Diffusion Model from Scratch
It addresses the learning challenge for undergraduate and graduate students interested in diffusion models, but it is incremental as it focuses on educational guidance rather than new research.
The paper tackles the complexity of diffusion generative models by providing a step-by-step guide from VAEs to DDPM, using mathematical derivations and problem-oriented analysis to build foundational understanding for students.
Diffusion generative models are currently the most popular generative models. However, their underlying modeling process is quite complex, and starting directly with the seminal paper Denoising Diffusion Probability Model (DDPM) can be challenging. This paper aims to assist readers in building a foundational understanding of generative models by tracing the evolution from VAEs to DDPM through detailed mathematical derivations and a problem-oriented analytical approach. It also explores the core ideas and improvement strategies of current mainstream methodologies, providing guidance for undergraduate and graduate students interested in learning about diffusion models.