DRIVE FaceID

Vehicle Security

 

Driver Authentication

Personalized Settings

 

Attendance tracking

Based on patented deep learning algorithms

What are the hardware requirements?

  • EDGE3 DRIVE is a hardware agnostic platform that can run on ARM or X86 platforms running Android, Linux, or Windows operating systems.

How does the system handle a new (unidentified) user?

  • Once the system identifies that the user is new, DRIVE queries the user for their name to authenticate and begin training.

How many faces are theoretically possible to ID?

  • This solution is architected in such a way that it can scale to service large commercial fleets, or individual families.

How much time does it take to train one face?

  • Training is an on-going, ever improving process, so there is not really a notion of training time in this application.
  • On average, initial training is completed in less than one minute.

What is the processing time to ID?

  • Very fast. Milliseconds

How many images of a face needed to train?

  • A few hundred images are generally needed to get an initial ID. However, because the human face is always changing, training is an ongoing process that is never complete.

How much of the face is needed to properly train?

  • The face doesn’t have to be positioned precisely in front of the camera. EDGE3 DRIVE can capture and use images from many angles.  Over time as more images are captured from the subject, an improved identifier can be created.

Are there lighting constraints?

  • No, the technology works in all lighting conditions: sunlight (direct or indirect), shadows, night, etc.