Skip to content

CS492(C): Diffusion and Flow Models

Minhyuk Sung, KAIST, Fall 2025


Teaser1

Time & Location

Time: Mon/Wed 10:30 a.m. - 11:45 a.m. (KST)
Location: E3-5 Room 210.

Description

Recent breakthroughs in generative AI have amazed people with the unprecedented quality of generated images and videos, as exemplified by SORA, Midjourney, StableDiffusion, and many others. These advancements have been achieved using diffusion models, which have become the new standard technique for generative models. Diffusion models offer numerous advantages, including superior performance in the quality of generated outputs, as well as capabilities in conditional generation, personalization, zero-shot manipulation, and knowledge distillation.

In this course, we will discuss the theoretical foundations and practical applications of diffusion models. While the goal is to cover both theory and practice, the focus will be on gaining hands-on experience by implementing diffusion model techniques in programming assignments and solving real-world problems in the course project.

Prerequisites

  • Solid background in machine learning and deep learning
  • Hands-on experience with neural network implementation
  • Recommended prior courses:
    • MAS.20050 Probability and Statistics
    • MAS.20001 Differential Equations and Applications
    • CS.30701 Introduction to Deep Learning

Course Staff

Instructor: Minhyuk Sung (mhsung@kaist.ac.kr)

Course Assistants:

Past Years

Grading

Useful Resources

Important Dates

ALL ASSIGNMENTS ARE DUE 23:59 KST.

(Subject to Change)

  • Project Team Sign-Up: Due Sep 30 (Tue)
  • 1st Programming Assignment: Due Oct 2 (Thu)
  • 2nd Programming Assignment: Due Nov 1 (Sat)
  • 3rd Programming Assignment: Due Nov 22 (Sat)
  • Image Generation Challenge Submission: Due Nov 15 (Sat)
  • Visual Generation Contest Submission: Due Dec 06 (Sat)

Schedule

(Subject to Change)

Week Mon Topic Wed Topic
1 Sep 01 Course Introduction
Slides		 Sep 03 Introduction to Generative Models
Slides
Recording
2 Sep 08 DDPM 1
Slides
Recording		 Sep 10 DDPM 2
Slides
Recording
3 Sep 15 Score-Based Models
Slides
Recording		 Sep 17 DDIM
Slides
Recording
4 Sep 22 Conditional Generation /
Latent Diffusion
Slides
Recording		 Sep 24 Assignment 1 Session
Slides
5 Sep 29 Diffusion Models in Continuous Time
Slides
Recording		 Oct 01 Demo Session
6 Oct 06 No Class (Chuseok) Oct 08 No Class (Chuseok)
7 Oct 13 ODE Solvers
Slides
Recording		 Oct 15 Assignment 2 Session
Slides
8 Oct 20 No Class (Midterm Week) Oct 22 No class (Midterm Week)
9 Oct 27 In-Class Test 1 Oct 29 Flow Matching 1
Slides
Recording
10 Nov 03 Flow Matching 2
Slides
Recording		 Nov 05 Assignment 3 Session
11 Nov 10 Inference-Time Guidance 1
Slides
Recording		 Nov 12 No Class (Break)
12 Nov 17 Inference-Time Guidance 2
Slides
Recording		 Nov 19 Score Distillation
Coarse Wrap-Up
Slides
Recording
13 Nov 24 Guest Lecture 1
Subham Sahoo
Recording		 Nov 26 Guest Lecture 2
Philipp Henzler
14 Dec 01 In-Class Test 2 Dec 03 No Class (Break)
15 Dec 08 Project Presentations 1 Dec 10 Project Presentations 2
16 Dec 15 No Class (Final Week) Dec 17 No Class (Final Week)

AI Coding Assistant Tool Policy

You are allowed (and even encouraged) to utilize AI coding assistant tools, such as ChatGPT, Copilot, Codex, and Code Intelligence, for your programming assignments and projects. Utilizing AI coding assistant tools will not be deemed as plagiarism. However, it is still strictly prohibited to directly copy code from the Internet or from someone else. Doing so will lead to a score of zero and a report to the university.


  1. Teaser image credits: Song et al., Score-Based Generative Modeling through Stochastic Differential Equations, ICLR 2021.
     ↩