TÖV606M - Machine Learning for Earth Observation powered by Supercomputers

Welcome to the course repository! This repository contains all lab notebooks and materials for building a complete ML pipeline for Earth Observation using Sentinel-2 satellite imagery and HPC infrastructure.

📚 Course Overview

• Credits: 6 ECTS
• Semester: Spring 2025-2026 (January - April)
• Modality: Mixed in-person/online (Iceland + Germany)
• HPC Resources: JURECA at Jülich Supercomputing Centre

🎯 Learning Outcomes

By completing this course, you will learn to:

• Access and manage HPC resources (Judoor, JURECA, SLURM)
• Acquire and preprocess satellite imagery (Sentinel-2 via Google Earth Engine)
• Build and train deep learning models for land cover classification
• Evaluate model performance using industry-standard metrics
• Fine-tune geospatial foundation models (TerraTorch, Prithvi)

🗂️ Repository Structure

├── notebooks/iceland-ml/     # Lab notebooks (work here!)
│   ├── lab1_judoor_hpc_access.ipynb
│   ├── lab2_jupyter_jsc_git.ipynb
│   ├── lab3_1_data_acquisition.ipynb
│   ├── lab4_1_preprocessing.ipynb
│   ├── lab4_2_preprocessing_patches.ipynb
│   ├── lab5_1_baseline_training.ipynb
│   ├── lab5_2_model_evaluation.ipynb
│   └── lab6_finetune.ipynb
├── docs/                     # Documentation
│   ├── iceland-ml/           # Course guides
│   └── units/                # Lab-specific docs
├── requirements.txt          # Python dependencies
└── pyproject.toml           # Project configuration

🚀 Getting Started

Step 1: Create Required Accounts

1. Judoor Account (do this first - takes 1-2 days): https://judoor.fz-juelich.de/register
2. Google Earth Engine (needed for Lab 3): https://earthengine.google.com/signup

Step 2: Join the Training Project

After your Judoor account is approved, join the training2600 project.

Step 3: Follow the Labs

Work through the notebooks sequentially starting with Lab 1.

📓 Lab Schedule

Lab	Topic	Notebook
1	Judoor & HPC Access	lab1_judoor_hpc_access.ipynb
2	Jupyter-JSC & Git	lab2_jupyter_jsc_git.ipynb
3	Sentinel-2 Acquisition	lab3_gee_sentinel2_acquisition.ipynb
4	Data Preprocessing	lab4.1_preprocessing.ipynb
5	Patch Extraction	lab4.2_preprocessing_patches.ipynb
6	Model Training	lab5.1_baseline_training.ipynb
7	Model Evaluation	lab5.2_model_evaluation.ipynb
8	TerraTorch Fine-tuning	lab6_finetune.ipynb

📖 Documentation

• Course Overview - Detailed course information
• Lab Summary - Quick reference for all labs

🔧 Prerequisites

• Python programming (intermediate level)
• Basic machine learning concepts
• Linux command line basics
• SSH client installed on your machine

💬 Getting Help

• Slack Channel: Check with instructors for invite link
• Email: [email protected]
• Office Hours: By appointment (3 days advance notice)

📧 Instructors

• Gabriele Cavallaro ([email protected]) - Course Lead
• Rocco Sedona ([email protected]) - Technical Lead
• Samy Hashim ([email protected]) - Lab Instructor
• Ehsan Zandi ([email protected])

📄 License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.

---

Good luck with your learning journey! 🛰️🌍