Incorporating Pre-trained Diffusion Models in Solving the Schr\"odinger Bridge Problem

TL;DR

This paper unifies diffusion models and the Schr"odinger Bridge problem through reparameterization techniques, enabling faster, more stable training and leveraging pre-trained models for improved generative performance.

Contribution

It introduces three reparameterization methods and novel initialization strategies using pre-trained diffusion models to enhance Schr"odinger Bridge-based generative modeling.

Findings

Reparameterization techniques accelerate training stability.

Pre-trained diffusion models improve Schr"odinger Bridge performance.

Extensive experiments validate the effectiveness of the proposed methods.

Abstract

This paper aims to unify Score-based Generative Models (SGMs), also known as Diffusion models, and the Schr\"odinger Bridge (SB) problem through three reparameterization techniques: Iterative Proportional Mean-Matching (IPMM), Iterative Proportional Terminus-Matching (IPTM), and Iterative Proportional Flow-Matching (IPFM). These techniques significantly accelerate and stabilize the training of SB-based models. Furthermore, the paper introduces novel initialization strategies that use pre-trained SGMs to effectively train SB-based models. By using SGMs as initialization, we leverage the advantages of both SB-based models and SGMs, ensuring efficient training of SB-based models and further improving the performance of SGMs. Extensive experiments demonstrate the significant effectiveness and improvements of the proposed methods. We believe this work contributes to and paves the way for future research on generative models.