International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 12 Issue: 05 | May 2025
p-ISSN: 2395-0072
www.irjet.net
Evo Haazri: A Dual Face Recognition System for Individual and Group Attendance Abhishek Kumar1, Pradeep Kumar2, Akash Sikarwar3, Akella Vandana4 Department of Physics and Computer Science Dayalbagh Educational Institute Deemed to be University, Dayalbagh Agra, Uttar Pradesh ------------------------------------------------------------------------***------------------------------------------------------------------------Abstract: Particularly in educational institutions, enterprises, and smart campuses, human inefficiencies, inaccurate data, and time-consuming procedures are common problems with attendance management systems. Such constraints may result in mistakes, decreased efficiency, and trouble keeping accurate attendance records. A state-of-the-art solution utilizing facial recognition technology is provided by EvoHaazri to tackle these problems. The application makes use of MongoDB for the safe management and storing of attendance data, TensorFlow for precise face recognition, and a smartphone camera for face capture. EvoHaazri reduces errors, saves time, and does away with manual intervention by simplifying the procedure. In order to guarantee scalability and adaptability across a variety of use cases, future improvements will include offline capabilities, group face recognition, and enhanced security measures. For contemporary attendance systems, EvoHaazri's user-friendly interface and sturdy design provide a dependable, effective, and convenient experience.
Keywords: Face Recognition, Group Attendance, React Native, Authentication, Mobile App, Tensorflow 1.
Introduction
Evo Haazri is a next-generation attendance management system that leverages the power of facial recognition technology to streamline and secure the process of marking presence. Designed for educational institutions and workplaces alike, Evo Haazri eliminates the need for traditional roll calls, ID cards, or manual registers. By simply scanning faces, the app ensures accurate, real-time attendance while reducing time and human error. Built using React Native, TensorFlow, and a MongoDB backend, it supports both single and group attendance modes. Evo Haazri stands out for its speed, accuracy, and ease of use, making it an ideal solution for modern-day attendance tracking. The development of Evo Haazri is rooted in the vision of enhancing operational efficiency through smart technology. Unlike conventional systems, Evo Haazri employs artificial intelligence to identify and verify faces with high precision. It works in diverse lighting conditions and is robust against common issues like photo manipulation or proxy attendance. With the capability to recognize multiple faces at once, it is especially suitable for classrooms and large gatherings. Data privacy is also a key concern; hence, the system is designed with secure storage and encrypted communication. The app’s intuitive interface ensures that users, regardless of technical skill, can operate it effortlessly. Keywords: Face Recognition Technology, Real-Time Data, Data Security, Group Face Recognition, Machine Learning, Database
2. Literature Review Over the past decade, significant research and development have taken place in the domain of biometric-based attendance systems, especially those using facial recognition technologies. Traditional attendance systems such as manual registers, RFID cards, and biometric fingerprint scanners have been found to be time-consuming, prone to human error, and vulnerable to manipulation. In contrast, face recognition offers a non-intrusive and contactless solution. According to Jain et al. (2011), facial biometrics are among the most accepted and scalable forms of authentication due to their uniqueness and the ease with which they can be captured without physical interaction. Several studies have explored the implementation of facial recognition in educational institutions and workplaces. For instance, Zhao and Chellappa (2012) emphasized that deep learning models significantly improve recognition accuracy in varied lighting and orientation. Furthermore, commercial solutions like Microsoft Azure Face API and Amazon Rekognition have demonstrated the feasibility of cloud-based face recognition, although concerns about data privacy and internet dependency remain. In response, researchers like Parkhi et al. (2015) proposed on-device models using deep
© 2025, IRJET
|
Impact Factor value: 8.315
|
ISO 9001:2008 Certified Journal
|
Page 329