Emotion Tracking
Reading the face
Identifying facial expressions
In this real world example, you see that we keep track of the head and then detect the emotion based on the facial expression. We measure all 7 universal core emotions simultaneously and then detect which one has the highest likelihood. This way we don’t miss anything, not even microexpressions.
Micro expression detection
Micro expressions are very short facial expressions that last less than 500 milliseconds. These facial expressions are so special because they can not be controlled or suppressed due to their speed.
That also means that they normally escape the human eye but our system detects them.
No Problem.
How good is it?
The accuracy of our emotion tracking algorithm compared to human annotators is 90%. Plus, our system doesn’t get tired and can analyse a video faster than any human could.
Specs
| Accuracy | 90% |
| Sampling frequency | 30 fps |
Developed with ♥ in The Netherlands
contact@mind-trace.com
