We analyze emotions not in discrete categories, but as a continuous space of expressions that can be visualized as a disk.
We measure 3 emotional attributes or dimensions:
In addition to the above emotion attributes, we measure several other useful parameters, including:
Head pose helps determine whether a person is looking at the camera. In combination, the above metrics can be used to determine the attention or engagement of people with the camera or screen.
Our algorithms have been trained with a database comprising hundreds of thousands of faces and expressions. Based on extensive testing, we are able to achieve an accuracy of above 90 percent for emotion recognition.