Aviad Levis
Email / CV / Bio / Google Scholar / ORCiD /
CSC2539: Physics-Informed Neural Representations for Visual Computing
2025 Winter, 2025 Fall
​
Physics-informed neural representations combine the strengths of data-driven learning and physics-based models to solve forward and inverse problems in vision, graphics, imaging, and simulation. Neural fields, operator learning, and differentiable solvers enable compact, differentiable, continuous representations that can incorporate priors and physical constraints.
This AI-in-the-loop seminar blends weekly paper discussions with hands-on coding labs. We use GitHub as our collaborative backbone and deliberately integrate AI tools (e.g., ChatGPT) as junior collaborators—useful for brainstorming and iteration, while maintaining scientific rigor through verification and reflection.
Latest offering webpage
CSC320: Introduction to Visual Computing
2025 Winter, 2025 Fall
This course is a first-principles introduction to the acquisition and computational processing of 2D images. It is aimed at undergraduates interested in learning about computer vision, digital photography and computer graphics. The course serves as a stepping stone for tackling more advanced courses in those subjects
Latest offering webpage