Camera, Gesture Tracking and Calibration Notice

Last updated: 16 September 2026 This is the specific privacy and safety notice for FaceCards camera-based tracking and calibration. It is intended to be available before or when a user enables the relevant functions. FaceCards uses computer vision to translate deliberate gestures into configured controller actions. The camera feature is not an identity-recognition system. 1. What the camera is used for The camera provides live video frames so FaceCards can estimate facial, eye, head and hand landmarks/features and decide whether a configured gesture has occurred. Camera access is used for gesture control and calibration, not for recording study sessions. 2. Raw camera frames stay local Raw camera frames and live camera video are intended to be processed on the device. The FaceCards backend is not designed to receive or store raw camera video, face photographs, eye images or continuous facial-landmark history. Live frames and transient landmark/feature values may exist briefly in memory while tracking is active. They are discarded as the live processing advances rather than retained as a camera history. 3. Derived calibration can be synchronised When you save or sync a calibration, FaceCards may send derived measurements to your account so the calibration can be reused. These may include neutral eye/head baselines, noise values, gesture thresholds, axis multipliers, gesture measurement/confidence information, tracking quality, camera/device labels and timestamps. Derived calibration is not a photograph, but because it is linked to your account it is treated as personal data in our Privacy Policy. 4. No biometric identity purpose FaceCards does not use the calibration profile to identify you, verify your legal identity, unlock your account by face recognition or compare you against a database of faces. It is used to tune gesture recognition for your controller. If this purpose changes in the future, we will carry out a fresh legal/privacy assessment and update the notices before using calibration data for identity recognition. 5. MediaPipe FaceCards uses Google MediaPipe Tasks and related vision models/runtime assets. Google’s current MediaPipe Tasks Privacy Notice states that input data such as images and video is processed on-device and is not sent to Google servers. It also states that MediaPipe Tasks APIs send performance and utilisation metrics to Google for purposes including measuring usage/performance, debugging, maintenance and improvement. FaceCards does not intentionally configure MediaPipe to transmit raw camera frames, study content or saved calibration profiles to Google. Google states that developers are responsible for obtaining informed consent to Google’s processing of MediaPipe metrics where applicable law requires it; any required consent is separate from the ordinary device camera permission. 6. Camera permission and active state You must grant camera permission before camera-based tracking can start. Operating systems and browsers may display their own camera indicator. FaceCards is designed to provide an obvious active state and a stop control. Stopping tracking should stop camera capture and input dispatch for that session. If a device or browser fails to release the camera correctly, close FaceCards or revoke camera permission through the operating-system/browser settings. 7. Calibration accuracy Calibration measures how your movements appear to a particular camera in particular conditions. It can become less accurate if the camera, position, lighting, glasses, distance, posture or device changes. A “good” or similar quality label is a technical quality indicator, not a guarantee that every gesture will be recognised correctly. 8. Accidental and missed gestures Computer vision can produce false positives and false negatives. FaceCards includes hold, cooldown, neutral/re-arm and calibration controls intended to reduce accidental triggers, but they cannot eliminate them. Test mappings in a low-risk context before using them with another application. 9. Unsafe contexts Do not use camera gesture control to operate vehicles, machinery, medical devices, clinical systems, emergency controls, weapons, financial-trading or payment confirmation interfaces, security-critical administration, irreversible data-deletion controls, or any other high-consequence system where an unintended gesture could cause serious harm. 10. Saved calibration management FaceCards currently allows saved calibrations to be listed, renamed and deleted. Account storage is currently limited to the most recent 50 profiles. Local copies may also exist on individual devices and can require device-level clearing or uninstalling. 11. Age restriction FaceCards is intended only for users aged 18 or over. People under 18 must not create an account or use the camera-based controller. If we become aware of under-18 use, we may restrict access and handle associated personal data as described in the Privacy Policy and Age Eligibility and Under-18 Privacy Notice. 12. Questions Camera/calibration privacy or safety questions: facecardssupport@gmail.com