Sampling and diffusion models

December 2, 2024 · 1 min read · Machine Learning

Group meeting slides and reading materials on score-based models, inverse problems, and sampling theory (2024).

Contents / 目录

This page collects my group meeting slides and reading materials on sampling and diffusion models from August to December 2024.

Score-based generative modeling: SMLD and SDE

August 2024 · Group meeting presentation

A reading presentation on Yang Song’s Generative Modeling by Estimating Gradients of the Data Distribution and Score-Based Generative Modeling through Stochastic Differential Equations, covering score-based generative models and their SDE formulation.

Slides: SMLD and SDE (PDF)

Score-based variational inference for inverse problems

September 2024 · Group meeting presentation

Reading material for Score-Based Variational Inference for Inverse Problems, on the use of diffusion models in inverse problems.

Reading presentation (PDF)

Sampling theory with minimal data assumptions

December 2024 · Group meeting presentation

A presentation of Sampling is as easy as learning the score: theory for diffusion models with minimal data assumptions, focusing on convergence guarantees for diffusion-model sampling.

Slides: sampling theory (PDF)

For background on Bayesian inference and sampling, see Bayesian signal processing.

← All posts