CLASSICAL RELIEF

Elective Fall 2021

Collaboration with Bohao Xiong
Professors Ali Rahim + Caleb White


The project started as a convolutional neural network style transfer made from two images of existent architecture. By editing the code variables for the transfer process, multiple passes were made to create images with legible geometries. The selected image morphs the conventional forms of classical architecture into new compositional orders. The images were collaged and modeled into 3D where certain abstractions had to take place in order to make the model printable. The 3D printed model consists of 12 different pieces assembled for the final 18”x24” relief.