Webcam emotion facial coding

Facial coding tracks the participants’ facial expressions through their webcams to measure their emotions. At each frame, an algorithm determines how likely it is that the participant is expressing each of the six basic and universal Eckman emotion facial codes.

The facial coding algorithm is derived from the cutting edge, peer-reviewed & benchmark-beating deep learning research: “Deep Learning the Dynamic Appearance and Shape of Facial Action Units” by¬†Shashank Jaiswal and Michel Valstar.

You can find a link to one of their key papers here.

After basic facial coding, we apply our own proprietary methods to clean and aggregate the data for analysis.