Diffusion Model Inference Acceleration

Follow this topic Ask about this topic

Momentum

40 papers in the last four weeks, up 100% on the four weeks before. 0.4% of all new papers.

Jul 13Week of Sep 28

Latest papers 335

All topics

DAVET: Denoising-Aware Visual Evidence Trajectory Allocation for Diffusion Vision-Language Models
Aug 3, 2026Yongkang Zhou, Xiang Xia, Cheng Yan +2Vision-Language Models Efficient VLM Inference
University of Science and Technology of China Hefei, China
REFLEX: Rethinking MoE Inference as Refinement-Aware Compute Allocation in Diffusion Language Models
Aug 3, 2026Xiang Xia, Cheng Yan, Yiming Zhang +3Expert Load Balancing Mixture-of-Experts Inference
University of Science and Technology of China, Hefei, China
Disagree to Accelerate: Closing the Loop on Diffusion Feature Forecasts
Aug 3, 2026Yanchao Li, Jiaqing Xie, Ben Gao +6Diffusion Sampling Diffusion Model Caching
1Nanjing University · 2Shanghai Artificial Intelligence Laboratory · 3North University of China +2
TurboClear: One-Step Object-Effect Removal via Region-Calibrated Distribution Matching and Fusion
Aug 2, 2026Jiawei Guo, Junxian Li, Yixin Tang +4Diffusion Model Distillation Diffusion-Based Image Editing
1Shanghai Jiao Tong University · 2Honor Device Co., Ltd · 3Imperial College London
DeltaFlow: Noise-Adaptive Bidirectional Gated Delta Networks for Embedded Language Flows
Aug 2, 2026Guangfu Guo, Xiaoqian Lu, Linsey Pang +4Gated Delta Networks Diffusion Language Models
1Clemson University · University of Ottawa · 3PayPal +1
DiffusionGemma Technical Report
Jul 31, 2026DiffusionGemma Team, Adrien Ali Taïga, James Assiene +41Blockwise Diffusion Language Models Diffusion Model Fine-Tuning
OnlineCache: Learning Dynamic Caching Policies with Error Correction for Efficient Diffusion Inference
Jul 31, 2026Zhikang Xie, Xichen Ye, Yifan Wu +5Diffusion Model Caching Diffusion Model Inference Acceleration
College of Computer Science and Artificial Intelligence, Fudan University · School of Data Science, Fudan University · 3Shanghai Key Laboratory of Intelligent Information Processing
FA-RDP: A Frequency-Adaptive Reactive Diffusion Policy for Contact-Rich Manipulation
Jul 30, 2026Lifeng Zhuo, Wendi Chen, Han Xue +4Contact-Rich Robotic Manipulation Visuomotor Policy Learning
1Shanghai Jiao Tong University · 2Shanghai Innovation Institute · 3Noematrix Ltd.
FeatFix: Reuse What You Verify through Local Exact-Feature Correction for Faster Cached Diffusion Inference
Jul 30, 2026Hanshuai Cui, Zhiqing Tang, Zhi Yao +3Diffusion-Based Image Generation Diffusion Model Caching
School of Artificial Intelligence, Beijing Normal University, Beijing 100875, China · Institute of Artificial Intelligence and Future Networks, Beijing Normal University, Zhuhai 519087, China
FARI: Robust One-Step Inversion for Watermarking in Diffusion Models
Jul 29, 2026Jindong Yang, Han Fang, Weiming Zhang +2Diffusion Model Inversion Diffusion Model Watermarking
University of Science and Technology of China · 2Anhui Province Key Laboratory of Digital Security
Noise-Free One-Step LoRA for Task-Driven Image Restoration with Diffusion Priors
Jul 28, 2026Jaeha Kim, Kyoung Mu LeeDiffusion Models for Image Restoration Low-Rank Adaptation
Dept. of ECE & ASRI, Seoul National University, Seoul, Korea · IPAI, Seoul National University, Seoul, Korea
MXAttention: Data-Free Optimal Scaling and Pre-Normalization Quantization for MXFP4 Attention
Jul 27, 2026Jianlin Yu, Jing Lin, Linghui Kong +13Video Diffusion Models Diffusion Model Quantization
Huawei Technologies Co., Ltd.
Sol-Attn: Accelerating Video Generation Inference via On-the-Fly Attention Sparsification
Jul 27, 2026Haopeng Li, Yitong Li, Junsong Chen +8Video Diffusion Models Sparse Attention
NVIDIA
OmniCache: Multidimensional Hierarchical Feature Caching For Diffusion Models
Jul 26, 2026Zhaoyuan He, Muhammad Muaz, Lili QiuVideo Diffusion Models Diffusion Transformer
Department of Computer Science, The University of Texas at Austin, Austin, TX, USA
TRaM-VSR: Importance-Aware Token Routing and Merging for One-Step Diffusion Video Super-Resolution
Jul 24, 2026Sicheng Gao, Zhuyun Zhou, Yixuan Liu +3Diffusion Transformer Video Super-Resolution
Advanced Micro Devices Inc. · Computer Vision Lab, CAIDAS & IFI, University of Würzburg
