04. 03. 2026, 13:00

Deep learning in the search for dark matter: an overview

Deep learning has quickly become a valuable tool in the quest to understand dark matter, helping researchers explore faint signals across a range of experiments, from high-energy colliders to direct and indirect searches. By sifting through vast datasets and uncovering subtle patterns, these techniques can reveal signs of dark matter that traditional approaches might miss. In this talk, I will provide a broad overview of how deep learning and related machine-learning approaches are reshaping the search for dark matter, with an emphasis on key breakthroughs and ongoing challenges. I will also outline several promising future directions.

23. 03. 2026, 13:00

Super-resolution microscopy for dynamic structural cell biology

Optical superresolution microscopy ideally complements electron microscopy for studying protein structure and dynamics in cells. The high contrast of the fluorescence label allows investigating individual structures in various conformational states without averaging, and live cell superresolution microscopy can directly probe dynamics. I will showcase the power of super-resolution microscopy by discussing two projects from the lab. First, I will show how single-molecule localization microscopy allowed us to understand the structural organization and dynamics of the complex machinery that drives endocytosis, an essential cellular process for the uptake of molecules. Second, I will show how we use MINFLUX to measure conformational changes of proteins in the living cell.

The lecture will be held in English.

24. 03. 2026, 11:00

Controlling light with light

Our planet is crisscrossed with optical fibres that carry enormous amounts of information every second at the speed of light. These data travel as extremely short laser pulses to data centres, where they are converted into electrical signals, stored, processed, and then transformed back into light signals that continue their journey through the optical network to internet users. Because the ever growing volume of data flowing across the web is driving a rapid increase in electricity consumption in data centres – primarily for converting light into electricity and back again – scientists are searching for ways to replace electricity with light inside data centres and computers. In other words: how can we redirect light using light itself, and how can we compute using light? Over the past decade, this challenge has given rise to the rapidly advancing field of silicon photonics, which integrates seamlessly with modern microelectronics. In the lecture, I will present the paths and pitfalls of a more unconventional approach, in which we harness the self organising properties of soft matter and liquid crystals to steer light with light.

The lecture will be held in Slovenian.

25. 03. 2026, 13:00

Metals under the pressure of sustainable development

Due to the growth of the global population and the rapid development of modern and green technologies, the consumption of metals is continuously increasing. The life cycle of metals begins with the extraction of natural resources, continues with processing and the manufacture of metal products, and concludes with their recycling or disposal. Corrosion plays a key role in this cycle, as most metals gradually deteriorate when exposed to the environment. Corrosion products form on the surface, differing from the base metal in their chemical, physical, and mechanical properties. As a result, the functional performance of metal products or structures declines, along with their energy efficiency, productivity, and ultimately their safety. Metal corrosion also represents a major economic challenge. Therefore, the efficient use of natural resources and the extension of the service life of metallic materials are essential conditions for a more sustainable future. In the lecture, I will present the main methods of corrosion protection, which play a central role in extending product lifetime and reducing the environmental impact of industry.

The lecture will be held in Slovenian.

26. 03. 2026, 13:00

How AI learns to see, generate, and judge fairly

In the first part of the lecture, I will explore how we teach AI to generate videos without relying on detailed annotations or object‑specific labels. By training on collections of similar videos – such as faces or human bodies – the model learns to generalize across an entire category. Building on this idea, we developed a Learnable Game Engine (LGE) that learns from simple monocular videos to keep track of scenes and objects and to re‑render them from different viewpoints. Much like a real game engine, it captures basic physics and logic, allowing users to control the scene or guide virtual agents through high‑level language instructions. The second part of the lecture turns to the safety and fairness of generative AI. Most existing approaches look only for predefined types of bias, but real‑world systems can exhibit unexpected ones. To address this, we introduce OpenBias, a method that uncovers and measures previously unknown biases in text‑to‑image models without relying on any preset list. Our experiments show that OpenBias aligns well with established methods and with human judgment, offering a more flexible way to assess fairness in generative systems.

The lecture will be held in English.

Last lectures