The main goal of the FIPU Lab is to encourage and develop research activities in information technologies. Research at the lab focuses on creating new and improving existing technologies, algorithms, and software solutions.
The FIPU Lab actively takes part in national and international conferences, including SISY, MIPRO, and THECUC. SISY, MIPRO i THECUC.
This paper explores the detection and classification of moving objects at the pixel level using only a front-facing camera in the CARLA simulator, while evaluating generalization across different cameras using Mask R-CNN and YOLOv7 for instance segmentation.
Goran Oreški
Lucija BabićThis study presents an automated framework that transforms UML class models into open-source end-user databases with a web interface, facilitating easier use and learning for less-experienced software engineers, especially students.
Alan Bubalo
Nikola TankovićThis research introduces an automated method for generating BPMN models from text descriptions using advanced NLP and deep learning tools, achieving up to 96% accuracy with GPT-4 in visualizing complex business processes.
Josip Tomo LicardoNikola Tanković
Darko EtingerThis research focuses on classifying visual perception brain signals using an OpenBCI headset to capture the EEG signals from the participants.
Alesandro Žužić
Robert ŠajinaNikola TankovićThis research analyzes the effect of different loss functions on the performance of semantic segmentation across 23 different traffic-related object classes.
Goran OreškiSimon AničićThis research evaluated the performance of 14 novel object detection models on a traffic dataset to reveal significant differences between the selected models. The goal was to generate recommendations for best-performing models for researchers coming into the domain.
Lucia VareškoGoran Oreški