Time series generation is widely used in real-world applications such as simulation, data augmentation, and hypothesis testing. Recently, diffusion models have emerged as the de facto approach to time series generation, enabling diverse synthesis scenarios. However, the fixed standard-Gaussian diffusion prior may be ill-suited for time series data, which exhibit properties such as temporal order and fixed time points. In this paper, we propose TimeBridge, a framework that flexibly synthesizes time series data by using diffusion bridges to learn paths between a chosen prior and the data distribution. We then explore several prior designs tailored to time series synthesis. Our framework covers (i) data- and time-dependent priors for unconditional generation and (ii) scale-preserving priors for conditional generation. Experiments show that our framework with data-driven priors outperforms standard diffusion models on time series generation.
Underdamped Diffusion Bridges with Applications to Sampling
Karlsruhe Institute of Technology(卡尔斯鲁厄理工学院);NVIDIA(NVIDIA公司);Zuse Institute Berlin(柏林泽尼克研究所);dida Datenschmiede GmbH(dida数据安全公司);FZI Research Center for Information Technology(信息技术研究中心)
Bridge-IF: Learning Inverse Protein Folding with Markov Bridges
Liangzhu Laboratory, Zhejiang University(浙江大学良渚实验室);College of Pharmaceutical Sciences, Zhejiang University(浙江大学药学院);Alibaba Cloud Computing(阿里云计算);School of Public Health, Zhejiang University(浙江大学公共卫生学院);School of Artificial Intelligence and Data Science, University of Science and Technology of China(中国科学技术大学人工智能与数据科学学院);The Second Affiliated Hospital Zhejiang University School of Medicine(浙江大学医学院附属第二医院)
Sparsely-Supervised Data Assimilation via Physics-Informed Schrödinger Bridge
Department of Industrial \& Systems Engineering, KAIST, Daejeon, Repulic of Korea;Department of Mathematical Sciences, KAIST, Daejeon, Repulic of Korea
School of AIA, Huazhong University of Science and Technology(华中科技大学AIA学院);Department of Computer Science and Engineering, Shanghai Jiao Tong University(上海交通大学计算机科学与工程系);Center for Applied Statistics and School of Statistics, Renmin University of China(中国人民大学应用统计中心及统计学院);School of EIC, Huazhong University of Science and Technology(华中科技大学EIC学院)
Exploiting the Prior of Generative Time Series Imputation
Imperial College London, Department of Electronic and Electrical Engineering(帝国理工学院伦敦校区电子与电气工程系);University of Science and Technology of China(中国科学技术大学);Tsinghua University, Department of CST(清华大学计算机科学与技术系);Shengshu AI(盛数人工智能)
RefineBridge: Generative Bridge Models Improve Financial Forecasting by Foundation Models
Department of Electrical and Electronic Engineering, Imperial College London, UK(帝国理工学院伦敦分校电子与电气工程系);Department of CST, Tsinghua University, Beijing, China(清华大学计算机科学与技术系)
Dequantified Diffusion-Schr{ö}dinger Bridge for Density Ratio Estimation
The School of Mathematics, South China University of Technology, Guangzhou 510006, China(华南理工大学数学学院);The School of Electronic and Information Engineering, South China University of Technology, Guangzhou 510006, China(华南理工大学电子与信息工程学院);The Department of Electrical Engineering, Columbia University, New York, NY 10027, USA(哥伦比亚大学电气工程系)
Force-Guided Bridge Matching for Full-Atom Time-Coarsened Dynamics of Peptides
Tsinghua University(清华大学);Institute for AI Industry Research (AIR), Tsinghua University(清华大学人工智能产业研究院(AIR));Gaoling School of Artificial Intelligence, Renmin University of China(中国人民大学高瓴人工智能学院)