DSML Reading Club #4: sampling for Lévy–Itô diffusion models

The day after tomorrow, Assel Yermekova will present a paper she co-authored, “Improved Sampling Algorithms for Lévy-Itô Diffusion Models”.

Recent work showed that Lévy–Itô diffusion models with isotropic α-stable noise improve image generation on imbalanced data. However, existing sampling algorithms solve only approximate reverse equations, which reduces quality. In this paper, we propose a family of stochastic differential equations with identical marginal distributions and show that parameter selection improves quality with few reverse-diffusion steps. We also demonstrate Lévy–Itô models across different domains and the advantages of text-to-speech models on highly imbalanced data.

At the meeting, we will discuss:

• What is a diffusion model?
• Core formulations of diffusion models
• The limitations of classical Gaussian-process diffusion and why Lévy diffusion is needed
• Sampling methods for Lévy diffusion
• Thursday, May 22, 11:00 Kazakhstan time
• Add to calendar
• Google Meet
• Host: Yelaman Abdullin