01:22

Force Classification

Forces play a crucial role in the study of physics and engineering. They are essential in describing the motion, behavior, and equilibrium of objects in the physical world. Forces can be classified based on their origin, type, and direction of action.
Contact and non-contact forces are two of the most widely used categories of forces. As the name suggests, contact forces require physical contact between two objects to act upon each other. Examples of contact forces include frictional,...
Masking and Demasking Agents01:19

Masking and Demasking Agents

EDTA titrations may necessitate masking and demasking agents to temporarily protect a particular metal ion in a mixture from the EDTA reaction. These agents facilitate the sequential analysis of the metal ions by forming stable complexes with some—but not all—metal ions during certain steps.
There are many masking agents, such as cyanide, fluoride, triethanolamine, thiourea, and 2,3-bis(sulfanyl)propan-1-ol (formerly 2,3-dimercapto-1-propanol), with the masking agent chosen based on the metal...

You might also read

Related Articles

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

Sort by
Same author

Ethics and bias in emotional AI.

Frontiers in artificial intelligence·2026
Same author

Effectiveness of Low-Level Laser Therapy Combined With Eccentric Exercise in Treating Midportion Achilles Tendinopathy: A Randomized Controlled Trial.

Cureus·2024
Same author

The Accuracy of Casts Obtained Using Different Impression Techniques and Impression Materials in Combined Parallel and Angulated Implants: An In Vitro Study.

Cureus·2024
Same author

Optimized Deep Learning for the Classification of Parkinson's Disease Based on Voice Features.

Critical reviews in biomedical engineering·2023
Same author

Identification of Multi-Potent Protein Subtilisin A from halophilic bacterium Bacillus firmus VE2.

Microbial pathogenesis·2021

This study introduces a three-class face mask compliance detection system using MobileNetV3, achieving 98.90% accuracy. The advanced model effectively identifies correct, improper, and no mask usage, outperforming traditional methods in surveillance environments.

Area of Science:

  • Computer Science
  • Artificial Intelligence
  • Machine Learning

Background:

  • Face mask compliance is crucial for public health surveillance.
  • Traditional binary mask detection methods often overlook nuanced compliance levels.
  • Developing robust systems for multi-class mask detection is essential for health-sensitive environments.

Purpose of the Study:

  • To develop and evaluate a multi-class face mask compliance detection system.
  • To address the limitations of binary classification by including 'improper mask use'.
  • To compare the proposed system's performance against existing state-of-the-art models.

Main Methods:

  • A Deep Convolutional Neural Network (DCNN) based on MobileNetV3 was employed.
  • The model was adapted with a lightweight fully connected layer and Squeeze-and-Excitation (SE) blocks.
Keywords:
COVID-19DCNNDeep LearningMobileNet V3Multi-Class Mask DetectionPublic Health MonitoringSqueeze and Excitation Block

Related Experiment Videos

  • A balanced, custom-curated dataset of 2,079 images was used with a 70:15:15 train-validation-test split.

    • Main Results:

    • The MobileNetV3 system achieved 98.90% accuracy and an F1-score of 0.989.
    • The three-class system (correct, improper, no mask) demonstrated superior performance.
    • The model proved competitive or superior to MobileNetV2, VGG16, and YOLO-based models.

    Conclusions:

    • The proposed MobileNetV3 system offers a highly accurate solution for multi-class face mask compliance monitoring.
    • The inclusion of improper mask use enhances the system's practical applicability in surveillance.
    • Future work includes real-time monitoring and explainable AI integration for improved implementation.