12/10/2025 Best Junior Paper Award – 2025
The Director of ÚTIA, Jiřina Vejnarová, in cooperation with the Institute Board, has announced the results of the annual Best Junior Paper Award, a long-standing competition recognizing outstanding scientific work by the Institute’s young researchers. The award aims to support and acknowledge the work of doctoral and postdoctoral researchers as well as students supervised at ÚTIA, whose contributions enhance the
Institute’s
About the competition
The competition is open to works whose Main Author is a Young Author – an employee classified as a doctoral or postdoctoral researcher, or a student supervised by a ÚTIA researcher. Eligible works are publications released in the year of the competition or the previous year, with the Main Author affiliated with ÚTIA. Submissions typically cover a broad range of research topics, from machine learning and optimization to practical applications and socially relevant scientific problems.
This Year’s Edition
An integral part of the competition is the seminar “Young Authors’ Day”, where participants present their work to an expert committee. The committee evaluates both the scientific quality of the work and the authors’ ability to present their research clearly and convincingly.
In this year’s edition, several outstanding contributions were presented, spanning a wide array of contemporary research areas. After the presentations, the committee selected three works that stood out for their originality and scientific impact.
Knowledge Transfer in Deep Reinforcement Learning via an RL-Specific GAN-Based Correspondence Function
IEEE Access, vol. 12 (2024), pp. 177204–177218
This work introduces an innovative method for knowledge transfer in deep reinforcement learning using GAN-based architectures, offering a significant contribution to effective agent training in environments with limited data.
Modular-topology optimization for additive manufacturing of reusable mechanisms
Computers and Structures, vol. 307, 107630
Spatial-temporal source term estimation using deep neural network prior and its application to Chernobyl wildfires
Journal of Hazardous Materials, vol. 448, 137510
The authors present an innovative approach to back-calculating source terms using deep neural networks. The method was applied to real data from the Chernobyl wildfires, providing important insights for environmental safety research.
ÚTIA greatly values the contributions of its young researchers and is proud to support their work. We look forward to the next edition of the competition and to the new scientific contributions it will bring.