Intel RealSense ID Solution F450, Datasheet

Page: 22 / 38

Software Definition
22 Rev: 001
4 Software Definition
The Intel® RealSense™ ID Solution F450 module executes in a closed system running
a real time operating system. The face authentication uses image pre-processing,
anti-spoofing checking, and the SOC matrix acceleration engine embedded to extract
the face features of the user. These face features are then stored during enrollment
of a new user or compared during authentication to determine whether the user is
valid.
4.1 Intel® RealSense™ ID SDK for Facial
Authentication
The Intel® RealSense™ ID SDK for Facial Authentication is a library providing Facial
Authentication capabilities to applications using the Intel® RealSense™ ID Solution
F450 or F455. It is open source and available on
https://github.com/IntelRealSense/RealSenseID
4.2 Face Authentication
The Intel® RealSense™ ID Solution F450 optical system and the face authentication
algorithms are optimized for functional and performance profiles targeted at access
control devices. In this use case, one or more users will walk up to the device in the
runtime or standby state, the host will notify the
Intel® RealSense™ ID Solution F450
module to perform face authentication and the Intel® RealSense™ ID Solution F450
module will respond back the result. The host can then decide how to feedback status
and result back to the end user.
Table 4-1: Functional Metrics
METRIC VALUE METRIC
Number of Users 1000 Users
Authentication Range 55 to 100 cm
Anti-Spoofing Range 55 to 100 cm
Ambient Light 5 to 50k* (up to 1.5m) lux
Ambient Light Dark 0 (up to 0.8 – 1.0m) lux
Ambient Temperature 0 to 50 o C
(*) Reduced performance may be exhibited in semi-shade or direct sunlight.
Table 4-2: Face Authentication Metrics
METRIC DESCRIPTION VALUE
False Acceptance Rate (FAR) Recognize someone else as you. 1:1M
True Acceptance Rate (TAR) Recognize you as you. 99.7%