Attendance Management System Using Face Detection
Year: 2025
Type: article
Abstract: This paper presents a Hybrid Multi-Stage Face Detection Algorithm that integrates traditional and deep learning methods for improved accuracy and efficiency. The process begins with Preprocessing and Enhancement to refine image quality. Fast Face Candidate Selection (Haar + HOG + SVM) quickly detects potential faces, followed by Precise Localization using MTCNN to refine detections and extract facial landmarks. Deep Learning Verification (RetinaFace/YOLO) eliminates false positives, ensuring reliability. Finally, Face Tracking (Kalman Filter + SORT) maintains consistency in video streams. This approach provides a robust and adaptable solution for real-world face detection applications. Keywords : Face Detection, Hybrid Algorithm, Deep Learning, Haar Cascade, HOG + SVM, MTCNN, RentinaFace, YOLO, Face Tracking, Real-Time Processing
DOI: 10.55041/ijsrem43195
Authors: Subash E, Hariprasath M, Aathithya S, Kalaiselvan M, Narendran R, Dinesh S, Krishna Nandhini B
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4408949805