Imagen de Portada

OmniCache: A Trajectory-Oriented Global Perspective on Training-Free Cache Reuse for Diffusion Transformer Models

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Chu, Huanpeng, Wu, Wei, Fen, Guanyu, Zhang, Yutao
Formato:	Preprint
Publicado:	2025
Materias:	Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

QNCD: Quantization Noise Correction for Diffusion Models
por: Chu, Huanpeng, et al.
Publicado: (2024)

AdaCorrection: Adaptive Offset Cache Correction for Accurate Diffusion Transformers
por: Liu, Dong, et al.
Publicado: (2026)

Rethinking Token-wise Feature Caching: Accelerating Diffusion Transformers with Dual Feature Caching
por: Zou, Chang, et al.
Publicado: (2024)

FastCache: Fast Caching for Diffusion Transformer Through Learnable Linear Approximation
por: Liu, Dong, et al.
Publicado: (2025)

CacheQuant: Comprehensively Accelerated Diffusion Models
por: Liu, Xuewen, et al.
Publicado: (2025)

BWCache: Accelerating Video Diffusion Transformers through Block-Wise Caching
por: Cui, Hanshuai, et al.
Publicado: (2025)

DisCa: Accelerating Video Diffusion Transformers with Distillation-Compatible Learnable Feature Caching
por: Zou, Chang, et al.
Publicado: (2026)

Accelerating Diffusion Transformers with Token-wise Feature Caching
por: Zou, Chang, et al.
Publicado: (2024)

H2-Cache: A Novel Hierarchical Dual-Stage Cache for High-Performance Acceleration of Generative Diffusion Models
por: Sung, Mingyu, et al.
Publicado: (2025)

OmniDiT: Extending Diffusion Transformer to Omni-VTON Framework
por: Zeng, Weixuan, et al.
Publicado: (2026)

SpeCa: Accelerating Diffusion Transformers with Speculative Feature Caching
por: Liu, Jiacheng, et al.
Publicado: (2025)

RT-Cache: Training-Free Retrieval for Real-Time Manipulation
por: Kwon, Owen, et al.
Publicado: (2025)

DreamCache: Finetuning-Free Lightweight Personalized Image Generation via Feature Caching
por: Aiello, Emanuele, et al.
Publicado: (2024)

Fast Autoregressive Video Diffusion and World Models with Temporal Cache Compression and Sparse Attention
por: Samuel, Dvir, et al.
Publicado: (2026)

FRDiff : Feature Reuse for Universal Training-free Acceleration of Diffusion Models
por: So, Junhyuk, et al.
Publicado: (2023)

Efficient Long-Horizon GUI Agents via Training-Free KV Cache Compression
por: Zhou, Bowen, et al.
Publicado: (2026)

AirCache: Activating Inter-modal Relevancy KV Cache Compression for Efficient Large Vision-Language Model Inference
por: Huang, Kai, et al.
Publicado: (2025)

Real2SAM2Real: Generative 3D Caches as Complementary Context for Video Diffusion
por: Wu, Jiayi, et al.
Publicado: (2026)

FreqCa: Accelerating Diffusion Models via Frequency-Aware Caching
por: Liu, Jiacheng, et al.
Publicado: (2025)

What Kind of Visual Tokens Do We Need? Training-free Visual Token Pruning for Multi-modal Large Language Models from the Perspective of Graph
por: Jiang, Yutao, et al.
Publicado: (2025)

Learning Generalized and Flexible Trajectory Models from Omni-Semantic Supervision
por: Zhu, Yuanshao, et al.
Publicado: (2025)

WorldCache: Content-Aware Caching for Accelerated Video World Models
por: Nawaz, Umair, et al.
Publicado: (2026)

VideoMLA: Low-Rank Latent KV Cache for Minute-Scale Autoregressive Video Diffusion
por: Yesiltepe, Hidir, et al.
Publicado: (2026)

LeMiCa: Lexicographic Minimax Path Caching for Efficient Diffusion-Based Video Generation
por: Gao, Huanlin, et al.
Publicado: (2025)

A Survey on Cache Methods in Diffusion Models: Toward Efficient Multi-Modal Generation
por: Liu, Jiacheng, et al.
Publicado: (2025)

Motion-Aware Caching for Efficient Autoregressive Video Generation
por: Xu, Jing, et al.
Publicado: (2026)

FlashBlock: Attention Caching for Efficient Long-Context Block Diffusion
por: Chen, Zhuokun, et al.
Publicado: (2026)

Model Reveals What to Cache: Profiling-Based Feature Reuse for Video Diffusion Models
por: Ma, Xuran, et al.
Publicado: (2025)

RoPECraft: Training-Free Motion Transfer with Trajectory-Guided RoPE Optimization on Diffusion Transformers
por: Gokmen, Ahmet Berke, et al.
Publicado: (2025)

Accelerating Diffusion-based Video Editing via Heterogeneous Caching: Beyond Full Computing at Sampled Denoising Timestep
por: Liu, Tianyi, et al.
Publicado: (2026)

EIDT-V: Exploiting Intersections in Diffusion Trajectories for Model-Agnostic, Zero-Shot, Training-Free Text-to-Video Generation
por: Jagpal, Diljeet, et al.
Publicado: (2025)

Cached Multi-Lora Composition for Multi-Concept Image Generation
por: Zou, Xiandong, et al.
Publicado: (2025)

KVCapsule: Efficient Sequential KV Cache Compression for Vision-Language Models with Asymmetric Redundancy
por: Huang, Yingbing, et al.
Publicado: (2026)

Multi-Cache Enhanced Prototype Learning for Test-Time Generalization of Vision-Language Models
por: Chen, Xinyu, et al.
Publicado: (2025)

From Reusing to Forecasting: Accelerating Diffusion Models with TaylorSeers
por: Liu, Jiacheng, et al.
Publicado: (2025)

Fast Sampling Through The Reuse Of Attention Maps In Diffusion Models
por: Hunter, Rosco, et al.
Publicado: (2023)

Chipmunk: Training-Free Acceleration of Diffusion Transformers with Dynamic Column-Sparse Deltas
por: Silveria, Austin, et al.
Publicado: (2025)

Cross-Self KV Cache Pruning for Efficient Vision-Language Inference
por: Pei, Xiaohuan, et al.
Publicado: (2024)

FAIRT2V: Training-Free Debiasing for Text-to-Video Diffusion Models
por: Zhong, Haonan, et al.
Publicado: (2026)

Trajectory-Consistent Calibration for Cache-Accelerated Diffusion Models
por: Liang, Mingyu, et al.
Publicado: (2026)