FaceEngine settings

This section describes the FaceEngine parameters that are used to configure FaceStream.

Configuration of FaceEngine parameters is performed only by editing the configuration file "faceengine.conf". The configuration file is mounted into the FaceStream container when it is launched (see the FaceStream launch command in the installation manual).

In this section you can set the settings for face and body recognition. It is recommended to change the parameters in this configuration only in consultation with VisionLabs employees.

FaceDetV3

This parameter sets the settings for the face detector.

Note. This section is used only when the "use-face-detector" parameter is enabled. When working with bodies, parameters from the "HumanDetector" section will be used.

minFaceSize

The parameter sets the scaling factor of the source frame before processing.

Images are resized by minFaceSize/20 times. For example, if the value is minFaceSize=50, then the image size will be resized by 2.5 times (minFaceSize=50/20).

For more information about scaling, see "Frame scaling".

HumanDetector

This section sets the settings for the face and body association detector.

Note. This section is used only when the "use-body-detector" parameter is enabled. When working with faces, parameters from the "FaceDetV3" section will be used.

ImageSize

The parameter sets the maximum frame size after scaling along the largest side of the frame.

This parameter works similarly to the "scale-result-size" parameter from the TrackEngine settings.

For more information about scaling, see "Frame scaling".