About

This course provides an introduction to optimal transport and its applications in generative modeling, with a particular focus on domain translation.

Syllabus

1. Introduction to Optimal Transport (OT)
2. Entropy-Regularized Optimal Transport and Schrödinger Bridges
3. Schrödinger Bridges and Entropic Optimal Transport (ENOT)
4. Schrödinger Bridges via Score-Based Models
5. Schrödinger Bridge Matching
6. Flow Models and Optimal Transport

Coursework

• Two homework assignments focused on practical implementations of generative models based on optimal transport theory
• A paper reading and presentation session
• A final project

Grading

• Homework assignments: 40 points total (20 + 20)
• Paper presentation/report session: 20 points
• Project: 30 points
• Final exam: 20 points

Final grade: min(round(total points / 10), 10)

Prerequisites

• Probability Theory and Statistics
• Deep Learning
• A prior course in Generative Modeling