By bringing together in one place the largest number of Leonardo da Vinci works ever assembled, The Louvre organized a real event exhibition. For this unique occasion, Accenture wished to make a real gift to its privileged customers by inviting them to a private night opening, where they had the opportunity to experiment installations that combine art and creative technology.
We designed and developed a creation experience which had as its subject La Cène and wanted to allow users to paint the famous painting from the sketch phase to the final work, not with brushes but with their brain waves. We made it by means of a helmet which allows the measurement of the concentration. In order to create the visual, we developed a custom application to extract the painting Data.
The piece is first analysed using a Corner-Harris algorithm in order to extract the “features” of the image, then a triangulated shape is generated from the given set of points. This shape represents the first level of revelation. The second level of the revelation is created by another triangulated shape in 3D, its level of elevation is set from a noise effector and each triangle color is based on the median color beneath it. The final composition combines the three levels such as the stroke triangulated shapes, the 3D triangulated shapes and the final piece.
In order to measure concentration we used a Muse EEG helmet wich is capable of giving us in realtime the value of the 5 principal brainwaves but not the concentration rate, only raw data that we have to interpret.
After reading a lot of scientific papers about EEG and concentration levels, we first noticed that most cases now use deep learning to do that, which means that each person has first to do a training session before being able to start the experience. Using AI wasn’t possible because in an event context we don’t have time to train the software each time a new user comes. That being said, by experience and research, we understood that analysing the variations of the Alpha and Beta waves could give us a good approximation of the concentration rate of a person, and that’s what we finally did.
To give more impact to the visual, the shape is placed in a 3D space using a physical based rendering engine. By this method we can add more contrast to the final composition by using more accurate lights and reflections models. We also retrieved the real-life material sensation of the piece, which is painted on a wall, by adding a micro-surface using a normal map texture.
Please switch your device orientation to continue.