Use the Demo App for Android
The SDK ships with a simple example application demonstrating how to set up an object tracker, how to feed it with video frames from a camera, and how to receive information about detected and recognized persons from the object tracker.
Note: This demo application requires an active SAFR Cloud account in order to function in online mode. Search for the USER_IDENTIFIER and PASSWORD words in the source and replace them with your user identifier and password, or enter credentials via sign-in dialog.
The demo app features include:
- Camera feed processing — Face tracking, face detection, and face recognition.
- Frame overlay is drawn on top of the tracked face; the level of details depends on current settings.
- Age, gender, and sentiment info are drawn on top of the overlay.
- Name and the face-quality metrics are drawn below the overlay.
- Frame color depends on the person's attributes.
- Heart icon is displayed when detected face is confirmed to be alive; a broken heart icon when detected face is confirmed not to be alive. (i.e. The subject failed an RGB liveness detection test.)
- Face image sharing — Tapping the tracked face displays the native share picker; the shared image includes the extracted face.
- Photo import — Face detection and face recognition.
- Selected photos are processed and displayed in another dialog.
- Frame overlays are drawn on top of the face.
- Frame color is set at random.
- The face-quality metrics are displayed in the bottom-right corner of the image.
- If recognized, the person's name is displayed below the frame.
- Image sharing — Clicking the share button displays the native share picker; the original image with rendered overlay is shared.
- Person (re)name — Tap the face to display the rename dialog.
- Configuration.
- License - Dialog to enter the SDK license provided by RealNetworks. (Note that RGB liveness detection won't work without the SDK license.)
- Sign-in — Allows entering SAFR credentials. Pressing Skip automatically transitions to Offline mode.
- Camera facing - Front or back camera.
- Face recognizer settings — Select or clear to enable or disable specific features:
- Identity recognition.
- Age recognition.
- Gender recognition.
- Mask recognition.
- Occlusion recognition.
- Sentiment recognition.
- If all the face recognizer setting features are unchecked, the SDK automatically transitions to Offline mode.
- Detector input size.
- RGB liveness detector.
- Minimum required face size.
- Liveness detection threshold.
- Fake detection threshold.
- Detection scheme.
- Frames evaluation count.
- Minimum confirmations required.
- Event reporter.
- Event reporter switch.
- RGB Liveness action - When enabled, events will be generated indicating detection of liveness for specific recognized persons.
- User interface.
- Show face info/quality attributes.
- Background service for RTSP stream processing - Enabled from the code.
- Run RTSP camera stream - Enabled from the code.
See Also