MUMBAI, India, Aug. 31 -- Intellectual Property India has published a patent application (202621081983 A) filed by Symbiosis International Deemed University on July 02, 2026, for Web-Based Automated Student Attendance System Using Face Recognition With Confidence-Gated Ensemble Classification.
Inventors include Dr. Shreyas Rajendra Hole; Ayushi Nasare; Tabish Sharose Mansoori; and Shubhanshu Manjare.
The application for the patent was published on August 28, 2026, under issue no. 35/2026.
Abstract: ABSTRACT WEB-BASED AUTOMATED STUDENT ATTENDANCE SYSTEM USING FACE RECOGNITION WITH CONFIDENCE-GATED ENSEMBLE CLASSIFICATION A web-based automated student attendance system (100) employing a three-stage machine learning pipeline for face recognition on standard CPU hardware. The system comprises a face detection module (110) using a Haar Cascade Classifier to localize faces in webcam frames, a feature extraction module (120) generating normalized 1,024-dimensional pixel-based feature vectors from detected face regions, and a classification module (130) using a Random Forest ensemble classifier with a plurality of decision trees for student identity prediction through majority voting. A confidence gate module (140) applies a configurable threshold to the classifier's native probability estimation to reject uncertain predictions, thereby preventing false attendance entries without a separate calibration model. The system further includes a web application server (150), an enrolment module (160) for browser-based webcam training image capture with automatic model retraining, a relational database module (170) for timestamped attendance logging, and a reporting module (180) for attendance filtering and CSV export. [
Disclaimer: Curated by HT Syndication.
