Project with Elizabeth Christoforetti
Winners of the 4th prize of Re:humanism Artist Call
This project opens up creative architectural production to imagine
a new era of design of the built environment in which any individual may have
an authorial stake in the imagination and production of their built context. In
this project, any body may be cast as building; collectively, the production of
this work will make up a new type of urbanism, a neighborhood and place that is
imagined as a direct translation of the accumulative identity of its
inhabitants. The identity, role, and agency of the designer is scrambled into a
new set of horizontal relationships, having slipped out of its historical
top-down orientation.
![]()
![]()
![]()
The minimal exhibition footprint consists of a single screen and a
camera above it, that captures the face of a visitor and translates its edges
and details into a computationally generated facade. Before starting, the
visitor picks a city (one out of 3) they would like to reside in, which then
dictates the style or facade type that will be created. It thus learns both
from the visitor’s features and a styleGAN model trained on the city they’ve
chosen. Computationally, the result will find minute differences in visitors’
facial characteristics -- based upon feature proportion, relative
relationships, and expression exaggeration rather than skin color -- while also
making room for some randomness. The aim is to allow for an emergence of a rich
variety of homes, which historically gets diluted with physical and stylistic
segregation and the drive of the real estate market toward market-ready
generics.
The minimal exhibition footprint consists of a single screen and a camera above it, that captures the face of a visitor and translates its edges and details into a computationally generated facade. Before starting, the visitor picks a city (one out of 3) they would like to reside in, which then dictates the style or facade type that will be created. It thus learns both from the visitor’s features and a styleGAN model trained on the city they’ve chosen. Computationally, the result will find minute differences in visitors’ facial characteristics -- based upon feature proportion, relative relationships, and expression exaggeration rather than skin color -- while also making room for some randomness. The aim is to allow for an emergence of a rich variety of homes, which historically gets diluted with physical and stylistic segregation and the drive of the real estate market toward market-ready generics.
