What is a truly realistic portrait? This question inspired a project about going beyond physical appearances and attempting to render the interior world of a human being, through memories, emotions, relationships and personal story. All of it organized around a portrait painting metaphor: brushes loaded with meanings instead of paint would render a portrait using my "generative video painting" techniques (vimeo.com/24065726).
Painting a Videorative portrait (a generative, narrative and interactive video portrait) starts with collecting personal videos of the person portrayed, tagged by him/her with relevant concepts and descriptions. Then, using a custom developed tool, the artist "paints with meanings" and generates a video portrait, subtitled with generative personal narratives. In the interactive installation version of the work, the viewer can "navigate" through the subject's mind, opening his/her video memories, accessing their thoughts and revealing hidden connections between the meanings, using real time access to Wikipedia to infer related emotional states. These "data visualizations" create new and unexpected interpretations of the portrait.
The subject portrayed can add more video memories, tags and descriptions to “increase the likeness” of the portrait over time, all his/her life, creating a never-ending collection of personal documentaries.
Videorative portraits are the result of my experiments to build a more “realistic” contemporary portrait of the physical and the psychological. A comment on the art of portrait painting. A visual metaphor of the memory, heavily distorted, chaotic, fragmented, obsessively replayed. A window to the subject's personal experience and intimate world. A picture of memories. A map of emotions.
My immense gratitude to filmmaker Randall Okita for accepting the challenge of exposing himself in these portraits.
All video sources, by Randall Okita (randallokita.com)
Music: "I Am a Man Who Will Fight for Your Honor", by Chris Zabriskie. (chriszabriskie.com)
Made with Processing, using GSVideo, generativedesign and controlP5 libraries.