Neuromorphic Diffusion Language Models: Addressing Compute and Memory Bottlenecks via Sparsity and Block Denoising
Jul 24, 2026Dengyu Wu, Clement Ruah, Jiechen Chen +2Diffusion Language Model Decoding Neuromorphic Computing
Department of Engineering, King’s College London, London, UK · Institute for Intelligent Networked Systems (INSI), Northeastern University London, London, UK
Ms. Forcing: Efficient Streaming Video Generation with Multi-Scale Patchification and Attention
Jul 23, 2026Zekun Li, Xiaoyan Cong, Hongyu Li +5Video Diffusion Models Long-Horizon Video Generation
1Brown University · 2Massachusetts Institute of Technology · 3Meta
Evolving Cache Schedules for Fast Diffusion Policy Inference
Jul 22, 2026Siying Wang, Kangye Ji, Di Wang +1Diffusion Policy Diffusion Model Caching
School of Telecommunications Engineering, Xidian University, Xi’an, China · Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen, China · School of Computer Science and Technology, Xidian University, Xi’an, China +1
HeadCast: Casting Attention Heads for Efficient Autoregressive Video Generation
Jul 22, 2026Jinliang Shen, Lianghao Su, Zheming Li +4KV-Cache Management Autoregressive Video Generation
1Peking University · 2KlingAI Research
Speculative Correction: Draft-then-Refine Decoding for Diffusion Language Models
Jul 21, 2026Brian K Chen, Chong Wu, Kenji KawaguchiLanguage Model Decoding Speculative Decoding for Diffusion Language Models
National University of Singapore · City University of Hong Kong
Surprise Forcing: What to Remember, When to Skip in Long Video Generation
Jul 20, 2026Shuwei Shi, Zhen Li, Muyao Niu +4Long-Horizon Video Generation Memory-Augmented Video Generation
Alaya Lab · The University of Tokyo
FlowBlock: Wavefront-Parallel Decoding for Self-Correcting Diffusion Language Models
Jul 20, 2026Bing Tian, Haikun Liu, Xiaocheng Zhong +5Blockwise Diffusion Language Models Parallel Decoding
† Huazhong University of Science and Technology · § Xiaohongshu Inc., China
FVAttn: Adaptive Sparse Attention with Runtime Load Balancing for Video Generation
Jul 17, 2026Hao Liu, Chenghuan Huang, Ye Huang +7Video Diffusion Models Diffusion Transformer
Sun Yat-sen University · WeChat HPC, Tencent Inc. · Peking University +1
Adaptive Multi-Step Lookahead Decoding for Diffusion Language Models
Jul 17, 2026Yingqian Cui, Wei Deng, Lantao Mei +4Masked Diffusion Models Language Model Decoding
Michigan State University · Morgan Stanley · IBM T.J. Watson Research Center
DiTango: Cost-Effective Parallel Diffusion Generation with Selective Attention State Reuse
Jul 17, 2026Yuyang Chen, Runxin Zhong, Zan Zong +3Diffusion Transformer Efficient Attention
Shanghai Jiao Tong University, Shanghai AI Laboratory Shanghai, China · Tsinghua University Beijing, China · University of Science and Technology Beijing Beijing, China +1
Kaleido: Algorithm-Hardware Co-Design for Video Diffusion Transformers by Exploiting Latent Space Correlations
Jul 15, 2026Wenxuan Miao, Haosong Liu, Weiming Hu +9Video Diffusion Models Diffusion Transformer
Shanghai Jiao Tong University · Shanghai Jiao Tong University, Shanghai Qi Zhi Institute · Huawei Technologies +1
ACID: Adaptive Caching for vIDeo generation
Jul 14, 2026Om Agrawal, Saurabh Agarwal, Aditya AkellaVideo Diffusion Models Diffusion Model Caching
UT Austin
Reducing Temporal Redundancy for Efficient Vision-Language-Action Inference
Jul 14, 2026Yuzhou Wu, Yuxin Zheng, Muchun Niu +6Efficient VLA Model Inference Efficient VLA Models
1Tianji KernalMind co ltd · 2Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen, China · 3Shanghai Jiao Tong University, Shanghai, China +2
FlashDiff: Efficient Regional Execution and Scheduling for Diffusion Model Serving
Jul 13, 2026Yaqi Qiao, Ping He, Songrun Xie +4Diffusion Model Serving Diffusion Model Inference Acceleration
University of Illinois Urbana-Champaign · Vanderbilt University · HKUST +2
Bridging Diffusion Pruning and Step Distillation with Teacher-Aligned Repair
Jul 7, 2026Jincheng Ying, Li Wenlin, Minghui Xu +1Diffusion Model Distillation Few-Step Diffusion Sampling
School of Big Data and Artificial Intelligence Guangdong University of Finance and Economics Guangzhou, China · School of Computer Science Shandong University Qingdao, China · GDUFE-UCM Joint Institute Guangdong University of Finance and Economics Guangzhou, China

