From 10 Minutes to 10 Seconds: AI Smart Attendance Using Group Photo Recognition | EHazri Hackathon скачать в хорошем качестве

From 10 Minutes to 10 Seconds: AI Smart Attendance Using Group Photo Recognition | EHazri Hackathon

EHazri – AI Smart Attendance System for Rural Schools | Hackathon Demo In many rural schools, attendance is still recorded manually using paper registers. This process takes 5–10 minutes every class, is prone to human error, and provides no real-time visibility to parents, which can increase dropout risk. Existing automated solutions like biometric or RFID systems are often expensive, require dedicated hardware, stable internet, and continuous power, and are difficult to maintain in rural environments. Because of this, they fail to be affordable, scalable, and user-friendly for everyday classroom use. To address this, we built EHazri — an offline-first AI-powered attendance system designed to be low-cost, simple, and fast. Teachers capture a single group photo, and our model uses face detection and face embeddings to recognize students and mark attendance within seconds. The data then syncs to a central dashboard enabling parent alerts, attendance analytics, and early dropout-risk identification. EHazri transforms a slow manual task into a smart, scalable digital process, helping make attendance more efficient, affordable, and accessible for rural education. Key Features Offline attendance with mobile app AI group photo auto-attendance using face recognition Admin dashboard with analytics and reports Parent SMS/WhatsApp alerts Early dropout-risk detection Tech Stack: Python, OpenCV, TensorFlow, Flutter, Kotlin, Firebase, SQLite, Figma Vision: Building a complete AI-powered smart school ecosystem with real-time dropout prediction, performance tracking, and government integration.