Documentation for Face PSIM 1.0.1.

Previous page Video camera mounting and setup requirements for STC recognition module  Configuring the frames skip of the STC recognition module Next page

The STC recognition module is configured as follows:

  1. Activate the STC recognition module (see Activation of the recognition modules in Face PSIM).
  2. Go to the settings panel of the STC recognition module object.
  3. From the Select profile drop-down list (1), select the face recognition profile:
    • Dataset cleaner − "heavy" profile, highest quality, but very slow.
    • Default − base profile. If you specify a profile other than Default, then the parameters of the basic profile are partially replaced by the parameters from the selected profile.
    • Kamaz − very fast profile, but low quality.
    • Nist − contains a limited set of available algorithms.
    • Onepass − profile that is commonly used for remote authentication tasks. It is a simple (cascading) face detector. The recognition algorithm is fast, but has a low recognition accuracy.
    • Profile wiki − not a profile, but a set of basic blocks for various algorithms.
    • Progressive − compromise profile, usually showing higher quality than others, but a little slower.
    • Safe city − profile for a safe city. Works with faces from 120 pixels. The recognition algorithm is not the fastest, but more accurate.
    • Stadium ix − profile for stadiums. Works with faces from 180 pixels. The recognition algorithm is not the fastest, but more accurate.
    • Stadium xeon − profile for stadiums. In version 4.2, it is similar to the Stadium ix profile.
    • Transport safety − profile for transport safety. Works with faces from 100 pixels. The recognition algorithm is faster, but has lower accuracy.
    • Turnstile − profile for turnstiles. Works with faces from 300 pixels. The recognition algorithm is not the fastest, but more accurate.
  4. Click the Apply button (2).

Configuring the STC recognition module is complete.

  • No labels