Scholarlink

The COVID-19 pandemic has created a global health crisis, and mask-wearing is recommended by the WHO to control the virus’s spread. Monitoring mask compliance in crowded areas is challenging. This study proposes an automated system using transfer learning with the ResNet-152V2 architecture to detect masked and unmasked faces in images and videos. The system also integrates other pre-trained models and the Haar classifier to enhance detection accuracy. The ResNet-152V2 model achieved a 99.5% detection accuracy, with significant improvements in precision and recall over existing methods. The advanced ResNet-152V2 architecture enhances mask detection accuracy and reliability, making the system effective for real-time video surveillance and aiding in COVID-19 prevention in public spaces.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Preventing Viral Infection with an Automated Face Mask Detection System Using Transfer Learning-Based Neural Networks

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Raman Kumar,
  • Rupali Gill,
  • Meenu Khurana,
  • Navdeep Kaur,
  • Shivani Pant,
  • Ahmad Radee Alawadi

摘要

The COVID-19 pandemic has created a global health crisis, and mask-wearing is recommended by the WHO to control the virus’s spread. Monitoring mask compliance in crowded areas is challenging. This study proposes an automated system using transfer learning with the ResNet-152V2 architecture to detect masked and unmasked faces in images and videos. The system also integrates other pre-trained models and the Haar classifier to enhance detection accuracy. The ResNet-152V2 model achieved a 99.5% detection accuracy, with significant improvements in precision and recall over existing methods. The advanced ResNet-152V2 architecture enhances mask detection accuracy and reliability, making the system effective for real-time video surveillance and aiding in COVID-19 prevention in public spaces.