Aug 3, 2026, cs.CVJ/K move · Enter open · S save

DAVET: Denoising-Aware Visual Evidence Trajectory Allocation for Diffusion Vision-Language Models

Yongkang Zhou, Xiang Xia, Cheng Yan, Fan Xu+1

University of Science and Technology of China Hefei, China

Vision-Language Models Efficient VLM Inference Diffusion Model Inference Acceleration

Open paper AskPDF arXiv

Diffusion vision-language models (dVLMs) iteratively denoise masked responses while conditioning each denoising step on visual evidence, making visual conditioning a substantial recurring inference cost. Unlike autoregressive decoding, diffusion generation repeatedly revisits the entire response as uncertainty evolves. Our analysis reveals that visual evidence demand is strongly step-dependent, motivating adaptive allocation across denoising steps. Existing inference acceleration methods operate through decoding-side strategies or visual token compression via pruning and merging, but do not explicitly treat visual evidence as a resource whose demand evolves across the diffusion process. Therefore, we present Denoising-Aware Visual Evidence Trajectory Allocation (DAVET), a training-free framework that allocates visual evidence according to the evolving generation state. Starting from a phase-conditioned evidence trajectory, the proposed allocation policy uses operation demand to set an evidence reserve whose allocation at each denoising step is modulated by trajectory risk. DAVET realizes the resulting budgets through a hierarchy of evidence views constructed from a single visual encoding, separating when and how much evidence is needed from how the evidence views are constructed. Evaluated on two representative dVLMs, LLaDA-V and LaViDa, across multiple visual-understanding benchmarks, DAVET achieves an average speedup of 1.55

\times

with an average relative performance drop of 1.86%, showing that denoising-aware visual evidence allocation can reduce visual conditioning cost while largely preserving generation quality.

Momentum

Related topics

Latest papers 335