Jono Brandel
Des Objets Reconnus, (2019)
On the right side, hundreds of famous oil paintings cycle in and out. On the left are keywords perceived by an object recognition algorithm used by many of the top technology corporations in the world. As the keywords change, the computer distorts the oil painting to reveal the part of the painting that triggers hidden metadata. The methodical pace runs at one half of one percent the speed the computer takes to recognize objects in an image. This expanded view frames the computational process in terms conducive to human perception. In addition, the wispy motion of the paintings is directly tied to how uncertain the algorithm is. Lastly, a sonic layer matches predominant colors present in the painting to underscore the pace. Des Objets Reconnus transcodes oil paintings through object recognition in order to juxtapose representation and meaning. In this way, Des Objets Reconnus questions our personal relationships with paintings as much as it questions state-of-the art object recognition algorithms.
Back To The Gallery