Diffusion Model Acceleration

Follow this topic Ask about this topic

Diffusion model acceleration studies how to make iterative denoising models faster and cheaper to train and run across images, video, language, and scientific applications. Common approaches reduce the number of denoising steps through distillation, consistency learning, or adaptive solvers, and reduce work per step through caching, token or attention sparsity, quantization, and efficient architectures. Research also targets training throughput and deployment, using optimized kernels, parallelism, and compact models on settings such as text-to-image benchmarks, long-video generation, and medical imaging. The central challenge is balancing real wall-clock speed and memory savings against fidelity, diversity, and temporal consistency, since effective shortcuts can depend on the model, task, and denoising stage.

Latest papers

141 papers21 papers in the last four weeks, up 133% on the four weeks before. 0.1% of all new papers.

SpeeDiff: Scalable Pixel-Anchored End-to-End Latent Diffusion Model
Oct 26, 2026Bingliang Zhang, Wenda Chu, Yizhuo Li +5Latent Diffusion Models Diffusion Transformer
ByteDance Seed · California Institute of Technology · Meta
LEAP: Accelerating ODE-based Diffusion Sampling with Optimal Linear Leaps
Oct 19, 2026Haowei Zheng, Zhenyu Zhou, Jiawei Chen +2Diffusion Model Sampling Diffusion Model Acceleration
State Key Laboratory of Blockchain and Data Security, Zhejiang University. · Hangzhou High-Tech Zone (Binjiang) Institute of Blockchain and Data Security. · University of California, Berkeley, California, United States
Which Tokens to Merge? Diffusion Dynamics for Efficient Image Generation
Oct 15, 2026SeungJu Cha, Ye-Chan Kim, HyunGee Kim +2Visual Token Merging Diffusion Model Acceleration
Hanyang University, South Korea
Representation Conditioning for Generation
Oct 14, 2026Saeed Razavi, Xingjian Leng, Liang ZhengDiffusion-Based Image Generation Diffusion Transformer
Australian National University
MiniDog: Dog-Breed Text-to-image Flow Matching from Pretraining to Fine-Tuning on 4 Consumer GPUs
Oct 13, 2026Reyhaneh Esmailizadeh, Xingjian Leng, Zhanhao Liang +1T2I Generation T2I Diffusion Models
Heuristics are Not Enough: Calibrated and Corrected Caching for Video Diffusion Acceleration
Oct 12, 2026Jiyue Jiang, Zhiyu Huo, Yuhan Chen +3Diffusion Model Caching Diffusion Model Acceleration
Department of Computer Science and Engineering, The Chinese University of Hong Kong · Theory Lab, 2012 Labs, Huawei Technology Co. Ltd
A Step in the Right Place: Understanding and Optimizing Timestep Placement in Continuous Diffusion Language Models
Oct 11, 2026Shutian Zheng, Zhihong Yu, Linye Wei +2Continuous Diffusion Language Models Diffusion Timestep Scheduling
College of Engineering, Peking University · Weiyang College, Tsinghua University · Institute for Artificial Intelligence, Peking University +3
Competition Domains Matter: Selective Head Globalization for Training-Free Sparse Attention in Video Diffusion Transformers
Oct 10, 2026Tiantian Liu, Meng Wan, Zhiqiang Liang +3Sparse Attention Diffusion Model Acceleration
University of Chinese Academy of Sciences, Beijing, China
TGuide: Faster Autoregressive Video Diffusion via Historical Trajectory Guidance
Oct 10, 2026Zhuo Chen, Bin Zou, Fanyue Wei +3Autoregressive Video Generation Diffusion Model Acceleration
University of Electronic Science and Technology of China · University of Hong Kong · National University of Singapore
High-Throughput Generative Video Super-Resolution on a Single Consumer GPU
Oct 9, 2026Georgios Panaretou, Sotirios ChatzisVideo Super-Resolution Temporal Consistency in Video Generation
Ethical AI Novelties Limassol, Cyprus · Cyprus University of Technology Limassol, Cyprus
APEX: Asynchronous Prefix Denoising for Extreme Compression
Oct 9, 2026David Dinkevich, Nisan Chiprut, Yoav HaCohen +1Latent Diffusion Models Diffusion Model Compression
Lightricks · Hebrew University of Jerusalem
Parametric Trajectory Distillation for Few-Step Video Generation
Oct 8, 2026Lan Feng, Peter Karkus, Maximilian Igl +6Diffusion Model Distillation Video Diffusion Models
EPFL · NVIDIA · Stanford University
Denoising Blocks, Not Tokens: Efficient Compressed Continuous Diffusion with Branching Token Realization
Oct 7, 2026Xinsong Feng, Peng Du, Zhizhuo Yang +3Diffusion Language Models Diffusion Language Model Inference
Georgia Institute of Technology · Writer AI Research · William & Mary
Adaptive Second-Order Solvers for Fast Stochastic Diffusion Sampling
Oct 2, 2026Ella Kemperman, Luca AmbrogioniDiffusion Sampling Diffusion Timestep Scheduling
Radboud University
TACD: Distilling Efficient Text-to-Motion Models via Terminal Amplification Control
Oct 2, 2026Wei-Jin Huang, Yuan-Ming Li, Kun-Yu Lin +8Diffusion Model Distillation Diffusion Models for Motion Generation
Sun Yat-sen University · Nanyang Technological University · Wuhan University +2
Just Align $\bm{x}$ : Aligning Predictions, Not Representations
Sep 30, 2026Yuyao Zhang, Yuwei Hu, Ziyang Mai +1Diffusion Model Alignment Diffusion Transformer
Dartmouth College
Rank-Aware Speculative Sampling for Diffusion Draft Trees
Sep 30, 2026Marcello Bullo, Yanxiao Liu, Öykü Sıla Güner +2Diffusion Sampling Diffusion Model Acceleration
Imperial College London
DC-SAE: Deep Compression Semantic Autoencoder for Faster Diffusion Convergence
Sep 30, 2026Xu Huang, Ye Huang, Zijun Liao +6Autoencoders Diffusion-Based Image Generation
Peking University · Singapore University of Technology and Design
SoL-Refiner: Speed-of-Light One-Step Refinement for High-Resolution Video
Sep 29, 2026Haozhe Liu, Tian Ye, Shuchen Xue +9Diffusion Model Acceleration Video Super-Resolution
NVIDIA
ReCaVSR: One-Step Streaming Diffusion Video Super-Resolution with Recycled Latents and Learned Cache Routing
Sep 29, 2026Xijun Wang, Xin Li, Suhang Yao +3Diffusion Transformer Streaming Video Generation
University of Science and Technology of China
PDMD: Projected Distribution Matching Distillation for Video Diffusion Models
Sep 28, 2026Zimo Wang, Junkun Yuan, Angtian Wang +9Video Diffusion Models Diffusion Model Distillation
University of California, San Diego · ByteDance Inc.
Scaffold Then Internalize: Representation Injection for Diffusion Transformers
Sep 28, 2026Han Fu, Jiacheng Chen, Baoquan Zhao +4Diffusion Transformer Diffusion Model Acceleration
Sun Yat-sen University · Video Rebirth · The Hong Kong Polytechnic University
GeoShrink: Accelerating Diffusion Transformers with Two Lines of Code
Sep 27, 2026Haosen Li, Wenshuo Chen, Shaofeng Liang +3Video Diffusion Models Diffusion Transformer
The Hong Kong University of Science and Technology (Guangzhou) · Griffith University · Data61, CSIRO +1
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
UltraTex: Unleashing 2K Multi-View Diffusion for 3D Texturing
Sep 19, 2026Yibo Zhang, Ze Yuan, Nan Cao +4Multi-View Image Synthesis 3D Texture Generation
Jilin University, Changchun, China · Shanghai Innovation Institute, Shanghai, China · The University of Hong Kong, Hong Kong, China +3
SparkDiffusion: Mitigating the High-Sparsity Trap --- A Unified Framework for up to $265\times$ Single-GPU Acceleration of Visual Generation
Sep 19, 2026Yuxi Liu, Haoyu Li, Zekun Zhang +12Diffusion Model Acceleration Video Diffusion Models
Peking University, Melon Group · Alibaba Group · Tsinghua University +3
Accelerating Diffusion Sampling via Speculative Draft Trees
Sep 15, 2026Marcello Bullo, Yanxiao Liu, Öykü Sıla Güner +2Diffusion Sampling Diffusion Model Acceleration
Imperial College London
MedDiME: Efficient Latent Diffusion with Adaptive Masking for Medical Counterfactual Generation
Sep 14, 2026Yan Zeng, Changlu Guo, Anders Nymark Christensen +2Counterfactual Explanations Latent Diffusion Models
Department of Applied Mathematics and Computer Science, Technical University of Denmark
Importance-Aware Low-Rank Distillation of Diffusion Transformers
Sep 4, 2026Denis Zavadski, Sebastian Heid, Damjan Kalšan +2Diffusion Model Compression T2I Diffusion Models
Heidelberg University · Zuse School ELIZA · TU Darmstadt
LeanGRPO: Eliminating Redundant Recomputation in Diffusion RL
Sep 3, 2026Sijie Wang, Zhiqiang Tan, Xinrui Yang +1RL for Diffusion Models RL Post-Training
School of Computer Science and Technology, Harbin Institute of Technology, Shenzhen

