Cryfe™ is software that uses artificial intelligence techniques to interpret signals from a person’s body language, facial expression and speech from a video. These different signals are then combined to identify a person’s internal states (comfort, discomfort, etc.), or more generally, their authenticity. Using state-of-the-art artificial intelligence technologies, Cryfe detects and highlights all verbal and visual signals emitted by a recorded (filmed) speaker (participant). It analyses all facial expressions, gestures, comfort and discomfort actions or changes in tone or energy in the voice. With Cryfe, look back on reported moments of congruence (the alignment between what I say, think and do) or incongruence (when authenticity is in doubt) and make informed decisions.
The support received by the STI should allow us to optimize the machine learning part, to strengthen Cryfe’s algorithms and to add new signals to detect.