AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
The Constraints of Calibration
We have more than half a million of data points that we've collected in our office over the past three years to just feed the facial muscle activity. The same model ended up being applied to the eye movement tracking, for example. That's exactly the same model. You were always looking at the same type of signal, which is electrical signal and then it's just a matter of how do you find the patterns within this signal? Yeah. How much variation is there between you and me and somebody else randomly picked off the street? So actually, not that much. It's much less variation than what you would have for voice recognition.