Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Lau, Kaikwan, Na, Andrew S., Wan, Justin W. L.
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2511.17634
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866913003270569984
author Lau, Kaikwan
Na, Andrew S.
Wan, Justin W. L.
author_facet Lau, Kaikwan
Na, Andrew S.
Wan, Justin W. L.
contents This paper presents a novel framework to accelerate score-based diffusion models. It first converts the standard stable diffusion model into the Fokker-Planck formulation which results in solving large linear systems for each image. For training involving many images, it can lead to a high computational cost. The core innovation is a cross-matrix Krylov projection method that exploits mathematical similarities between matrices, using a shared subspace built from ``seed" matrices to rapidly solve for subsequent ``target" matrices. Our experiments show that this technique achieves a 15.8\% to 43.7\% time reduction over standard sparse solvers. Additionally, we compare our method against DDPM baselines in denoising tasks, showing a speedup of up to 115$\times$. Furthermore, under a fixed computational budget, our model is able to produce high-quality images while DDPM fails to generate recognizable content, illustrating our approach is a practical method for efficient generation in resource-limited settings.
format Preprint
id arxiv_https___arxiv_org_abs_2511_17634
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Efficient Score Pre-computation for Diffusion Models via Cross-Matrix Krylov Projection
Lau, Kaikwan
Na, Andrew S.
Wan, Justin W. L.
Computer Vision and Pattern Recognition
This paper presents a novel framework to accelerate score-based diffusion models. It first converts the standard stable diffusion model into the Fokker-Planck formulation which results in solving large linear systems for each image. For training involving many images, it can lead to a high computational cost. The core innovation is a cross-matrix Krylov projection method that exploits mathematical similarities between matrices, using a shared subspace built from ``seed" matrices to rapidly solve for subsequent ``target" matrices. Our experiments show that this technique achieves a 15.8\% to 43.7\% time reduction over standard sparse solvers. Additionally, we compare our method against DDPM baselines in denoising tasks, showing a speedup of up to 115$\times$. Furthermore, under a fixed computational budget, our model is able to produce high-quality images while DDPM fails to generate recognizable content, illustrating our approach is a practical method for efficient generation in resource-limited settings.
title Efficient Score Pre-computation for Diffusion Models via Cross-Matrix Krylov Projection
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2511.17634