Seminar: Blockseminar: Neural ODE and Generative Modelling für Master - Details

General information

Course name	Seminar: Blockseminar: Neural ODE and Generative Modelling für Master
Course number	MTH-1360; MTH-1410
Semester	SS 2025
Current number of participants	0
Home institute	Mathematische Bildverarbeitung
Courses type	Seminar in category Teaching
Veranstaltung findet in Präsenz statt / hat Präsenz-Bestandteile	Yes
Hauptunterrichtssprache	englisch

Lecturers

• Prof.Dr. Stefania Petra

Rooms and times

No room preference

Fields of study

• Mathematisch-Naturwissenschaftlich-Technische Fakultät > Institut für Mathematik
• Mathematisch-Naturwissenschaftlich-Technische Fakultät > Institut für Mathematik > Master Mathematical Analysis and Modelling
• Mathematisch-Naturwissenschaftlich-Technische Fakultät
• Mathematisch-Naturwissenschaftlich-Technische Fakultät > Institut für Mathematik > Master Wirtschaftsmathematik
• Mathematisch-Naturwissenschaftlich-Technische Fakultät > Institut für Mathematik > Master Mathematik

Comment/Description

Score-based generative models have demonstrated state-of-the-art performance in numerous applications in recent years. The central concept behind these models involves the gradual injection of noise into the training data, followed by learning the reverse process to generate new samples. The training and sampling procedures can be conducted independently. The learning phase is facilitated by noise-conditional score networks, while sampling can be accomplished through various methods, including Langevin Monte Carlo approaches, stochastic differential equations, ordinary differential equations, and various combinations.

This seminar will overview generative models and common architectures, comparing different training objectives and sampling methods. We will examine normalizing flows, score matching, and Langevin dynamics. Additionally, we will discuss the use of stochastic differential equations (SDEs) in score-based models and conclude with comparisons to other diffusion models and potential enhancements in sample generation.