Prefigures - BFI Luminous
September 2022
A hybrid performance that seamlessly intertwines the movements of a dance trio on stage with BFI's national archive of dance footage, reanimated through a state-of-the art AI model for depth estimation. Developed in collaboration with Alexander Whitley Dance Company and presented at BFI’s Luminous event.
Description
This hybrid piece marries archive film, live performers and digital avatars to create an interactive link between past and future representations of the body in the moving image.
We worked with dance-related footage from BFI's national archive which we processed using a state-of-the-art Neural Network for the estimation of depth information from monocular video. This AI model enabled us to reconstruct the full three-dimensional scenes shown in the archival material, giving a relief-like 3D depth to the 2D moving imagery.
In these reconstructed worlds the digital avatars of the live dancers coexist with the performers of the BFI films, and interact with the footage that surrounds them introducing turbulent flows of energy within a bespoke fluid simulation. The resulting visuals were projected onto translucent screens in front of three live dancers, whose choreography was, in part, led by the movements of those long past dancers on screen, breathing life into the archive, and creating a new contemporary narrative.