Denoising diffusion models (DDMs) offer a flexible framework for sampling from high dimensional data distributions. DDMs generate a path of probability distributions interpolating between a reference Gaussian distribution and a data distribution by incrementally injecting noise into the data. To numerically simulate the sampling process, a discretisation schedule from the reference back towards clean data must be chosen. An appropriate discretisation schedule is crucial to obtain high quality samples. However, beyond hand crafted heuristics, a general method for choosing this schedule remains elusive. This paper presents a novel algorithm for adaptively selecting an optimal discretisation schedule with respect to a cost that we derive. Our cost measures the work done by the simulation procedure to transport samples from one point in the diffusion path to the next. Our method does not require hyperparameter tuning and adapts to the dynamics and geometry of the diffusion path. Our algorithm only involves the evaluation of the estimated Stein score, making it scalable to existing pre-trained models at inference time and online during training. We find that our learned schedule recovers performant schedules previously only discovered through manual search and obtains competitive FID scores on image datasets.
Disentangling Total-Variance and Signal-to-Noise-Ratio Improves Diffusion Models
BIFOLD – Berlin Institute for the Foundations of Learning and Data Machine Learning Group, Technische Universität Berlin(柏林学习与数据基础研究所机器学习组,柏林技术大学);Technische Universität Berlin(柏林技术大学);Google DeepMind(谷歌DeepMind);Department of Artificial Intelligence, Korea University(韩国大学人工智能系);Max-Planck Institute for Informatics(马克斯·普朗克信息研究所);RIKEN Center for Advanced Intelligence Project(理化学研究所高级智能项目中心)
Noise Schedule Design for Diffusion Models: An Optimal Control Perspective
ECE & CSL University of Illinois Urbana-Champaign(电子工程与计算机科学实验室,伊利诺伊大学厄巴纳-香槟分校);Computer Science Department Carnegie Mellon University(计算机科学系,卡内基梅隆大学);ECE, CSL & NCSA University of Illinois Urbana-Champaign(电子工程、计算机科学实验室及国家计算科学中心,伊利诺伊大学厄巴纳-香槟分校)
A Comprehensive Review on Noise Control of Diffusion Model
University of Pittsburgh(匹兹堡大学);Northeastern University(东北大学);Columbia University(哥伦比亚大学);Northwestern University(西北大学);University of California, Berkeley(加州大学伯克利分校)
Class-frequency Guided Noise Schedule for Diffusion Models
Hefei University of Technology(合肥工业大学);University of Science and Technology of China(中国科学技术大学);The University of Hong Kong(香港大学);The Chinese University of Hong Kong(香港中文大学);Nanyang Technological University(南洋理工大学)
Equivariant Asynchronous Diffusion: An Adaptive Denoising Schedule for Accelerated Molecular Conformation Generation
Shanghai Academy of Artificial Intelligence for Science, SAIS(上海人工智能科学研究院,SAIS);College of Chemistry and Chemical Engineering, Xiamen University(厦门大学化学与化工学院);Artificial Intelligence Innovation and Incubation (AI 3 ) Institute, Fudan University(复旦大学人工智能创新与孵化(AI 3)研究院)
Hierarchical Schedule Optimization for Fast and Robust Diffusion Model Sampling
School of Computer Science and Engineering, Macau University of Science and Technology(澳门科学技术大学计算机科学与工程学院);Beijing Institute of Technology(北京理工大学);Zhejiang University(浙江大学)
ADiff4TPP: Asynchronous Diffusion Models for Temporal Point Processes
RBC Borealis(加拿大皇家银行博雷利斯);School of Computing Science, Simon Fraser University(西蒙菲莎大学计算科学学院);Department of Computer Science, University of British Columbia(不列颠哥伦比亚大学计算机科学系)
Center for Advanced Systems Understanding (CASUS)(先进系统理解中心);Helmholtz-Zentrum Dresden-Rossendorf e. V. (HZDR)(德累斯顿-罗斯托克亥姆霍尔茨研究中心);Department of Computing, Imperial College London(伦敦帝国理工学院计算机系);Bladder Infection and Immunity Group (BIIG), UCL Centre for Kidney and Bladder Health(膀胱感染与免疫组(BIIG),UCL肾与膀胱健康中心);Division of Medicine, University College London(伦敦大学学院医学系);Institute of Computer Science, University of Wrocław(沃斯克大学计算机科学研究所)
Generative Diffusion Surrogates with Analytical Variance Schedule
University of Oxford(牛津大学);Princeton University(普林斯顿大学);Max Planck Institute for Security and Privacy(马克斯·普朗克安全与隐私研究所);University of Cambridge(剑桥大学);Gonville & Caius College(贡维尔与凯斯学院)
OTCache: Optimal Transport for Geometry-Aware Caching in Diffusion Models
Data Science & Artificial Intelligence Research Institute, China Unicom(中国联通数据科学与人工智能研究院);Unicom Data Intelligence, China Unicom(中国联通联通数据智能);National Key Laboratory for Novel Software Technology, Nanjing University(南京大学计算机软件新技术国家重点实验室)
Gaoling School of Artificial Intelligence, Renmin University of China(中国人民大学首都人工智能学院);Beijing Key Laboratory of Research on Large Models and Intelligent Governance(北京大模型与智能治理重点实验室);Engineering Research Center of Next-Generation Intelligent Search and Recommendation, MOE(下一代智能搜索与推荐工程技术研究中心,教育部);Huawei Noah’s Ark Lab(华为诺亚实验室)