In a 3-room parkour an artifical portrait is generated.
To read more about the real experiment we developed in relation to this installation see EEG-workshop.
To see 3D footage from the installation, visit our blog.
In room 1, "face-recognition" the visitor is instructed to have their photo taken using software based on Chinese facial recognition.
In room 2, an artist is hooked up to a professional 32-channel EEG headset that measures the visual reaction of the artist to real photographic or synthetically generated portraits via EEG studying the image-data process.
In room 3, a generative portrait of the last visitor is created with the help of a deep neural network system. In a gamified "result" the visitor gets a social score based on how fast the portrait was generated. The network constructs the portrait “from scratch”: the photo of the visitor is not used directly (as in the case of image processing), but only for comparison. The network thus takes on the role of the portraitist.