Select Images for Training

Add one or more training images to each face that you want to recognize.

Micro Focus recommends using images that are representative of how you will use face recognition. For example, if you intend to use face recognition in a situation that is not controlled for facial expression, add several images to each face showing different facial expressions.

It is better to use fewer high-quality images, rather than many low-quality images, because low-quality images can reduce accuracy.

A good training image for a face:

  • should contain only one face. An image that contains multiple faces is accepted only if one of the faces covers more than double the image area of the others. The largest face is used to train Media Server and the others are ignored.
  • must be rotated correctly (the face cannot be upside down).
  • contains a face that is wider than approximately 150 pixels; the face should constitute a large proportion of the image.
  • contains a front view of the face, so that both eyes are visible.
  • is not blurry.
  • has even illumination, because dark shadows or bright highlights on the face reduce recognition accuracy.
  • has not been converted from another file format to JPEG. Converting images to JPEG introduces compression artifacts. If you have JPEG images, avoid making many rounds of edits. Each time that a new version of the image is saved as a JPEG, the image quality degrades.

  • shows a face that is not partially obscured (for example a person standing behind a microphone).