AI Columns was an exploration of artificial intelligence driven geometry. The project revolved around the creation of a digital exhibit for the University of Michigan Museum of Art. The team chose to digitally recreate the main hall of the museum. We then extended this hall replacing the current columns with the AI created columns. The columns are generated using a 2d to 3d style transfer technique. The 2d images used are the museums own collection so behind each column the original artwork is shown as well as the generated 2 dimensional AI representation of that same art piece.
Collaborators: Emre Guven, SoYoung Lee, and Mike Shin
Advisor: Matias del Campo
Each object has three corresponding parts. On the far left is the original input image. The second image shows the 2-dimensional AI representation of the first image. This image has all of the 3-d data of the final object embedded within itself in a way the AI can interpret. The final image shows the transformed 3-dimensional object.