Few-Step Diffusion Sampling

Follow this topic Ask about this topic

Momentum

19 papers in the last four weeks, up 280% on the four weeks before. 0.2% of all new papers.

Jul 13Week of Sep 28

Latest papers 114

All topics

Two Halves are More than One: Phase-wise Velocity Distillation for Fast and High-Quality Image Generation
Oct 6, 2026Zhen Guo, Rongyuan Wu, Qiaosi Yi +3Diffusion Model Distillation Diffusion Distillation
Department of Computing, The Hong Kong Polytechnic University · OPPO Research Institute
DMAD: Distribution Matching as Adversarial Distillation for Fast Visual Generation
Oct 1, 2026Zhengming Yu, Junkun Yuan, Haotian Yang +8Diffusion Model Distillation Few-Step Diffusion Sampling
Texas A&M University · Intelligent Creation, ByteDance
Learned End-to-End Guidance Schedules for Diffusion Models
Oct 1, 2026Aneesh Barthakur, Mathias Niepert, Luiz F. O. ChamonDiffusion Model Guidance Few-Step Diffusion Sampling
University of Stuttgart · École polytechnique, Institut Polytechnique de Paris
Uncertainty-Aware Consistency Distillation for Few-Step Video Generation
Sep 30, 2026Lingyu Liu, Yaxiong Wang, Li Zhu +1Video Diffusion Models Few-Step Diffusion Sampling
School of Software Xi’an Jiaotong University Xi’an, China. · School of Computer and Information Science Hefei University of Technology Hefei, China. · Faculty of Science and Technology, and Institute of Collaborative Innovation University of Macau Macau, China.
Visualizing Distribution Coverage in Generative Diffusion Models
Sep 30, 2026Yifei Wang, Xiaoyu Wu, Tsu-Jui Fu +4Diffusion Distillation Few-Step Diffusion Sampling
Rice University · Apple
Acceleration of Diffusion Language Model through Discrete Average Generator
Sep 29, 2026Yidong Ouyang, Zhengyan Wan, Themis Haris +8Few-Step Diffusion Sampling Diffusion Language Model Inference
University of California, Los Angeles · Google · Work done as a research intern at Google
DMA $^2$ : Pixel-space Distribution Matching with Adversarial and Anchor Losses
Sep 29, 2026Xin Lin, Zhifei Zhang, Yuqian Zhou +6Diffusion Model Distillation Pixel-Space Diffusion Models
UC San Diego · Adobe Research · University of Virginia +1
MeanFlowAdvantage: Stable Reward Fine-Tuning for Few-Step Average-Velocity Generators
Sep 29, 2026Haocheng Tang, Tianchi Xie, Xingqiao LinFlow Matching MeanFlow
Northeastern University Boston, MA 02115, USA · Tsinghua University Beijing, 100084, PRC · Carnegie Mellon University Pittsburgh, PA 15213, USA
E-MoE: Enhanced Mixture-of-Experts for Non-Factorized Diffusion Language Models
Sep 29, 2026Arseny Ivanov, Alexander Kolesov, Alexander Korotin +2Masked Diffusion Models Few-Step Diffusion Sampling
AXXX, Moscow, Russia · Applied AI Institute, Moscow, Russia
Improved Distributional Diffusion Models
Sep 29, 2026Tommaso Martorella, Alexandre Galashov, Felix Krause +4Few-Step Diffusion Sampling Diffusion-Based Image Generation
Munich Center for Machine Learning (MCML) · Google DeepMind
Salt++: Context-Aligned Post-Training for Few-Step Streaming Multimodal Generation
Sep 29, 2026Xingtong Ge, Yutong Wang, Lunjie Zhu +7Audio-Video Generation Video Diffusion Models
The Hong Kong University of Science and Technology · Vivix Group Limited · The University of Sydney +1
Unlocking Few-Step Diffusion for Faithful Previews
Sep 28, 2026Jing Jia, Sifan Liu, Guanyang WangDiffusion Sampling Few-Step Diffusion Sampling
Department of Computer Science, Rutgers University · Department of Statistical Science, Duke University · Department of Statistics, Rutgers University
Simple Diffusion Language Models Are More Effective Few-Step Generators Than Reported
Sep 27, 2026Hasan Amin, Ming Yin, Rajiv KhannaMasked Diffusion Models Language Model Generation Evaluation
Purdue University
Spectral Amplitude Purification in Distribution Matching for Diffusion Distillation
Sep 24, 2026Zhenyu Zhou, Can Wang, Chun Chen +2Diffusion Model Distillation Few-Step Diffusion Sampling
Zhejiang University · University of California, Berkeley
ViRDM: Taming Representation Distribution Matching for Few-Step Causal Video Generation
Sep 24, 2026Zichong Meng, Chongjian Ge, Chun-Hao P. Huang +2Diffusion Model Distillation Distribution Matching
Northeastern University · Adobe Research
