Dr. Joanna Komorniczak and Dr. Tomasz Szandała from WIT are among the 16 researchers from Wrocław University of Science and Technology who received this year's scholarships from the Minister of Science and Higher Education. This is the second-best result in Poland. The awards are granted to early-career researchers with outstanding achievements in scientific research.
This year, 1,640 applications were submitted to the competition and evaluated by a dedicated committee using a points-based system within individual scientific and artistic disciplines. The winners were selected according to the principle of equal treatment of scientific and artistic disciplines, which meant awarding a similar number of scholarships in each discipline. As a result, researchers competed within their respective disciplines rather than across entire fields.
Ultimately, scholarships were awarded to 230 outstanding early-career researchers representing 54 scientific and artistic disciplines. As many as sixteen of them are from Wrocław University of Science and Technology. This is the second-best result in Poland.
The competition winners will receive scholarships of PLN 5,390 per month for three years.
The scholarships were awarded to:
Dr. Joanna Komorniczak
In her research, Dr. Joanna Komorniczak has developed methods for analysing variability in data streams and investigated the potential of various meta-features in detecting concept drifts.
The results of this research have been published in international journals and presented at conferences.
Her research interests include the analysis of non-stationary data streams, detection and quantification of distribution changes, difficulty evaluation of supervised machine learning tasks, and open-set recognition.
In the future, she plans to develop methods and experimental protocols for data streams in which previously unknown classes appear.
Dr. Tomasz Szandała
He received the Minister of Science and Higher Education scholarship for his research on trustworthy artificial intelligence.
His work focuses on AI explainability and deepfake detection, i.e. images manipulated by artificial intelligence. Among the recognised achievements were two of his methods. PRISM makes it possible to verify how a model recognises objects and whether it analyses them similarly to a human. Fair-FLIP, developed together with Prof. Silvia Giordano and Dr. Omran Ayoub, improves the effectiveness of deepfake detection across different ethnic groups, genders and age groups, while requiring only minimal modifications to the model.
Previously, Dr. Szandała completed a one-year Swiss Government Excellence Scholarship at SUPSI University in Lugano. For the past 10 years, he has combined scientific research with work in the IT industry as a DevOps and Cloud consultant. In the future, he plans to develop AI solutions supporting the creation of more reliable systems while preserving the key role of humans.
