For my MFA thesis, I presented integrated mixed media projects that utilized XR (extended reality), projection mapping, generative AI imagery+3D environments, and an interactive depth camera installation. I developed custom AI workflows to create scenic projections, VR-ready environments, and real-time depth-based imagery for performance. These media included two-dimensional scenic assets, three-dimensional meshes, and real-time animation pipelines. By training models on 2,000+ hand captioned public-domain images, I embedded consistent stylistic vocabularies across formats ensuring continuity between projected scenography, VR navigation, and interactive projections.

The video above employs a “painting” version of the Aura LoRA (Low Rank Adaptation Model), while the one below employs a 3d “sculptural” LoRA

“Theatroscope” created an interactive VR environment with 3d models that were generated from images created by training 16gb model with the neural net associated in the latent space by experimenting with a system of custom tokens