Skip to main content

Spring Seminar Series for Publication Award Winners at HUN-REN SZTAKI

3 March 2025

As per tradition, the seminar series featuring publication award recipients is held again this spring.

"Sharing research findings and promoting scientific dialogue are of utmost importance to HUN-REN SZTAKI" stated Prof. Csaba Benedek, one of the main organisers of the event and the deputy director of research coordination at HUN-REN SZTAKI.

Dániel Horváth, a researcher at HUN-REN SZTAKI, gave the first lecture of the series, titled "Bridging the Gap: Sim2Real Transfer Learning for Computer Vision in Robotics." The presentation lies at the intersection of artificial intelligence, robotics, and image processing, focusing on how synthetic images can be generated to extract knowledge that is also relevant in the physical world.

"Generating synthetic training samples is particularly important because modern deep-learning-based neural networks rely on vast amounts of data. For instance, in robotic assembly, accurately identifying parts is essential, but capturing and labelling real-world images is time-consuming and, in several cases, unfeasible," explained Dániel Horváth.

Based on the research findings presented in the lecture, a practical industrial framework is currently under development, led by Dániel Horváth, with the involvement of SZTAKI researchers and master's students. This framework aims to offer solutions for a wide range of industrial image-processing tasks—from object detection to six-degree-of-freedom pose estimation.

Publication Award-Winning Papers: