Područje interesa Odjela za računalnu inteligenciju je teorija, oblikovanje, primjena i razvoj biološki i lingvistički motiviranih računalnih paradigmi s naglaskom na neuronske mreže, konekcijske sustave, genetičke algoritme, evolucijsko programiranje, neizrazite sustave i hibridne inteligentne sustave koji su temeljeni na ovim paradigmama.
Odjel za računalnu inteligenciju Hrvatske sekcije IEEE poziva Vas na predavanje
"Automating inspection in the manufacturing industry"
koje će održati Juraj Fulir iz Frauenhofer ITWM-a, u ponedjeljak, 25. svibnja 2026. godine, s početkom u 14.00 sati u prostoriji A-201 Fakulteta elektrotehnike i računarstva Sveučilišta u Zagrebu, Unska 3.
Predavanje će se održati na hrvatskom jeziku, a predviđeno trajanje s raspravom je 60 minuta. Predavanje je otvoreno za sve zainteresirane, a posebno pozivamo studente.
Više informacija o predavanju i predavaču pročitajte u opširnijem sadržaju obavijesti.
Sažetak predavanja:
The manufacturing industry is increasingly using AI-based machine vision methods for automated quality control, to increase the speed, robustness and flexibility of manufacturing lines. However, developing AI models in this environment is difficult due to slow data collection, as the defects and anomalies are rare by nature. This can produce inspection systems which are not fit for detecting critical defects that were not observed during the data collection phase.
I will present our approach to this problem which uses computer graphics, through light transport simulation and procedural surface modelling. This allows us to model a digital twin of the product and defects for the generation of missing data. However, the domain gap between synthetic and real data can impede AI development, if data modelling is incorrect. We experimentally study this problem to estimate the severity of the domain gap, to gain practical knowledge about its sources and develop interpretative methods for overcoming it.
However, before AI development can begin, the design of inspection systems is the first critical step. With hundreds of optical and illumination devices to choose from, and multiple ways of positioning them, the task of system construction is challenging. I will present the research field of virtual inspection planning which aims to automate inspection system development. It uses digital twins and various optimization algorithms to find inspection solutions that achieve a complete product inspection. I will conclude the lecture with the open questions in this field and the many opportunities that lie within.
Životopis predavača:
Juraj Fulir is a researcher at Fraunhofer ITWM and a PhD candidate at RPTU Kaiserslautern-Landau, Germany. He works in applied research for the manufacturing and maintenance industries, on the topic of “Automated Visual Inspection Planning using Synthetic Datasets”. His research is suspended between Machine Vision, Computer Graphics and Automated Inspection Planning. He finished his master’s degree in Computer Science at the University of Zagreb, Faculty of Electrical Engineering and Computing, on the topic of Evolutionary Algorithms. After graduation he temporarily started a doctorate in the field of Computer Graphics on the topic of “Interactive Environment Modelling for Serious Gaming”, before switching to RPTU.