10.3K ·GitHub
What is this?
A curated list of techniques for deep learning with satellite and aerial imagery, covering architectures, models and algorithms for tasks like classification, segmentation and object detection.
A curated list of techniques for deep learning with satellite and aerial imagery, covering architectures, models and algorithms for tasks like classification, segmentation and object detection.So it is not a library I install — it is a collection of references?Right: it is a single curated README of links to papers and GitHub repositories, not installable software. You browse it and follow the links to the actual code.
Right: it is a single curated README of links to papers and GitHub repositories, not installable software. You browse it and follow the links to the actual code.How is the material organized?By technique and task under a Techniques section, with subsections like classification, segmentation of water, coastlines, rivers and floods, building and road extraction, change detection and pansharpening.
By technique and task under a Techniques section, with subsections like classification, segmentation of water, coastlines, rivers and floods, building and road extraction, change detection and pansharpening.How do I find something specific in such a long list?The README tells you to search the page directly: use Command+F on Mac or CTRL+F on Windows to look for a technique name such as 'SAM'.
The README tells you to search the page directly: use Command+F on Mac or CTRL+F on Windows to look for a technique name such as 'SAM'.What are some example techniques it lists?Concrete repos like a PyTorch U-Net for road and building extraction, cloud detectors, change detection networks, and pix2pix models that convert satellite images into map images.
Concrete repos like a PyTorch U-Net for road and building extraction, cloud detectors, change detection networks, and pix2pix models that convert satellite images into map images.Does it cover specific satellite programs like Sentinel-2?Yes. For example it lists KappaMask, a cloud detector built for Sentinel-2 Level-1C and Level-2A products, and Sentinel-related topics are tagged on the repo.
Yes. For example it lists KappaMask, a cloud detector built for Sentinel-2 Level-1C and Level-2A products, and Sentinel-related topics are tagged on the repo.Is each entry just a paper citation, or does it point to runnable code?Most entries link to a GitHub repository, very often the official implementation of a paper (U-Net road extraction, SAM-based farmland mapping, Siamese change detection networks), so you can clone and run them.
Most entries link to a GitHub repository, very often the official implementation of a paper (U-Net road extraction, SAM-based farmland mapping, Siamese change detection networks), so you can clone and run them.Are competition-winning approaches included?Yes. The list references the xView object detection challenge top entries and the winners of the STAC Overflow competition on mapping floodwater from radar imagery.
Yes. The list references the xView object detection challenge top entries and the winners of the STAC Overflow competition on mapping floodwater from radar imagery.Who gets the most out of this list?Researchers and engineers who already know deep learning basics and want to apply them to remote sensing data. Entries assume familiarity with terms like U-Net, segmentation and multispectral imagery.
Researchers and engineers who already know deep learning basics and want to apply them to remote sensing data. Entries assume familiarity with terms like U-Net, segmentation and multispectral imagery.What are similar projects?deep_learning_object_detection, awesome-semantic-segmentation, deep-learning-for-image-processing, CVPR2024-Paper-Code-Interpretation, VisionHOPE, Deep-Learning-Papers-Reading-Roadmap, fashion-mnist, CVPR2026-Papers-with-Code
deep_learning_object_detection, awesome-semantic-segmentation, deep-learning-for-image-processing, CVPR2024-Paper-Code-Interpretation, VisionHOPE, Deep-Learning-Papers-Reading-Roadmap, fashion-mnist, CVPR2026-Papers-with-CodeWhich topics is it in?convolutional-neural-networks, image-segmentation, gis, geospatial, datasets, deep-neural-networks, image-classification, object-detection, papers, dataset, reference, computer-vision
convolutional-neural-networks, image-segmentation, gis, geospatial, datasets, deep-neural-networks, image-classification, object-detection, papers, dataset, reference, computer-vision- Size
- 29.4 MB
- Stars
- 10.3K
- Age
- 8y
- Commits · 90 Days