Face selection is the most important task to achieve healthy identities that yield accurate recognition results.
To help with this, Nirovision classifies faces at the server level based on quality metrics such as pose, sharpness, brightness and pixel variance, in addition to size. Faces that meet our quality standards are classified as High Quality, whereas those that fall below any of our minimum thresholds are classified as Low Quality.
Once you've paired a camera to Nirovision, you can access the calibration tool to get real-time feedback about the faces the camera is detecting. This tool will help you assess the need to reposition a camera or adjust its settings, by pushing one frame per second and displaying the quality values for every face in frame.
<aside> ⚠️ Video coming soon!
</aside>