DART: Distillation-Aware Reparameterization for Training-Free LoRA Reuse in Few-Step Video Diffusion Models
Sep 17, 2026Shihong Li, Juntao Xu, JinCao +3Video Diffusion Models Diffusion Model Distillation
University of Electronic Science and Technology of China · Tsinghua University · Harbin Institute of Technology +1
Beyond Random Couplings: Contrastive Noise Alignment in Generative Flows
Sep 16, 2026Lennart Wittke, Vinicius AzevedoFlow Matching Cross-Modal Alignment
ETH Zürich · Disney Research | Studios
Representation-based Masked Diffusion Model
Sep 14, 2026Yangrong Hu, Ding Huang, Xueyu Zhou +1Non-Autoregressive Text Generation Masked Diffusion Models
Department of Data Science and Artificial Intelligence The Hong Kong Polytechnic University, Hong Kong SAR, China · Department of Applied Mathematics The Hong Kong Polytechnic University, Hong Kong SAR, China · ByteDance Seed, Beijing, China
CrossDistill: Balancing Quality and Diversity via Trajectory-Level Hybrid Few-Step Distillation
Sep 13, 2026Yuxi Liu, Haoyu Li, Yixiang Cai +8Video Diffusion Models Diffusion Model Distillation
Peking University, Melon Group · Alibaba Group · Tsinghua University
Uncertainty DMD: Restoring Diversity in Few-Step Autoregressive Video Distillation
Sep 11, 2026Zixuan Duan, Xunzhi Xiang, Yabo Chen +6Diffusion Model Distillation Few-Step Diffusion Sampling
Nanjing University, Institute of Artificial Intelligence, China Telecom (TeleAI), China · Institute of Artificial Intelligence, China Telecom (TeleAI), China · Fudan University, Institute of Artificial Intelligence, China Telecom (TeleAI), China
SelfLift: Accelerating Few-Step Diffusion via Self-Recovering Resolution Transition
Sep 2, 2026Tingyan Wen, Chenqian Yan, Xurui Peng +4Few-Step Diffusion Sampling Diffusion Model Inference Acceleration
1Tsinghua University · 2ByteDance
Efficient and High-Quality Depth Estimation via Pixel-Space Diffusion with Linear Attention
Aug 31, 2026Bingde Liu, Wu Ran, Jinglei Zhang +2Monocular Depth Estimation Pixel-Space Diffusion Models
MoE Key Lab of Artificial Intelligence, Institute of AI, Shanghai Jiao Tong University, Shanghai, China · Zhiyuan College, Shanghai Jiao Tong University, Shanghai, China
GeoFlow: Efficient Driving Video Generation via Geometry-Aligned Priors
Aug 12, 2026Jiazheng Liu, Hang Li, Jiawei Zhang +5Few-Step Diffusion Sampling Video Generation
School of Computer Science and Engineering, Beihang University · Macquarie University · Tianyi Transportation Technology Co., Ltd, Suzhou 215133, China +3
DUET: A Diversity-Quality Duet of Distillation Experts for Two-Step Video Generation
Aug 10, 2026Zian Li, Litong Gong, Borui Liao +6Video Diffusion Models Diffusion Model Distillation
Institute for Artificial Intelligence, Peking University, Beijing, China · School of Intelligence Science and Technology, Peking University, Beijing, China · *Work done as an intern at Alibaba Group +3
RL-Native Distillation: Exploiting Scored Trajectories for Few-Step Image Generation
Aug 10, 2026Yuhan Li, Fangao Zeng, Sicong Kang +7Diffusion Model Distillation Few-Step Diffusion Sampling
Shanghai Jiao Tong University · Taobao & Tmall group of Alibaba
Latent-Kernel Discrete Flow Maps for Few-Step Generation
Jul 29, 2026Mansoor Ahmed, Yue-Tsz Fan, Hemanth Venkateswara +1Latent Variable Models Few-Step Diffusion Sampling
Georgia State University, Atlanta, GA, USA · Georgia Institute of Technology, Atlanta, GA, USA
Parallel Decoding Distillation for Fast Image and Video Generation
Jul 28, 2026Neta Shaul, Chao Liu, Arash Vahdat +1Video Diffusion Models Diffusion Model Distillation
2Weizmann Institute of Science · 1NVIDIA
Manifold-Constrained Noise Optimization for Diverse Diffusion Sampling
Jul 27, 2026Qitan Shi, Cheng Jin, Ziyuan Liu +1Diffusion Model Sampling Image Generation
Tsinghua University
Multi-Mask Diffusion Language Models for Few-Step Generation
Jul 22, 2026Sijin Chen, Yinuo Ren, Heyang Zhao +3Masked Diffusion Models Diffusion Model Distillation
ByteDance Seed · Princeton University · Stanford University +2
Spatial Transport of Integration Error in Generative ODEs
Jul 17, 2026Songheng YinDiffusion Model Sampling Few-Step Diffusion Sampling
Columbia University

