Go With the Flow: Fast Diffusion for Gaussian Mixture Models

Bewaard in:
Bibliografische gegevens
Gepubliceerd in:arXiv.org (Dec 24, 2024), p. n/a
Hoofdauteur: Rapakoulias, George
Andere auteurs: Pedram, Ali Reza, Tsiotras, Panagiotis
Gepubliceerd in:
Cornell University Library, arXiv.org
Onderwerpen:
Online toegang:Citation/Abstract
Full text outside of ProQuest
Tags: Voeg label toe
Geen labels, Wees de eerste die dit record labelt!

MARC

LEADER 00000nab a2200000uu 4500
001 3144197775
003 UK-CbPIL
022 |a 2331-8422 
035 |a 3144197775 
045 0 |b d20241224 
100 1 |a Rapakoulias, George 
245 1 |a Go With the Flow: Fast Diffusion for Gaussian Mixture Models 
260 |b Cornell University Library, arXiv.org  |c Dec 24, 2024 
513 |a Working Paper 
520 3 |a Schr\"{o}dinger Bridges (SB) are diffusion processes that steer, in finite time, a given initial distribution to another final one while minimizing a suitable cost functional. Although various methods for computing SBs have recently been proposed in the literature, most of these approaches require computationally expensive training schemes, even for solving low-dimensional problems. In this work, we propose an analytic parametrization of a set of feasible policies for steering the distribution of a dynamical system from one Gaussian Mixture Model (GMM) to another. Instead of relying on standard non-convex optimization techniques, the optimal policy within the set can be approximated as the solution of a low-dimensional linear program whose dimension scales linearly with the number of components in each mixture. Furthermore, our method generalizes naturally to more general classes of dynamical systems such as controllable Linear Time-Varying systems that cannot currently be solved using traditional neural SB approaches. We showcase the potential of this approach in low-to-moderate dimensional problems such as image-to-image translation in the latent space of an autoencoder, and various other examples. We also benchmark our approach on an Entropic Optimal Transport (EOT) problem and show that it outperforms state-of-the-art methods in cases where the boundary distributions are mixture models while requiring virtually no training. 
653 |a Probabilistic models 
653 |a Diffusion rate 
653 |a Parameterization 
653 |a Controllability 
653 |a Dynamical systems 
653 |a Convexity 
653 |a Time varying control systems 
653 |a Dimensional analysis 
653 |a Optimization techniques 
653 |a Normal distribution 
653 |a Optimization 
700 1 |a Pedram, Ali Reza 
700 1 |a Tsiotras, Panagiotis 
773 0 |t arXiv.org  |g (Dec 24, 2024), p. n/a 
786 0 |d ProQuest  |t Engineering Database 
856 4 1 |3 Citation/Abstract  |u https://www.proquest.com/docview/3144197775/abstract/embedded/ZKJTFFSVAI7CB62C?source=fedsrch 
856 4 0 |3 Full text outside of ProQuest  |u http://arxiv.org/abs/2412.09059