A field guide to spotting what changed from orbit

What it does

This repository is a curated “awesome” list that compiles resources for remote sensing change detection. It organizes datasets, tools, methods, review papers, and competitions into dense markdown tables, linking directly to papers and data sources. The goal is to serve as a centralized index for a field that spans satellite optics, SAR, and aerial imagery.

The interesting bit

The dataset tables are unusually specific: each row lists the sensor (e.g., Sentinel-2, Maxar, Jilin-1), spatial resolution, image dimensions, geographic coverage, and task type—whether binary change detection, semantic segmentation, or disaster damage assessment. The table of contents also breaks methods down into modern deep-learning taxonomies like foundation models, diffusion models, and transformers, making the repo a quick pulse check on where the research momentum is.

Key highlights

• Dataset coverage spans optical, multi-modal, and SAR sources, with entries from 2020 through 2026.
• Task types include binary change detection, semantic change detection, building damage assessment, and remote sensing image change captioning.
• Methods are categorized into deep-learning branches—foundation models, diffusion models and GANs, transformers, and CNNs—alongside traditional approaches.
• Includes dedicated sections for review papers, competitions, and disaster-response satellite data resources.

Caveats

• The compilation is strictly a directory; there is no comparative analysis or guidance on which dataset or method fits a given problem, so you still need to read the original papers.
• The README consists almost entirely of wide markdown tables, which means browsing on mobile or narrow windows will involve significant horizontal scrolling.

Verdict

Worth bookmarking if you work in geospatial machine learning and need to locate a relevant dataset or catch up on recent architectural trends. Skip it if you are looking for runnable code or a tutorial-style walkthrough; this is a reference shelf, not a toolbox.