techniques
Techniques for deep learning with satellite & aerial imagery
What is techniques?
Techniques for deep learning with satellite & aerial imagery
Overview of techniques
What it is
This repository provides an overview of deep learning techniques for satellite and aerial imagery processing. It covers architectures, models, and algorithms for tasks such as classification, segmentation, and object detection. The content is organized as a browsable list of techniques and linked resources.
Key features
- Covers classification, segmentation, object detection, regression, cloud detection, change detection, time series analysis, crop classification, generative networks, autoencoders, few-shot learning, self-supervised learning, SAR, explainable AI, large vision and language models, AI agents, and foundational models.
- Includes references to numerous open-source projects and papers for each technique.
- Provides a searchable page (Command+F or Ctrl+F) to quickly find specific topics like 'SAM'.
- Lists example datasets such as UC Merced, EuroSAT, and Urban Atlas.
Use cases
- Land cover and land use classification from satellite imagery.
- Building detection, roof type recognition, and urban analysis.
- Crop classification, yield forecasting, and vegetation monitoring.
- Cloud detection and removal, change detection, and flood susceptibility mapping.
- Object detection for vessels, aircraft, and infrastructure.
- Self-supervised and few-shot learning for remote sensing.
Getting started
See the repository README for usage instructions.
License and pricing
Licensed under Apache-2.0. Open source.
Based on the project's public README and repository metadata.