Donnie Stewart
Computer Vision Researcher
Currently at Industrial Light & Magic, where I support and enhance the Global Render Pipeline to help create visually stunning films. Previously, I conducted research at UC Santa Cruz, earning an M.S. in Computer Science and publishing two papers in the process.
​
I collaborated with Professor Alex Pang and the AVIS lab on computer vision research aimed at community-focused applications. One example — featured in the section below — is a drone-based system we developed for real-time rip current detection.
​
This site showcases my technical background, including Publications, Projects, Education, and Experience in computer science. Feel free to explore and reach out via the contact form below if you’d like to connect​​.
Research Demo
This video playing on the right is a recording from a pilot application developed in part for my research project Realtime ML-Assisted Data Collection using Drones with the AVIS lab. Clicking the title in the previous sentence will provide a more detailed explanation of the project. To capture this video I put a lightweight Machine Learning model on a cell phone that connects to the drone's camera. I then flew a drone over several beaches in Santa Cruz. Once the model detects a rip current, a recording automatically starts for 30 seconds (hence the timer in the center). The resulting video gets saved on the phone.
This work was a part of larger collection of drone research, including two publications, featured in the next section.
Note: "Back" and "Stop ML" are features of the app that won't work in this video.
Publications
RipScout: Realtime ML-Assisted Rip Current Detection and Automated Data Collection Using UAVs​
Published in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing. (2025)
RipViz: Finding Rip Currents by Learning Pathline Behavior,
Published in IEEE Transactions on Visualization and Computer Graphics. (2023)