Audience Measurement
Measure viewing time and gaze-based attention around a display, with facial expression estimates. Build the results into your own analytics.
Gaze & attention · Full source code
Gaze Tracker Engine
Estimate head pose and eye gaze direction from a webcam, with facial landmarks and eye openness measurements.
$325 USD · Digital source package
Checkout and downloads via Payhip. Applicable taxes calculated at checkout. Commercial license; no support or custom integration included.
ImplementationC++ / OpenCV
Browser demoWebAssembly
DeliveryDigital download
▶ Watch the recorded demonstration
Video loads only when you choose to play it.
What the engine does
- Detect and track faces with associated IDs.
- Locate facial and eye landmarks.
- Estimate head direction and eye gaze vectors.
- Estimate eye openness and apply filtering for smoother output.
Where it fits
For developers exploring gaze-based interaction, attention analytics or research prototypes. The engine provides a foundation you can adapt to your own geometry and interaction model.
What to validate before you buy
The output is a gaze direction estimate, not a calibrated screen coordinate. Mapping gaze to a screen or physical target requires your own calibration and integration. Validate accuracy for camera distance, pose and lighting.
Inside the download
- Complete C++ engine source
- Pretrained networks and required assets
- Code documentation
- Emscripten build instructions
- Precompiled OpenCV WebAssembly module
- Guide to building OpenCV for WebAssembly
Inspect the public code reference to understand the implementation before purchasing.
Integration essentials
Check your build environment.
- Required skills
- C++ programming, OpenCV integration and basic understanding of neural networks.
- Browser build
- Emscripten compiles the engine to WebAssembly. Follow the included instructions and use the supplied OpenCV module.
- Other platforms
- Desktop, server, mobile and edge builds require the appropriate OpenCV dependency and build configuration. Porting and platform-specific optimization are your responsibility.
- Performance
- Depends on camera resolution, face count, models, hardware and browser. Evaluate representative conditions; the browser demo is not a benchmark for every device.
License & handover
Clear terms from the start.
You may study and modify the code, integrate it into your own product that adds substantial value, and sell that product under the software license.
The supplied source must remain private. You may not resell the software as a standalone library, SDK or toolkit. Third-party components retain their own licenses.
Installation, integration, consulting, guaranteed fixes and ongoing maintenance are not included. Updates are discretionary. Review the support policy before checkout.
Digital sales are final except where required by law. Read the purchase FAQ.
Related engines
Explore adjacent capabilities.
Each engine is a separate purchase. Combining them requires integration.
Facial Expression Recognition
Classify visible facial expressions in a camera feed and explore the estimates in your own visualization or analytics.
Liveness Detection
Add action-based liveness checks to your face verification flow. Ask for blinks or head turns and check the response.