The Portrait of
Christine of Sweden

an ARTificial portrait

A game of reflections, mirrors and projections, “Portrait of Cristina” is an encounter with the historical figure, and with her myth, through the aspects that reverberate in our contemporaneity and through the portrait in which to find her gaze on herself and the sign and trait of her time projected towards the future.
“Portrait of Christina” is a meeting, or rather it is a device for meetings.
“In my work of interpretation I placed myself in front of the painting as in front of a mirror. I immersed myself in an expression, I went through the fibers of the canvas and the material of the pictorial backgrounds until I sought the spirit of the portrait, the one that, together with the painter, Christina chose to represent herself, I looked for her voice – the sonorous one. and the existential one. And at the same time we ignited and dissolved the hieratic nature of the painted gaze into emotions and thoughts, making the painting learn through artificial intelligence to feed on softness, life, my performance. “
“To bring the portrait of Christina to life we ​​have chosen to use and, once again, experiment, a different approach.
We wanted Christina herself to tell her story, for her to be the face that the spectators would see moving and talking.
We therefore chose to create an image painted four hundred years ago, the portrait of Christina of Sweden.
To achieve this effect, we used sophisticated machine learning algorithms, trained to recognize faces through the reproduction and analysis of hundreds of video footage. These algorithms are therefore able to recognize the characteristic elements of a face, trace its movements and create a model that can then be applied to an image to create an animation that reproduces the original video as closely as possible.
We started from Gaia’s acting and interpretative work to create a video, then broken into 30-second fragments, which were analyzed by the system. Each fragment generated an mathematical model of the animation which was then applied to the painting.
The procedure, although made, in part, simpler by the artificial intelligence system, was extremely long and demanding. Each second of the final video required about 3 minutes of work between filming, computational analysis, animation creation and post production. The final video has a duration of 9 minutes. “
So for Urban Experience’s “Human Landscapes” we worked to detect and reveal a fragment of genius loci, making Christina of Sweden’s presence appear and reverberate in the cloister of San Salvatore in Lauro, one of the emblematic places of her presence in Rome. We projected the portrait and triggered a meeting device. A multidimensional meeting, visual and auditory.
“The people saw, while many others tuned in live on heard, open the door through which it is said that Cristina met Decio Azzolino and followed my body that invited them to cross the spaces until they see the projection of the talking portrait and its references, I then welcomed the image of Cristina on me, I became a screen and a medium for the last part that offered itself as yet another new beginning and we closed the meeting by making the space resonate of the cosmogonic words of Ovid’s Metamorphoses. “
We are therefore close the performance with the great beginning.
Massimo Di Leo and Gaia Riposati