What happens when a system tries to understand you by looking at your face?
IRIS is an interactive installation that explores the relationship between observation, identity, and the systems we use to recognize people. Using real time body and facial tracking, IRIS observes its participant, maps their features, and gradually breaks their face down into individual fragments.
The system then attempts to reconstruct what it has seen.
At first, the process feels precise. Features are detected, organized, and placed into position, as if a person could be understood simply by collecting enough information about them. But as the reconstruction progresses, things begin to fall apart. Eyes shift from where they should be. Features are duplicated. Fragments move, reorganize, and form something that is recognizable, but never quite right.
The work takes its name from the iris, the part of the eye that makes each person visually distinct, while also controlling how we see. For IRIS, this becomes a metaphor for the act of looking itself. The system can see a person, measure them, and create a representation of them, but seeing does not necessarily mean understanding.
Rather than aiming for a perfect digital portrait, IRIS is interested in the moment when the reconstruction begins to fail.
Because perhaps a person cannot be reduced to the pieces that a system can see.
CAMERA ACCESS DENIED
IRIS needs your camera to see you. Nothing is recorded or uploaded.