Few-Step Diffusion Sampling

Follow this topic Ask about this topic

Few-step diffusion sampling asks how to produce images, videos, text, and other data with only a handful of denoising steps instead of hundreds or thousands. One major approach distills a many-step teacher into a faster student using consistency training, trajectory matching, or distribution matching; another improves existing samplers through better time schedules, ODE solvers, or error correction without retraining. Methods are tested on image benchmarks such as CIFAR-10 and ImageNet, text-to-image and video suites, and language-generation tasks, where speed is weighed against fidelity, prompt alignment, motion, and diversity. A central tension is that strong scores on individual samples can conceal lost distribution coverage or incorrect sampling, so evaluations increasingly need to measure both quality and diversity under matched compute budgets.

Latest papers

175 papers38 papers in the last four weeks, up 138% on the four weeks before. 0.2% of all new papers.

FIS-DiT: Breaking the Few-Step Video Inference Barrier via Training-Free Frame Interleaved Sparsity
May 12, 2026Jian Tang, Jiawei Fan, Qingbin Liu +1Video Diffusion Models Diffusion Transformer
Platform and Content Group, Tencent
STRIDE: Training-Free Diversity Guidance via PCA-Directed Feature Perturbation in Single-Step Diffusion Models
May 12, 2026Ankit Yadav, Arpit Garg, Ta Duc Huy +1Diffusion Model Guidance One-Step Generative Modeling
Australian Institute for Machine Learning, Adelaide University, Australia
Infinite Mask Diffusion for Few-Step Distillation
May 11, 2026Jaehoon Yoo, Wonjung Kim, Chanhyuk Lee +1Masked Diffusion Models Diffusion Model Distillation
Korea Advanced Institute of Science and Technology (KAIST)
When Few Steps Are Enough: Training-Free Acceleration of Identity-Preserved Generation
May 10, 2026Dongqi ZhengIdentity-Preserving Image Generation Few-Step Diffusion Sampling
FlashClear: Ultra-Fast Image Content Removal via Efficient Step Distillation and Feature Caching
May 9, 2026Yixin Tang, Jiawei Guo, Junxian Li +6Diffusion Model Distillation Few-Step Diffusion Sampling
Shanghai Jiao Tong University · Nanyang Technological University · Honor Device Co., Ltd
Geometry-Aware Discretization Error of Diffusion Models
May 8, 2026Samuel Hurault, Thomas Moreau, Gabriel PeyréDiffusion Sampling Few-Step Diffusion Sampling
Univ. Gustave Eiffel, CNRS, LIGM · Univ. Paris-Saclay, Inria, CEA · CNRS, ENS, PSL Université
Normalizing Trajectory Models
May 8, 2026Jiatao Gu, Tianrong Chen, Ying Shen +3Normalizing Flows Diffusion Model Distillation
Apple · UIUC
Stochastic Transition-Map Distillation for Fast Probabilistic Inference
May 8, 2026George Rapakoulias, Peter Garud, Lingjiong Zhu +1Diffusion Model Distillation MeanFlow
Department of Aerospace Engineering, Georgia Institute of Technology, Atlanta, GA · Department of Computer Science, Georgia Institute of Technology, Atlanta, GA · Department of Mathematics, Florida State University, Tallahassee, FL
LENS: Low-Frequency Eigen Noise Shaping for Efficient Diffusion Sampling
May 8, 2026Haewon Jeon, Si-Hyeon LeeDiffusion Model Distillation Few-Step Diffusion Sampling
School of Electrical Engineering, Korea Advanced Institute of Science and Technology (KAIST), Daejeon, South Korea
FlashMol: High-Quality Molecule Generation in as Few as Four Steps
May 7, 2026Xinyuan Wei, Zian Li, Shaoheng Yan +2Diffusion Distillation Few-Step Diffusion Sampling
Institute for Artificial Intelligence, Peking University · Yuanpei College, Peking University · School of Intelligence Science and Technology, Peking University +2
Continuous-Time Distribution Matching for Few-Step Diffusion Distillation
May 7, 2026Tao Liu, Hao Yan, Mengting Chen +8Diffusion Distillation Few-Step Diffusion Sampling
1VCIP, College of Computer Science, Nankai University · 2Alibaba Group · College of Artificial Intelligence, Jilin University
D-OPSD: On-Policy Self-Distillation for Continuously Tuning Step-Distilled Diffusion Models
May 6, 2026Dengyang Jiang, Xin Jin, Dongyang Liu +9Supervised Fine-Tuning Diffusion Model Distillation
1The Hong Kong University of Science and Technology · 2Z-Image Team, Alibaba Group · 4The Chinese University of Hong Kong +1
TOC-SR: Task-Optimal Compact diffusion for Image Super Resolution
May 4, 2026Sowmya Vajrala, Akshay Bankar, Manjunath Arveti +5Diffusion Distillation Few-Step Diffusion Sampling
Samsung Research Institute Bangalore, India
Hyper-DP3: Frequency-Aware Right-Sizing of 3D Diffusion Policies for Visuomotor Control
