People can recognize each other by looking at each other's faces. Besides enabling recognition of a person, faces can convey attributes such as age, gender and even emotions.
In our research, we create technologies to safely and conveniently using facial attributes. Attributes are useful for facial recognition, but changes in expression, the direction that a person is facing, and ambient lighting is all factors that increase the difficulty of recognition. To solve these problems, we use three-dimensional (3D) facial representations.
3D facial representations allow us to create virtualized faces whose appearances change under different conditions. By determining what features change under different conditions, unique individualistic changes can be identified.
A specialized 3D scanner is usually used to create 3D facial representations. However, we have pioneered a method to create 3D representations from two dimensional images. This technology incorporates models of human faces and knowledge about shadows created by uneven surfaces such as the eyes and nose, to quickly construct detailed 3D facial representations.
3D facial representations are used in the Walkthrough Face Recognition system, the first domestic walk-thru-type face recognition system to use 3D features. This system can use images from a security camera, and delivers secure access control with the convenience of recognition while walking, without any need to slow down or swipe cards. This system can recognize multiple faces in an image, easing congestion at entrances or elevators during peak hours.
Our facial imagery analysis can also estimate attributes such as age, gender, the presence of glasses, or purposely obstructed faces.
So what are these attributes good for? Attributes can be used in cases where a child is lost in a shopping mall or someone is purposely covering their face etc. By knowing these attributes, incident response times can be reduced, or incidents may even be preempted altogether.
By combining our 3D facial representation technology with upcoming research into tracking and behavior recognition, we aim to be able to automatically recognize the "who," "where," and "what" in an image.
At sporting venues and other large scale events, security has traditionally required large amounts of manpower, but automated recognition technologies are posed to help organizers by becoming a force multiplier. At SECOM, we continue to conduct field evaluations of facial recognition technologies and other advanced security technologies at events such as the 2017 Tokyo Marathon.