In our extensive testing, we have observed a false negative rate of less than 1% and a false positive identification rate of less than 0.1%

  • A false negative is when the system fails to recognize an enrolled user.
  • A false positive is when the system misidentifies an enrolled user as a different user.