Creative technologist working across generative art, machine learning, film, and sound design.
A web-based audio visualizer and exploratory application utilizing HTML5 and WebGL technologies. Provides pre-built audio journeys as well as access to oscillators and filters so that users can explore sound in a visual way and interact with the sound generation process to see changes reflected visually.
Staring into the map for too long, and holding iPad wondering: Wait, there might be some stories on here.
An automated interactive AI painter that can take your text prompts or camera inputs and turn them into paintings or images. It automatically captures the motion in front of the camera and uses it as inspiration for image generation. It constantly generates information based on available inputs, like a street-side portrait artist.
GAN neural network testing and training model for real-time interaction. Exploring speech-to-image and text-to-image generation pathways.
Technologies: VQGAN + CLIP
StyleGAN3 neural network directly linked to TouchDesigner for visual automatic generation. Real-time generation pipeline with interactive controls.
A semi-automatic deep learning model trained on vinyl album artworks. Generates new album cover designs automatically.
Adapted from a short story by Raymond Carver. Shot on ARRI SR2 + Blackmagic URSA Mini Pro 4.6k. Produced by Denison University Cinema Department.
Nan — Caitlin Wilson
Mike — Cedric Gegel
Director / DP / Color — Desmond Liang
Camera Assistant — Peter Mathes, Sam O'Malley
Sound — Peter Mathes
Original Music — Desmond Liang
Digital Intermediate — FotoKem
Weihang Zhou, Marc Wiskemann, Dave Bussan, Anthony Silvis, Ryan Swensen
GAN neural network training model with real-time interaction capabilities (speech-to-image conversion). Currently in maintenance mode.