wenhwu/awesome-remote-sensing-change-detection

A comprehensive and up-to-date compilation of datasets, tools, methods, review papers, and competitions for remote sensing change detection.

What it solves

This project addresses the challenge of finding and organizing high-quality resources for remote sensing change detection—the process of identifying differences in land cover or objects (like buildings or crops) over time using satellite or aerial imagery. It serves as a central hub for researchers and developers to discover datasets, tools, and methodologies without having to search across fragmented academic sources.

How it works

It acts as a curated, community-maintained "awesome" list. The repository categorizes vast amounts of information into structured tables and lists, including:

  • Datasets: Detailed breakdowns of optical, multi-modal, and SAR (Synthetic Aperture Radar) datasets, including information on image resolution, location, and specific tasks like semantic or binary change detection.
  • Tools: Links to open-source software toolboxes for training, inference, and benchmarking.
  • Methods: A collection of approaches ranging from traditional algorithms to modern deep learning techniques like foundation models, diffusion models, and transformers.
  • Additional Resources: Pointers to review papers, competitions, and satellite data resources for disaster response.

Who it’s for

This project is for researchers, data scientists, and developers working in remote sensing, computer vision, or geospatial analysis who need a reliable, up-to-date reference for building or evaluating change detection systems.

Highlights

  • Extensive catalog of datasets categorized by task (e.g., building damage, land cover, flood detection).
  • Curated list of open-source toolboxes for reproducing and developing change detection models.
  • Coverage of modern AI techniques including foundation models, diffusion models, and transformers.
  • Regularly updated with the latest research papers and competition benchmarks.

Related

  • Project
    datopian/portaljs

    PortalJS is an open‑source, AI‑assisted framework that lets you describe a data‑portal in natural language, then automatically recommends an architecture, scaffolds a plain Next.js site, and adds datasets, visualisations, and metadata. It works with modern lakehouse stacks (Git + Cloudflare R2 + Parquet + DuckDB) or classic CKAN back‑ends, and the generated code is fully editable with no lock‑in.

  • Project
    mcordts/cityscapesScripts

    The repository provides Python tools for downloading, visualizing, converting, and evaluating the Cityscapes street‑scene dataset, covering semantic, instance, panoptic, and 3‑D detection tasks.

  • Project
    DeepInsight-AI/DeepBI

    An AI-native data analysis platform that uses LLMs to allow users to explore, query, and visualize data from multiple sources via natural language conversations.

  • Project
    NVIDIA-NeMo/DataDesigner

    NVIDIA NeMo DataDesigner is an open‑source Python library for building synthetic‑data pipelines. It lets you define columns (categorical samplers, LLM‑generated text, image/audio generators), specify cross‑field dependencies, validate rows with Python/SQL/LLM judges, and run the whole workflow via a CLI or Python API. Plugins add extra data sources and generation capabilities. The tool is model‑agnostic (supports NVIDIA Build, OpenAI, OpenRouter, etc.) and is aimed at creating production‑grade multimodal datasets for AI training.