Facial recognition plays a crucial role in our lives. The simple act of recognizing a face is the DNA of social interactions. Empowering devices to recognize faces will help us tap into a huge reservoir of opportunities with the potential of disrupting the way we interact with technology.
FaceSafe enables OEMs to go beyond a simple face unlock. By knowing who is using them their devices could make educated decisions thus elevating the user experience. This decision making, for example, could change the way the UI looks or could use personalized photography or videography acquisition and processing.
20 years of experience in computer vision are at the core of FaceSafe. This is powered by FotoNation’s convolutional neural network based Face Analytics pack, a solution that includes face detection and tracking, face features detection, face classification and emotion detection.
Face is compatible with structured light and time of flight systems.
Print attack, replay / video attack and 3D mask.
A face covered up to 20% is still detected.
Solution will activate only when the user is looking at the device.
In a fully digital world, people need state-of-the-art experiences combined with security and ID protection in the form of contactless biometric solutions.
FaceSafe offers rich user experience, security and reliability.
FotoNation’s FaceSafe can be used to unlock your phone or set up multiple user profiles, each with a different access level, look and feel. It can also be used to offer the first access to a more secure zone, such as a payment zone. The technology translates the visual information from the enrollment process into secured key data and uses that data for verification of each user. The user’s face is not stored as a 3D scan or as a picture.
FaceSafe offers state-of-the-art anti-spoofing and liveness detection capabilities.
A spoofing attack can be translated as an unauthorized attempt to access your device by someone that does not have rights or privileges to do it. The attack can be carried out by using a photo (print attack), a video (replay attack) or a 3D mask of the phone’s owner. Anti-spoofing measures demonstrate the face recognition solution’s robustness in such an attack.
Liveness detection, which includes contextual information analysis, texture analysis, user interaction and depth analysis, all play a role in the anti-spoofing measures.