What if we could encourage people to make silly faces?
Skills
Physical Computing
Prototyping
Computer Vision
Tools
FaceOSC
Processing
Projector
Team / Instructors
Yase Dusu
Wayne Kim
Jacob Remin
Dennis Paul
Duration
4 days, 2024
Overview
Mirrored Canvas turns facial expressions and voice into live visual art projected on a translucent canvas. The canvas acts as both mirror and stage letting participants play and make faces.
How It Works
The system runs in real time using:
faceOSC for facial tracking — measuring x/y coordinates of key points like mouth height and eyebrow position
Processing for visual rendering and interaction logic
Webcam + microphone for multimodal input
Projection onto a translucent acrylic canvas
Each tracked expression maps to an output:
Smile → a photo of a joyful memory
Shock → abstract, reactive visuals
Absuridity can be contagious
We began bodystorming in front of mirrors: exaggerated faces, strange sounds, awkward movements. When the system felt serious, people froze. When it felt absurd, they leaned in and laughed. That shaped both the interaction and the visual language.
It held up in testing: watching prototypes, laughter spread before people even tried it themselves, curiosity pulled them in. The experience became less about introspection, more about shared permission to be loud, awkward, unpolished.
Reflection
Mirrored Canvas creates space for expression without judgment. By prioritizing play over precision, the installation invited movement, laughter, and curiosity. Emotion was not captured or classified. It was performed, shared, and briefly made visible before fading away.
making faces with Yase