idw – Informationsdienst Wissenschaft
Nachrichten, Termine, Experten
Nachrichten, Termine, Experten
The German Research Foundation (DFG) has admitted two group leaders from the Max Planck Institute for Informatics (MPI-INF) in Saarbrücken into the prestigious Emmy Noether Programme: Dr. Jan Eric Lenssen and Dr. Karol Węgrzycki. Their research focuses on algorithms and machine learning. The Emmy Noether Program offers outstanding early-career scientists the opportunity to qualify for a university professorship by independently leading a research group for a period of six years. Together, the two researchers will receive around 3.7 million euros in funding.
Dr. Jan Eric Lenssen will receive approximately 1.9 million euros in funding to establish the research group Spatial Modeling and Reasoning. The group aims to enable computers—especially AI systems based on machine learning—to interpret visual scenes in a way that resembles human understanding: In other words, the goal is to extract as much relevant information as possible about the environment from only a few visual "impressions", or data points. To this end, the group will pursue two intertwined lines of research:
Lenssen and his team will focus first on developing efficient spatial representations that can serve as a foundation for machine learning. This involves studying methods that allow a computer to infer the underlying causes of an observed scene, for example, reconstructing three-dimensional shapes, camera positions, and lighting conditions from a two-dimensional image. The second focus concerns the development of models that perform spatial reasoning across these representations. This includes the system localizing itself and objects in a 3D environment, identifying objects, recognizing interaction potentials, and forming hypotheses about unobserved parts of the surroundings. “Progress in these areas is particularly relevant for applications in cyber-physical systems that interact with their environments, such as robots or autonomous vehicles. Generative 3D modeling could also benefit,” explains Lenssen.
A recent achievement from Lenssen’s research in this area is the development of novel Spatial Reasoning Models (SRMs). These generative models gradually uncover complex visual relationships when generating images by relating pictured elements to each other and identifying dependencies. In doing so, Lenssen’s team significantly reduced “hallucinations”, or plausible but incorrect image content. To evaluate their performance, the researchers tested SRMs on difficult Sudoku puzzles as one of their testing methods: SRMs solved more than 50 percent correctly, whereas classical image-generating models largely failed.
Jan Eric Lenssen earned his PhD in February 2022 at TU Dortmund University under Professor Heinrich Müller and subsequently joined the MPI for Informatics as a postdoctoral researcher. Since June 2023, he has been a Senior Researcher at the Institute and leads the Geometric Representation Learning group in the Computer Vision and Machine Learning department, headed by Director Professor Bernt Schiele. For his dissertation titled “Differentiable Algorithms with Data-driven Parameterization in 3D Vision,” he received the 2024 PhD Award from the European Association for Computer Vision. In 2025, he was honored with the German Pattern Recognition Award from the German Association for Pattern Recognition (DAGM).
Dr. Karol Węgrzycki will receive a total of 1.8 million euros in funding to establish his research group. The aim of his group is to develop new solutions to so-called “packing problems,” which are widespread and highly relevant optimization tasks.
Packing problems arise in many practical situations. Examples include the efficient storage of luggage items of different sizes, shapes, or monetary values in a cargo hold to maximize transport capacity; cutting a block of raw material such as a tree trunk to minimize waste; and stacking containers of the same size on a ship to minimize loading and unloading times at successive ports.
The methods developed to date are often optimized for such specific problems, which limts their use in broader applications. Węgrzycki now aims to adapt solution techniques from other areas of computer science in order to tackle previously inaccessible problems or to determine the best variant for applications that could previously only be solved approximately. To this end, he and his group plan to use solution methods from cryptography, graph theory, and parameterized algorithms in the fields of optimization, algorithmic geometry, operations research, and logic. They hope that this rigorous, cross-disciplinary analysis will foster collaboration between fields and deepen understanding of complex problems.
Karol Węgrzycki studied physics and computer science at the University of Warsaw, where he earned his degree. In 2021, he received his doctorate with a thesis on “Provably Optimal Dynamic Programming,” which was recognized by the Polish Prime Minister as one of the 25 best doctoral theses in the country. Since 2021, he has worked as a postdoctoral researcher, and in 2024 he became a group leader in the Algorithms and Complexity department, headed by Director Professor Danupon Nanongkai, at the MPI for Informatics.
Contact:
Bertram Somieski
Joint Administration
Max Planck Institute for Informatics
Max Planck Institute for Software Systems
Tel: +49 681 9302-5710
E-Mail: somieski@mpi-klsb.mpg.de
Philipp Zapf-Schramm
Max Planck Institute for Informatics
Tel: +49 681 9325 4509
E-Mail: pzs@mpi-inf.mpg.de
https://geometric-rl.mpi-inf.mpg.de/people/lenssen.html Webpage Jan Eric Lenssen
https://www.mpi-inf.mpg.de/departments/algorithms-complexity/people/current-memb... Webpage Karol Węgrzycki
Dr. Jan Eric Lenssen
Source: Philipp Zapf-Schramm
Copyright: MPI für Informatik
Dr. Karol Węgrzycki
Source: Bertram Somieski
Copyright: MPI für Informatik
Criteria of this press release:
Journalists
Information technology
transregional, national
Contests / awards, Research projects
English
Dr. Jan Eric Lenssen
Source: Philipp Zapf-Schramm
Copyright: MPI für Informatik
Dr. Karol Węgrzycki
Source: Bertram Somieski
Copyright: MPI für Informatik
You can combine search terms with and, or and/or not, e.g. Philo not logy.
You can use brackets to separate combinations from each other, e.g. (Philo not logy) or (Psycho and logy).
Coherent groups of words will be located as complete phrases if you put them into quotation marks, e.g. “Federal Republic of Germany”.
You can also use the advanced search without entering search terms. It will then follow the criteria you have selected (e.g. country or subject area).
If you have not selected any criteria in a given category, the entire category will be searched (e.g. all subject areas or all countries).