May 2, 2026Jinhao Zhang, Zhexuan Zhou, Huizhe Li +5Visuomotor Policy Learning Diffusion Policy
Harbin Institute of Technology, Shenzhen · Shanghai Jiao Tong University
Consistent Diffusion Language Models
Apr 30, 2026Hasan Amin, Yuan Gao, Yaser Souri +4Few-Step Diffusion Sampling Diffusion Model Acceleration
Department of Computer Science, Purdue University · Microsoft
AdvDMD: Adversarial Reward Meets DMD For High-Quality Few-Step Generation
Apr 29, 2026Xu Wang, Zexian Li, Litong Gong +2Diffusion Model Distillation Few-Step Diffusion Sampling
Shanghai Jiao Tong University · Alimama Tech
VS-DDPM: Efficient Low-Cost Diffusion Model for Medical Modality Translation
Apr 24, 2026Nikoo Moradi, Gijs Luijten, Behrus Hinrichs-Puladi +4Image-to-Image Translation 3D Medical Imaging
Institute for Artificial Intelligence in Medicine (IKIM), Essen University Hospital (AöR), University of Duisburg-Essen, Essen, Germany · Institute of Computer Graphics and Vision, Graz University of Technology, Inffeldgasse 16, Graz, 8010, Austria · Center for Virtual and Extended Reality in Medicine (ZvRM), University Hospital Essen (AöR), Essen, Germany +7
Efficient Diffusion Distillation via Embedding Loss
Apr 24, 2026Jincheng Ying, Yitao Chen, Li Wenlin +2Diffusion Model Distillation One-Step Generative Modeling
School of Big Data and Artificial Intelligence Guangdong University of Finance and Economics Guangzhou, China · School of Computer Science Shandong University Qingdao, China
Guiding Distribution Matching Distillation with Gradient-Based Reinforcement Learning
Apr 21, 2026Linwei Dong, Ruoyu Guo, Ge Bai +3Diffusion Model Distillation Few-Step Diffusion Sampling
1Zhejiang University · 2Bytedance Inc. · 3Zhejiang Lab
NI Sampling: Accelerating Discrete Diffusion Sampling by Token Order Optimization
Apr 20, 2026Enshu Liu, Xuefei Ning, Yu Wang +1Few-Step Diffusion Sampling Diffusion Language Model Decoding
Department of EE, Tsinghua University · Microsoft Research
Optimally Bridging Semantics and Data: Generative Semantic Communication via Schrödinger Bridge
Apr 20, 2026Dahua Gao, Ruichao Liu, Minxi Yang +3Semantic Communication Few-Step Diffusion Sampling
School of Artificial Intelligence, Xidian University, Xi’an, Shaanxi 710071 China · Pazhou Lab, Huangpu, Guangdong 510555 China · Peng Cheng Laboratory, Shenzhen, Guangdong 518055 China +1
Reward-Aware Trajectory Shaping for Few-step Visual Generation
Apr 16, 2026Rui Li, Bingyu Li, Yuanzhi Liang +3Diffusion Model Alignment Diffusion Model Distillation
University of Science and Technology of China HeFei, China · TeleAI ShangHai, China
MixFlow: Mixed Source Distributions Improve Rectified Flows
Apr 10, 2026Nazir Nayal, Christopher Wewer, Jan Eric LenssenRectified Flow Image Generation
Max Planck Institute for Informatics, Saarland Informatics Campus, Germany
FastDiSS: Few-step Match Many-step Diffusion Language Model on Sequence-to-Sequence Generation--Full Version
Apr 7, 2026Dat Nguyen-Cong, Tung Kieu, Hoang Thanh-TungDiffusion Language Model Inference Continuous Diffusion Language Models
FPT Software AI Center, FPT Corporation · Department of Computer Science, Aalborg University, Denmark · Quantum AI and Cyber Security Institute, FPT Corporation
FlowLM: Few-Step Language Modeling via Diffusion-to-Flow Adaptation
Apr 6, 2026Runzhe Zhang, Letian Chen, Wenpeng Zhang +2Language Modeling Flow-Based Generative Modeling
Shanghai Jiao Tong University · ByteDance
Isokinetic Flow Matching for Pathwise Straightening of Generative Flows
Apr 6, 2026Tauhid KhanFlow Matching Flow-Based Generative Modeling
1.x-Distill: Breaking the Diversity, Quality, and Efficiency Barrier in Distribution Matching Distillation
Apr 5, 2026Haoyu Li, Tingyan Wen, Lin Qi +6Diffusion Model Distillation Diffusion Sampling
Tsinghua University · Central Media Technology Institute, Huawei
EFlow: Fast Few-Step Video Generator Training from Scratch via Efficient Solution Flow
Mar 28, 2026Dogyun Park, Yanyu Li, Sergey Tulyakov +1Video Diffusion Models Diffusion Model Acceleration
Snap Inc. · Korea University, Seoul, Republic of Korea
BiFM: Bidirectional Flow Matching for Few-Step Image Editing and Generation
Mar 26, 2026Yasong Dai, Zeeshan Hayder, David Ahmedt-Aristizabal +1Flow Matching Diffusion-Based Image Generation
Australian National University, Australia · Data61-CSIRO, Australia
InverFill: One-Step Inversion for Enhanced Few-Step Diffusion Inpainting
Mar 24, 2026Duc Vu, Kien Nguyen, Ngan Nguyen +5Image Inpainting T2I Diffusion Models
Qualcomm AI Research · Posts & Telecommunications Inst. of Tech., Vietnam