Oct 26, 2026J/K move · Enter open · S save

SpeeDiff: Scalable Pixel-Anchored End-to-End Latent Diffusion Model

Bingliang Zhang, Wenda Chu, Yizhuo Li, Linjie Yang+4

ByteDance Seed · California Institute of Technology · Meta

Latent Diffusion Models Diffusion Transformer Diffusion Model Acceleration

Open paper AskPDF arXiv

We present Scalable Pixel-anchored End-to-end Diffusion (SpeeDiff), a latent diffusion method that jointly trains the VAE and the diffusion model from scratch. In principle, joint training allows the diffusion loss gradient to directly guide the VAE encoder, encouraging the formation of a generation-friendly latent space and potentially yielding faster convergence than the conventional two-stage approach with a pretrained frozen VAE. However, a naive end-to-end implementation severely degrades performance, as unrestricted backpropagation of the diffusion loss leads to latent space collapse. Our main technical contribution is a simple yet effective Tweedie Pixel Reconstruction (TPR) loss, which provides additional pixel-level feedback by decoding a predicted clean latent from an intermediate noisy state using Tweedie's formula, thereby alleviating collapse. Furthermore, our method enables jointly scaling a fully transformer-based architecture and enhances representation alignment within the end-to-end framework. Our SpeeDiff-XL model achieves over 140x and 61x faster training compared to Vanilla SiT and REPA, respectively, while attaining an FID of 1.50 without guidance on ImageNet 256x256 generation. With a more efficient 32x compressed VAE, our model further reaches an FID of 1.53 without guidance on ImageNet 512x512 generation.