CME 296: Diffusion and Large Vision Models

Course Description

This class explores the evolution of computer vision from early classification models to modern generative systems powered by diffusion and large vision models. Through a mix of theory and practical insights, student will learn how these models operate and how they're applied in real-world scenarios. Topics include: introduction to multimodal learning with models like CLIP that connect vision and language; deep dive into generative models such as autoencoders, GANs, and diffusion models; controllability and practical applications: inpainting, segmentation, text-to-image, and video generation. Ideal for those with a background in calculus, linear algebra, and basic machine learning concepts.