FaceGuard

About this app

FaceGuard is an Android home-security pilot that gives phones separate Guarding and Watching roles. A Guarding phone runs the camera and on-device face pipeline, compares detected faces with enrolled household members, and records reviewable alerts when configured policy identifies an unknown visitor. A Watching phone can pair using a join code or QR setup flow and receive alerts from the home.

Alert details can include the event time, recognition evidence, authenticity and quality status, delivery state and reporting device. Face embeddings and sensitive setup credentials use local protected storage where supported. Remote pairing and delivery require a configured FaceGuard server and network access.

FaceGuard is an awareness aid, not a certified alarm, access-control or emergency-response system. Recognition accuracy varies with lighting, camera position, image quality and device performance.

Usage reporting: while FaceGuard is visible and connected, it sends a random app-specific installation ID, app version and whether camera guarding is healthy. Watching phones report guarding as inactive. This reporting excludes camera, household, alert and location content. Offline use cannot appear live. Optional feedback appears after tapping Stop following at least 60 continuous seconds of healthy visible camera analysis. Comments are private to the publisher unless you explicitly consent to public display; ratings contribute to the public average. Queued offline feedback retries while the app is open. Installation counts and app use are self-reported, not unique people or verified downloads.

What’s new

Optional once-per-version feedback after stopping a successful visible guarding session. Ratings contribute to the store average; comments remain private unless explicitly shared. Offline feedback retries while the app is open. The publisher dashboard shows visible live installations and healthy camera guarding, using minimal app-scoped reports every 30 seconds and a 90-second expiry.