BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Schrödinger bridges\, diffusion and SDEs  - Stratis Markou and Sh
 reyas Padhy
DTSTART:20230621T100000Z
DTEND:20230621T113000Z
UID:TALK202708@talks.cam.ac.uk
CONTACT:Isaac Reid
DESCRIPTION:In this talk\, we cover the mathematical groundwork for stocha
 stic differential equations (SDEs)\, introducing Ito and Stratanovich calc
 ulus. We then talk about how SDEs serve as a natural framework for unifyin
 g many generative modelling techniques\, such as score-matching\, denoisin
 g diffusion models\, and conditional flows. Finally\, we discuss some theo
 retical convergence results for SDEs\, and show improvements to generative
  modelling through the Schrodinger Bridge formulation\, which improves upo
 n convergence in SDEs by imposing boundary conditions. Finally\, we discus
 s some empirical techniques for solving Schrodinger Bridge problems for ge
 nerative modelling.\n\nReferences:\n[1] Song\, Yang\, et al. “Score-base
 d generative modeling through stochastic differential equations.” arXiv 
 preprint arXiv:2011.13456 (2020).\n[2] De Bortoli\, Valentin\, et al. “D
 iffusion Schrödinger bridge with applications to score-based generative m
 odeling.” Advances in Neural Information Processing Systems 34 (2021): 1
 7695-17709.
LOCATION:Cambridge University Engineering Department\, CBL Seminar room BE
 4-38.
END:VEVENT
END:VCALENDAR