May 12, 2026J/K move · Enter open · S save

FIS-DiT: Breaking the Few-Step Video Inference Barrier via Training-Free Frame Interleaved Sparsity

Jian Tang, Jiawei Fan, Qingbin Liu, Zheng Wei

Platform and Content Group, Tencent

Video Diffusion Models Diffusion Transformer Few-Step Diffusion Sampling Diffusion Model Inference Acceleration

Open paper AskPDF arXiv

While the overall inference latency of Video Diffusion Transformers (DiTs) can be substantially reduced through model distillation, per-step inference latency remains a critical bottleneck. Existing acceleration paradigms primarily exploit redundancy across the denoising trajectory; however, we identify a limitation where these step-wise strategies encounter diminishing returns in few-step regimes. In such scenarios, the scarcity of temporal states prevents effective feature reuse or predictive modeling, creating a formidable barrier to further acceleration. To overcome this, we propose Frame Interleaved Sparsity DiT (FIS-DiT), a training-free and operator-agnostic framework that shifts the optimization focus from the temporal trajectory to the latent frame dimension. Our approach is motivated by an intrinsic duality within this dimension: the existence of frame-wise sparsity that permits reduced computation, coupled with a structural consistency where each frame position remains equally vital to the global spatiotemporal context. Leveraging this insight, we implement Frame Interleaved Sparsity (FIS) as an execution strategy that manipulates frame subsets across the model hierarchy, refreshing all latent positions without requiring full-scale block computation. Empirical evaluations on Wan 2.2 and HunyuanVideo 1.5 demonstrate that FIS-DiT consistently achieves 2.11--2.41

\times

speedup with negligible degradation across VBench-Q and CLIP metrics, providing a scalable and robust pathway toward real-time high-definition video generation.