Oct 6, 2026, cs.CVJ/K move · Enter open · S save

Two Halves are More than One: Phase-wise Velocity Distillation for Fast and High-Quality Image Generation

Zhen Guo, Rongyuan Wu, Qiaosi Yi, Chenxi Xie+2

Department of Computing, The Hong Kong Polytechnic University · OPPO Research Institute

Diffusion Model Distillation Diffusion Distillation Few-Step Diffusion Sampling Diffusion-Based Image Generation T2I Diffusion Models

Open paper AskPDF arXiv

Recent diffusion-based image generation backbones have grown substantially in scale, making the network inference cost increase rapidly. While diffusion distillation techniques can reduce the number of inference steps, high-quality image generation within a single full-backbone-forward compute budget remains challenging. Existing one-step methods typically allocate this budget to a single evaluation of a monolithic student. However, approximating the heterogeneous coarse-to-fine transport with a single monolithic mapping is difficult and often leads to over-smoothed outputs. To address this issue, we propose Phase-wise Velocity Distillation (PVD), which partitions the generation timeline into a coarse and a fine phase, and models the transition within each phase via the average velocity. A dedicated half-sized expert is assigned to each phase, decoupling structural composition from detail refinement while keeping the cumulative computation equivalent to one full-backbone forward pass. We show that the use of two half-sized phase-specific experts outperforms a single full-size monolithic student. On class-conditional image generation, PVD achieves an FID of 1.48 on ImageNet 256 x 256. On more complex text-to-image (T2I) tasks, PVD-distilled models (Stable Diffusion 3.5-Medium, FLUX.1-dev, Qwen-Image) produce results competitive with their multi-step teachers, significantly outperforming prior distillation methods. Moreover, across the evaluated T2I backbones, PVD reduces active parameters by 49.10-50.89% and peak VRAM by 45.76-48.36% compared to the corresponding teachers. Source code and distilled models are available at https://github.com/PolyU-VCLab/PVD.

Momentum

Related topics

Latest papers 114