Real-time classroom attention tracking using computer vision. Monitor student focus, detect gaze direction, identify drowsiness, and improve learning outcomes.
A full suite of computer vision and AI features built for modern classrooms.
Computer vision models track where each student is looking — board, screen, phone, or away — at 30 fps.
Eye aspect ratio analysis detects prolonged eye closure and blink patterns indicating fatigue.
6-DoF head pose estimation identifies whether students are facing the instructor or looking elsewhere.
Smart alerts notify instructors when students are distracted for longer than configurable thresholds.
Aggregate dashboards show class-wide attention trends, peak engagement windows, and learning insights.
All processing runs on-device. No video is stored — only anonymized attention metrics are retained.
From raw camera frames to actionable classroom insights in under 50ms.
A single wide-angle classroom camera feeds into the CV pipeline.
YOLO-based face detector locates each student's face with 98%+ accuracy.
Landmark detection extracts 468 facial keypoints per face.
A lightweight MLP fuses gaze, pose, and blink into a 0–100 score.
Scores stream to the teacher dashboard in real time.
Live simulation of a classroom with 20 students. Attention scores update in real time.
Integrate with your favorite video conferencing platforms
ClassroomAct helps educators understand how students engage with their lessons — and act on that data in real time.