Open navigation

Facial Recognition - Best Practices


Best Practices

We recommend that you first upload a person's photo before using the face recognition feature. This photo, which serves as a reference image for recognizing faces of this person in images, should show the person's face and shoulders from the front, like the following example image.


Video: Using Facial Recognition



Further Information

Here are some other items to consider when selecting a reference image. 

More information on the functionality of our Facial Recognition system can be found here as well.


Metric  Fail Condition

Reason

Confidence c<98 A face must be used.
Quality
(brightness & sharpness)  
s<60 || b<30Use images that are bright and sharp.
Pose
(pitch, yaw && roll)
p<-30 || p>45
|| y<-45 || y>45
Use an image with a face that is within the recommended range of angles. The pitch should be less than 30 degrees face down and less than 45 degrees face up. The yaw should be less than 45 degrees in either direction. There is no restriction on the roll.
Eyes Open e>90 Use an image of a face with both eyes open and visible
Bounding Box
(left, top, width & height, all in ratio)
l<0.01 || t<0.01
|| l + w>0.99 || t + h>0.99
The image should contain the full head and shoulders of the person. It should not be cropped to the face bounding box.
w*h<0.01Use an image of a face that occupies a large proportion of the image. Images with a face that occupies a larger portion of the image, are matched with greater accuracy.
Emotion
(Happy: h; Calm: c; Other emotions)

{emotion not happy not calm} >90 || h<10 && c<10 

Use images of faces with neutral facial expressions with mouth closed with little to no smile for applications that require high precision.
Resolution
width & height in pixel
Width x Height <4000 Ensure that images are sufficiently large enough with respect to resolution.



Did you find it helpful? Yes No

Send feedback
Sorry we couldn't be helpful. Help us improve this article with your feedback.