Drowsiness detection
Eye-state monitoring detects prolonged eye closure and can trigger an alert before a study session turns into a nap.
FocusGuard turns your webcam into a privacy-first AI study companion. It watches for drowsiness, phone distractions and time away from your desk while keeping the experience simple and local.
FocusGuard combines computer vision, live session tracking and lightweight analytics without turning studying into another complicated app.
Eye-state monitoring detects prolonged eye closure and can trigger an alert before a study session turns into a nap.
FocusGuard identifies visible phones during a session and can immediately warn you when the device becomes a distraction.
Detects longer periods away from the camera while avoiding instant alerts from brief head movement.
Converts focused and distracted time into an easy-to-read session score, so progress is measurable rather than vague.
Choose your own session duration and track focused time, distracted time, remaining time and alerts in one place.
Use built-in warning beeps or choose your own sounds for phone, drowsiness and away alerts.
Choose how long you want to focus and start the monitoring session.
Your webcam feed is analysed for focus-related signals in real time.
End the session and review your focus time, distractions and score.
FocusGuard is designed around local webcam processing. The study-monitoring experience does not need a cloud video stream or an online account to work. Your focus tool stays focused on you—not on collecting your camera footage.
FocusGuard was created as a desktop AI project that combines computer vision with a modern study-session experience. Instead of adding more productivity clutter, it focuses on three common problems: drowsiness, phone distraction and leaving the desk.
The current release is built for Windows and uses local AI models to provide live feedback while you study.
The desktop build is being prepared for release. The website is ready for GitHub Pages and a future Windows download button.