Scholarlink

Effective attendance monitoring in educational institutions is crucial for shaping academic performance, particularly as student enrollment rises. Conventional methods, reliant on manual recording and paper-based sign-in sheets, are proving cumbersome and time-consuming. Existing automated systems, including mobile applications, RFID, Bluetooth, and fingerprint models, face challenges related to inaccuracies and inefficiencies. Hence these limitations can be overcome by adopting a smart attendance monitoring system utilizing facial recognition technology, offering a non-intrusive and promising solution. The proposed work introduces a time-integrated model that systematically records the attendance status periodically at regular intervals of time throughout lectures thus enhancing accuracy. A recommended multi-camera system ensures comprehensive coverage of the classroom. An experiment has already been conducted by considering the smart attendance monitoring system in a specific class at the University of KwaZulu-Natal (UKZN), based on facial recognition and the outcome clearly exhibited a reliable average accuracy rate of 98%, presenting a robust and efficient solution to address challenges posed by traditional attendance methods for academic institutions aiming to enhance attendance monitoring practices.

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

Multi-facial Recognition Using Multi-task Cascaded Convolutional Neural Networks

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Jayasri B. S,
  • A. Harshit Joshi,
  • Prateek Kumar Choubey,
  • Ritesh Bagati,
  • Sarthak Kumar Singh

摘要

Effective attendance monitoring in educational institutions is crucial for shaping academic performance, particularly as student enrollment rises. Conventional methods, reliant on manual recording and paper-based sign-in sheets, are proving cumbersome and time-consuming. Existing automated systems, including mobile applications, RFID, Bluetooth, and fingerprint models, face challenges related to inaccuracies and inefficiencies. Hence these limitations can be overcome by adopting a smart attendance monitoring system utilizing facial recognition technology, offering a non-intrusive and promising solution. The proposed work introduces a time-integrated model that systematically records the attendance status periodically at regular intervals of time throughout lectures thus enhancing accuracy. A recommended multi-camera system ensures comprehensive coverage of the classroom. An experiment has already been conducted by considering the smart attendance monitoring system in a specific class at the University of KwaZulu-Natal (UKZN), based on facial recognition and the outcome clearly exhibited a reliable average accuracy rate of 98%, presenting a robust and efficient solution to address challenges posed by traditional attendance methods for academic institutions aiming to enhance attendance monitoring practices.