Physics-informed Wave Simulation
I work at the boundary where physical law meets generative imagination, building software that turns equations into visual experiences, and turning canvases into interactive simulations.
To me, science and art are not opposites. They are two languages describing the same reality. My work lives at their intersection, using physics-informed models to create generative art that is grounded in the real world, yet free to imagine beyond it.
That attempt has moved across mediums. Stylus reframed music as something a diffusion model could re-paint in the spectrogram domain. Revisiting Your Memory turned EEG signals into audiovisual reconstructions of affect-contextualized memory. More recently, my work has moved toward 3D human motion, using physics-informed models to ground how bodies move in the real world.
Different surfaces, one question.
What is the world made of, and how do we show it?
A multimodal framework that reconstructs video and music contextualized by human affect, using brain signals as the seed of the imagery.
A training-free framework that re-purposes a pre-trained Stable Diffusion model to perform music style transfer in the mel-spectrogram domain.