Image & Model

Mimesis OS

Series
Model Face
Classification
Interactive installation; modified Polaroid camera, Raspberry Pi, CRT monitor, thermal printer, custom web software
Dimensions
Dimension variable
Ref.
RC-15
Date
2026

A Polaroid camera sits on an old CRT monitor.

Record

Inside its shell, a Raspberry Pi reads each photo locally, pixel by pixel, measuring brightness, contrast, highlights, shadows, and texture. It turns those five numbers into a negative prompt, the instructions that tell an AI image generator what to leave out, and a thermal printer prints it. Here, the thing to leave out is you.

The CRT runs Mimesis OS, a fictional operating system I designed as a webpage. It shows the visitor’s unmirrored photo, a version of their face that often feels unfamiliar, among windows of idealized faces and beauty ads. I wanted to recreate the discomfort of seeing your own appearance treated as something to correct.

The project grew from my research into how image generators learn what an attractive face looks like from photos and human ratings. The title refers to Plato’s mimesis: art as a copy of a world that is already a copy. Here, the ideal itself is built from copies.

A Polaroid camera on top of a CRT monitor showing a blue desktop
Fig. 1 Mimesis OS, installation view. Ryan Chen, 2026.
A thermal receipt headed WIST 80 listing luminance, contrast, highlights, shadows and texture
Fig. 2 Receipt printed by the installation. Ryan Chen, 2026.
A CRT monitor with a receipt coming out of a printer beside it
Fig. 3 The thermal printer and monitor. Ryan Chen, 2026.