Workshops

This workshop focuses on introducing the concept of federated learning and demonstrating how it differs from traditional centralized machine learning through hands-on model building.

During the workshop, we will use Jupyter Notebook, which allows us to build and test machine learning models in an interactive and accessible way. We will start by creating a centralized model, trained on data collected in a single dataset, to understand the traditional approach to model training. Then, we will move to the implementation of a federated model using TensorFlow Federated (TFF), which enable the simulation and construction of distributed learning systems. Step by step, we will demonstrate how such a model is built, how communication between the server and clients works, and how parameter aggregation is performed. We will discuss the key differences between centralized and federated approaches — especially in terms of data privacy, system architecture, training process, and performance. Finally, we will compare the results of both models and analyze when federated learning makes the most sense.

The workshop will be lead by dr Dominika Cywicka and dr hab. Pawel Netzel, prof. URK

Workshop's presentation/script:

This workshop will focus on the possibilities of analyzing raster data. Raster data is a spatial representation of remote sensing information. Participants will learn the fundamentals of raster algebra during the workshop. They will also learn how to use a raster calculator for advanced spatial analysis. With just one tool, participants will progress from simple operations to complex spatial analysis. All of this can be achieved without working in a GIS system. The only tool used for calculations is the standalone raster calculator, plMapcalc. Participants will also use QuantumGIS, but only for visualizing the results. All of the tools used in the workshop are free, open-source programs.

The workshop will be lead by dr hab. Pawel Netzel, prof. URK and dr Jacek Ślopek

Workshop's presentation/script (it needs plTools):

This workshop offers a concise introduction to quantum computing with a strong focus on practical applications in remote sensing. It is designed for participants without prior experience in quantum technologies. We begin with a brief overview of quantum computing fundamentals and the current landscape of quantum technologies, emphasizing concepts most relevant to machine learning and data analysis. Participants will then be introduced to key ideas in quantum machine learning (QML), including quantum kernels and parameterized quantum circuits, as tools for analyzing remote sensing data. The second part of the workshop is a hands-on tutorial, where participants will gain practical experience using quantum computing libraries to implement simple quantum machine learning models. In addition to basic QML circuit design, we will focus on the description, interpretation, and insights that can be drawn from the proposed circuits, helping participants understand both their structure and their practical implications for remote sensing tasks. We will also explore hybrid quantum-classical architectures combining neural networks with quantum layers. By the end of the workshop, attendees will have both a conceptual understanding of quantum computing in the context of remote sensing and practical coding experience to build upon in their own work. 

What You’ll Learn

• Introduction to Quantum Computing and the Landscape of Quantum Technologies
• Fundamentals of Quantum Machine Learning (QML)
• Getting Started with Quantum Computing Libraries
• Quantum Kernel SVM and Parameterized Quantum Circuits (PQC): Basic circuit design, interpretation and insights
• Hybrid Architectures: Combining Classical Neural Networks with Quantum Layers

The workshop will be lead by dr Artur Miroszewski

Workshop's presentation/script:

Workshop participants will learn about selected tools from the GDAL library. They will use these tools to process remote sensing satellite and aerial imagery, georeference data, and reproject it. The workshop will use open raster data. Participants will learn to work with the command line and use the GDAL raster calculator to calculate spectral indices based on data recorded by remote sensing satellites.

The workshop will be lead by dr Jacek Ślopek

Workshop's presentation/script: