auFacecount
auFaceCount™ is an AI-based application designed to count people in front of a screen at a distance of 0,2 to 3 meters for more than 2 seconds. The system detects faces by recognizing features such as eye, eyebrow, and nose, and it functions reliably even when the person is not facing the screen directly up to 45 degrees of head rotation.
The application has been tested in real-world environments using both mobile screen setups (Gallery robot) and stationary displays. Test results show that mobile screens receive five times more registered face views, and potentially up to ten times more visual impressions, compared to stationary screens. This is attributed to the robot’s movement, which attracts more attention and increases visibility
To ensure data reliability and privacy compliance, auFaceCount has been validated against Amazon Web Services’ facial recognition tool. This validation confirms that auFaceCount applies stricter criteria for facial recognition, ensuring that only individuals who actively engage with the screen are counted. This approach enhances data security and minimizes false positives, making the system suitable for GDPR-compliant analytics in public spaces.
auFaceCount is particularly well-suited for retail environments, exhibitions, and smart spaces where accurate and privacy-conscious audience measurement is essential.