Competition project
SmartClass AI
Real-time classroom engagement analysis on an ESP32-CAM, with no cloud dependency.
Bronze Medal · WICE 2026
Overview
An IoT and AI classroom monitor. An ESP32-CAM captures the room, computer vision estimates student emotion and engagement, and live metrics reach a web dashboard and a teacher’s OLED display.
The problem
Teachers can’t see engagement across a whole room in real time, and cloud video analysis is slow, costly and a privacy risk.
What I built
Real-time emotion and engagement analysis with OpenCV and DeepFace, running at the edge.
Reworked the ESP32-CAM to PC image transfer, removing a 10-second capture delay for near-real-time analysis.
Live metrics pushed to a web dashboard and an on-desk OLED display.
Outcome
Bronze Medal at the WICE 2026 national round, and selected for the international round in Malaysia.