Tutorial on Diffusion Models for Imaging and Vision

TL;DR

This tutorial explains the core principles of diffusion models, a powerful sampling mechanism driving recent advances in generative imaging and vision applications like text-to-image and text-to-video synthesis.

Contribution

It provides an accessible overview of diffusion models, highlighting their fundamental ideas and recent growth in generative imaging and vision tasks.

Findings

Diffusion models have revolutionized generative image and video synthesis.

They overcome limitations of previous generative approaches.

The tutorial serves as an educational resource for newcomers.

Abstract

The astonishing growth of generative tools in recent years has empowered many exciting applications in text-to-image generation and text-to-video generation. The underlying principle behind these generative tools is the concept of diffusion, a particular sampling mechanism that has overcome some shortcomings that were deemed difficult in the previous approaches. The goal of this tutorial is to discuss the essential ideas underlying the diffusion models. The target audience of this tutorial includes undergraduate and graduate students who are interested in doing research on diffusion models or applying these models to solve other problems.