Planet Biometrics took time out to speak to Mikael Fagerlund, Director Sales EMEA at Cognitec Systems, about the facial recognition company’s truly innovative new products.
Q: Can give us an overview of Cognitec’s new product for biometric photo capture?
A: Calls for biometric Entry/Exit systems around the world led us to develop a high-end device for capturing ISO-compliant facial images during border control processes. We needed to realize three main features: speed, high usability, and optimal image capture.We incorporated proven features used in our verification technologies for eGates. The camera position adjusts in real time according to person height, an animated face marker on the feedback screen shows the user where to move his face, and active LED lighting provides optimal illumination. All these features are packed into a super slim, light-weight device that can be installed on booths, counters, walls, or in free-standing pillars. Aside from border crossing scenarios, the device can support any application that needs to capture biometric ID photos.
Q: Can you also talk about the new face finder and matching algorithm your company has been working on?A: The past year posed a new challenge to all automated face recognition vendors: virus protection masks covering large parts of the face. Our previous face finding technology showed inconsistent detection rates when significant areas of the face were “deleted” by a mask. The accuracy of facial recognition systems depends on finding faces in images and successfully enrolling them, or, in other words, successfully creating a template. It proved imperative to advance our technology for face finding tasks, with or without masks. In addition, we updated our matching algorithm to meet the mask challenge, arriving at a new version that works for images with or without facial masks with almost equal accuracy.
Q: What other emerging trends has your research and development team been working on?
A: We have been able to further improve our age estimation technology, especially for younger and older people, and for faces of all ages with masks. Stores, casinos and bars, for example, are looking for touch-free methods to perform an initial underage check, either for age 18 or 21. We are continuously optimizing this algorithm to achieve higher accuracies within that age range, so users will receive fewer false alerts.
Q: What impact has COVID-19 had on your business strategy?
A: As long as wearing masks is required or common in everyday life, recognizing masked faces is becoming an expected feature for systems that unlock phones, track and count people, measure demographics, or identify persons of interest. Business and entertainment venues need to return to normal operations while implementing new methods to stay safe and healthy. Many are looking toward face recognition for touch-free access control and mask detection technologies. Airports and other travel services that had considered face recognition to solve growing passenger numbers are now evaluating contactless, “seamless journey” solutions for hygiene reasons, for decreasing the time people spend at the airport, and for reducing their contact with other humans. While our products already cater to these use cases and markets, we are adding features to address the current safety concerns and requirements. For example, we have created a special mask detection algorithm to identify people either wearing a mask or if the mask is missing. The software can alert staff in real-time to a person without the required facial covering. And in general, we will continue to work on the accuracy of our algorithms to ensure reliable face matching or facial analysis results for faces with virus protection masks and other occlusions.
