This material was prepared for the IEEE GRSS-USC MHI 2023 Remote Sensing Summer School, in conjunction with the 2023 IEEE International Geoscience and Remote Sensing Symposium (IGARSS)
Remote Sensing image Classification and Denoising: A Practical Guide
This repository contains a hands-on, easy-to-follow tutorial that provides an introduction to using Deep Learning techniques for Remote Sensing Image Classification and Denoising.
The code explores various deep learning architectures, from Convolutional Neural Networks (CNNs) to U-NETs, and demonstrates how these can be applied to the two tasks in RGB images from the UCMerced dataset.
The classification jupyter notebook provides an easy-to-follow example for remote sensing scene classification using ResNets.
The denoising jupyter notebook provides an easy-to-follow example for remote sensing scene denoising using UNETs.
- Big Data in Science
- Introduction to Machine Learning
- Applications in remote sensing
- Overview of data sources
- Introduction to Deep Learning
- Deep Learning architectures
- Application of supervised/discriminative learning in remote sensing
- Programming frameworks
- Inverse problems (denoising, super-resolution)
- Generative models (autoencoders and GANs)
- Self-supervised learning
- Class imbalance
- Deep Reinforcement Learning
- Hardware-in-the-loop
- Beyond State-of-the-art