For the sixth time, around 50 researchers came together for the Statistical Learning for Signal and Image Processing (SLSIP) workshop, which this year took place in the idyllic coastal town of Hanko, Finland, from May 29 to 31 2025. Once again, the Robust Data Science Group (RDS) participated, this time with six attendees. Thanks to the support of emergenCITY, I had the great opportunity to join as a newly appointed research associate, just three months into my position, and to gain my first international conference experience.

The cover picture of the presentation can be seen. Red curtains hang to the left and right of it, with a flipchart in the foreground.
© Christoph Löser

The presentation "False Discovery Rate Control for Complex-Valued High Dimensional Data with Applications in Array Signal Processing."

The workshop offered not only engaging talks from leading experts, but also the chance for direct, in-depth discussions and valuable exchanges on concrete research challenges. One of the highlights for me was the opportunity to present my current work on emergenCITY’s Scout Robot during a talk given by Fabian Scheidt (RDS).

Fabian opened his talk by introducing the complex-valued T-Rex selector, an extension of the T-Rex framework, and the first method to enable statistically error-controlled variable selection in high-dimensional complex-valued data. Following this, I presented how the T-Rex selector is practically applied within a novel localization algorithm that supports the Scout Robot in locating individuals in disaster scenarios.

Two people on bicycles can be seen from behind on a narrow path. There is a forest around them.
© Christoph Löser

Exploring the Hanko countryside by bike.

Beyond the outstanding academic contributions, including presentations from researchers I had previously only known from textbooks, I was particularly impressed by the open and collegial atmosphere of the workshop. Even outside the lecture halls, we made the most of our time: during lunch breaks, we explored the surroundings of Hanko. One especially memorable excursion was a bike trip to the southernmost point of Finland, where we crossed narrow wooden planks and rugged terrain to reach our destination.

A small ship is moored at a jetty. Blue water stretches out behind it, reflecting the sunset. Further back, more ships and parts of the city can be seen.
© Christoph Löser

The port of Hanko.

Looking back, the SLSIP workshop in Hanko was not only scientifically enriching, but also a formative experience early in my academic career. The exchange with international colleagues, the opportunity to present my own research in this setting, to discuss personal ideas and research questions, and to share the experience as a team with the RDS Group all left a lasting impression on me, thank you, emergenCITY!

Author:

Christoph Löser is a research associate at TU Darmstadt with the Robust Data Science Group (RDS) and the institute of Artificial Intelligent Systems in Medicine (KIS*MED). As part of his work, he is involved in the CPS program of emergenCITY.