Poor face quality

In the context of analytics report appearing imprecise, face quality is a real concern you must have. Face quality is clarity of a face within the camera, a stat that represent how available are the facial features of the visitor. If they are not available, the quality drops, and the precision/reliability of your analytics report drops.
 

Symptoms of poor face quality:

• Visitor is not detected - numbers go down.
• Double-counting - Multiple visitor IDs assigned per one visitor, numbers go up.
• Misgendering - visitor appear as members of the opposite gender.
• Increased amount of contacts - many contacts with little contact duration

• Contact duration drop.

   

How to solve poor face quality

1. Read the Camera Installation and Positioning guide.
2. Check camera resolution (above 960x720 pixels) and FPS (above 25FPS, preferably 30 or higher)
3. Check camera positioning - distance not exceeding 4 meters, upright angle. 
4. Check lighting conditions - avoid backlights, avoid natural light, aim for artificial lights in favor of the face of your customers, illuminating their faces, and not backs.