This repository serves as a comprehensive survey and curated list of diffusion models applied to time series and spatio-temporal data. It aims to systematically summarize recent advances, providing researchers and practitioners with a centralized resource for papers, code, applications, and reviews in this rapidly evolving field.

How It Works

The repository categorizes diffusion models based on their application to time series data (prediction, generation, imputation, anomaly detection, classification, causal inference, event prediction, foundation models) and spatio-temporal data. It also includes a section for tabular data applications. The core of the resource is a meticulously compiled list of research papers, each linked to its corresponding code and further details, facilitating a deep dive into specific methodologies.

Quick Start & Requirements

This repository is a curated list of research papers and does not have a direct installation or execution command. Users can access a comprehensive, categorized list of papers and related resources via a Google Sheet linked in the README.

Highlighted Details

• Extensive categorization of diffusion model applications across various data types and tasks.
• Links to official code repositories for many cited papers.
• Includes Chinese interpretations and discussions of the survey paper.
• Regularly updated with new research in the field.

Maintenance & Community

The project is maintained by its authors and welcomes community contributions through issues or pull requests for missed resources or corrections. A citation for the survey paper is provided.

Licensing & Compatibility

The repository itself does not specify a license, but it aggregates links to research papers, each with its own licensing and usage terms.

Limitations & Caveats

As a survey and list, this repository does not provide executable code or a unified framework. Users must refer to individual papers for implementation details and dependencies. The breadth of the survey means some niche applications might be less detailed.