The first experiment revolves around the question: Can we objectively quantify people’s experience of a ride in the Autonomous Vehicle Lab?
As a we study to what extent we can relate people’s facial expressions to situations on the road, such as a sudden acceleration, bumps in the road or sharp turns. To carry out the research, we put a small smart camera in the car to get a good look at the passenger’s face. We try to quantify the experienced emotion of passengers and relate it to comfort characteristics of the ride. The analysis takes place via edge computing directly on the camera, so that we do not need to store or send privacy-sensitive data. This guarantees the privacy of passengers. The camera reduces faces to a number of “facial landmarks”, a point cloud that maps all the movable parts of the face. We will develop analysis methods to quantify the expressions in emotional valence (positive – neutral – negative) and arousal (‘activation level’ as it were).
Status: Start June 2021
Partners: Leaseplan, Corporate Innovation (Valorisation Centre – TU Delft), Smart Vehicle lab (CITG), Urban Mobility Centre (CITG)