Geo Face Attendance System

AI-Powered Attendance Management using Face Recognition & Geolocation

ADITYA YADAV

B.TECH COMPUTER SCIENCE ENGINEERING

The Problem with Traditional Attendance

Manual attendance systems are error-prone, time-consuming, and easily exploited. Modern institutions need smarter solutions.

❌ Traditional Attendance

• Proxy attendance is easy to exploit
• Manual record-keeping is slow and error-prone
• No location or identity verification
• Time-consuming for faculty
• Human errors in data entry

✅ Smart Attendance

• Face recognition prevents proxy attendance
• Automated digital records eliminate errors
• GPS geolocation confirms presence
• Instant, real-time data capture
• Secure, tamper-proof database

Proposed Solution

A unified system combining biometric identity verification with real-time location checks — ensuring every attendance record is authentic and tamper-proof.

Face Verification

Biometric match via face-api.js

Attendance Saved

Secure MongoDB record stored

User Login

Authenticate with credentials

Location Verification

GPS geolocation on-site check

Each step adds a layer of security, ensuring only verified, on-site students can mark attendance.

Face Recognition

GPS Verification

JWT Auth

Secure Database

Admin Dashboard

Student Dashboard

System Architecture

How It All Connects

The frontend communicates with the backend through secure RESTful API calls. Face recognition and geolocation are validated client-side before submission, reducing server load and improving response times.

All API communication is protected via JWT tokens, ensuring only authenticated users can submit attendance.

Technology Stack

Frontend

• React (Vite)
• Axios
• HTML & CSS
• Face-api.js

Backend

• Node.js
• Express.js
• JWT
• bcrypt.js

Database

• MongoDB Atlas
• Mongoose ODM

Deployment

• Render (Backend)
• Vercel (Frontend)
• GitHub (Version Control)

Key Features

Every feature is designed to ensure accuracy, security, and a seamless user experience.

Face Recognition

AI-powered biometric verification prevents proxy attendance.

Geolocation Verification

GPS checks confirm students are physically on campus.

Secure Login

JWT tokens and bcrypt hashing protect user credentials.

Admin Dashboard

Full control over attendance records, users, and reports.

Student Dashboard

Personal attendance history and real-time status updates.

Responsive Design

Works seamlessly across desktop, tablet, and mobile devices.

Project Workflow

A clear, step-by-step journey from registration to verified attendance — fully automated and secure.

Each stage is validated before the next begins, ensuring data integrity at every step of the pipeline.

Challenges & Solutions

⚡ Challenges Faced

Face Detection Accuracy

Varying lighting and angles affected recognition rates.

MongoDB Atlas Connection

Initial connection strings and whitelist configuration caused delays.

JWT Authentication

Token expiry and refresh logic required careful middleware design.

CORS & Camera Permissions

Cross-origin requests and browser camera access needed explicit handling.

Deployment

Environment variables and build configuration differed across platforms.

✅ Solutions Implemented

Face-api.js Models

Pre-trained TinyFaceDetector improved speed and accuracy.

MongoDB Atlas

Cloud connection with IP whitelisting resolved connectivity issues.

Express Middleware

Custom JWT middleware handled token validation and refresh cleanly.

Axios Interceptors

Centralized CORS and permission handling at the API layer.

Environment Variables

Separate .env configs for Render and Vercel ensured smooth deployment.

Results & Future Scope

🏆 Project Results

• Secure, tamper-proof attendance records
• Fast face verification under 2 seconds
• Fully cloud-deployed on Render & Vercel
• Responsive UI across all devices
• RESTful API with documented endpoints

🚀 Future Enhancements

• AI Liveness Detection to prevent photo spoofing
• QR Code fallback attendance mode
• Native Mobile Application (React Native)
• Email & SMS attendance notifications
• Advanced attendance analytics dashboard
• AI-generated performance reports

2s

Verification Time

3

Security Layers

100%

Cloud Deployed

0

Proxy Cases

Thank You

This project demonstrates how modern web technologies — face recognition, geolocation, and cloud infrastructure — can solve real-world institutional challenges.

🔗 GitHub Repository

Full source code available for review

☁️ Live Deployment

Backend on Render · Frontend on Vercel

⚙️ MERN Stack

MongoDB · Express · React · Node.js

💬 Q&A

Open for questions and feedback

Aditya Yadav · B.Tech Computer Science Engineering · Geo Face Attendance System