A Survey on Cache Methods in Diffusion Models: Toward Efficient Multi-Modal Generation

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Liu, Jiacheng, Wang, Xinyu, Lin, Yuqi, Wang, Zhikai, Wang, Peiru, Cai, Peiliang, Zhou, Qinming, Yan, Zhengan, Yan, Zexuan, Shi, Zhengyi, Zou, Chang, Ma, Yue, Zhang, Linfeng
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866909881576980480
author Liu, Jiacheng
Wang, Xinyu
Lin, Yuqi
Wang, Zhikai
Wang, Peiru
Cai, Peiliang
Zhou, Qinming
Yan, Zhengan
Yan, Zexuan
Shi, Zhengyi
Zou, Chang
Ma, Yue
Zhang, Linfeng
author_facet Liu, Jiacheng
Wang, Xinyu
Lin, Yuqi
Wang, Zhikai
Wang, Peiru
Cai, Peiliang
Zhou, Qinming
Yan, Zhengan
Yan, Zexuan
Shi, Zhengyi
Zou, Chang
Ma, Yue
Zhang, Linfeng
contents Diffusion Models have become a cornerstone of modern generative AI for their exceptional generation quality and controllability. However, their inherent \textit{multi-step iterations} and \textit{complex backbone networks} lead to prohibitive computational overhead and generation latency, forming a major bottleneck for real-time applications. Although existing acceleration techniques have made progress, they still face challenges such as limited applicability, high training costs, or quality degradation. Against this backdrop, \textbf{Diffusion Caching} offers a promising training-free, architecture-agnostic, and efficient inference paradigm. Its core mechanism identifies and reuses intrinsic computational redundancies in the diffusion process. By enabling feature-level cross-step reuse and inter-layer scheduling, it reduces computation without modifying model parameters. This paper systematically reviews the theoretical foundations and evolution of Diffusion Caching and proposes a unified framework for its classification and analysis. Through comparative analysis of representative methods, we show that Diffusion Caching evolves from \textit{static reuse} to \textit{dynamic prediction}. This trend enhances caching flexibility across diverse tasks and enables integration with other acceleration techniques such as sampling optimization and model distillation, paving the way for a unified, efficient inference framework for future multimodal and interactive applications. We argue that this paradigm will become a key enabler of real-time and efficient generative AI, injecting new vitality into both theory and practice of \textit{Efficient Generative Intelligence}.
format Preprint
id arxiv_https___arxiv_org_abs_2510_19755
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A Survey on Cache Methods in Diffusion Models: Toward Efficient Multi-Modal Generation
Liu, Jiacheng
Wang, Xinyu
Lin, Yuqi
Wang, Zhikai
Wang, Peiru
Cai, Peiliang
Zhou, Qinming
Yan, Zhengan
Yan, Zexuan
Shi, Zhengyi
Zou, Chang
Ma, Yue
Zhang, Linfeng
Machine Learning
Artificial Intelligence
Computer Vision and Pattern Recognition
Diffusion Models have become a cornerstone of modern generative AI for their exceptional generation quality and controllability. However, their inherent \textit{multi-step iterations} and \textit{complex backbone networks} lead to prohibitive computational overhead and generation latency, forming a major bottleneck for real-time applications. Although existing acceleration techniques have made progress, they still face challenges such as limited applicability, high training costs, or quality degradation. Against this backdrop, \textbf{Diffusion Caching} offers a promising training-free, architecture-agnostic, and efficient inference paradigm. Its core mechanism identifies and reuses intrinsic computational redundancies in the diffusion process. By enabling feature-level cross-step reuse and inter-layer scheduling, it reduces computation without modifying model parameters. This paper systematically reviews the theoretical foundations and evolution of Diffusion Caching and proposes a unified framework for its classification and analysis. Through comparative analysis of representative methods, we show that Diffusion Caching evolves from \textit{static reuse} to \textit{dynamic prediction}. This trend enhances caching flexibility across diverse tasks and enables integration with other acceleration techniques such as sampling optimization and model distillation, paving the way for a unified, efficient inference framework for future multimodal and interactive applications. We argue that this paradigm will become a key enabler of real-time and efficient generative AI, injecting new vitality into both theory and practice of \textit{Efficient Generative Intelligence}.
title A Survey on Cache Methods in Diffusion Models: Toward Efficient Multi-Modal Generation
topic Machine Learning
Artificial Intelligence
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2510.19755