ABOUT THE GROUP
The Research Group for Computing and Human-Computer Interaction (IG03) operates within the Faculty of Electrical Engineering, Computer Science and Information Technology at Josip Juraj Strossmayer University of Osijek, Croatia. The group brings together researchers and experts dedicated to advancing computational methods at the intersection of artificial intelligence, image processing, and immersive technologies, with a strong emphasis on real-world applicability and societal impact.
The group's main objective is to develop robust, interpretable, and scalable intelligent systems by combining deep learning, medical and visual data analysis, and generative modelling. A significant part of IG03's research focuses on advanced methods for medical image analysis based on CT and MRI, targeting automated segmentation, quantitative analysis, and extraction of clinically relevant biomarkers. The group also develops extended reality (XR) systems for medical education, surgical training, and procedural planning, as well as intelligent assistive systems for blind and visually impaired individuals.
Beyond medical and assistive applications, IG03 investigates the use of large language models for Croatian and other low-resource languages, with applications in education, language teaching, and domain-specific knowledge systems. The group further explores multi-agent architectures for autonomous task orchestration and explainable AI that provides users with transparent and comprehensible decision-making processes. Through its work, IG03 aims to contribute to scientific excellence and the broader community through publications, open-source tools, data repositories, and interdisciplinary collaborations.
Active in both fundamental research and applied development, IG03 emphasises interdisciplinary approaches, bridging artificial intelligence with medicine, education, linguistics, law, and accessibility to address real-world challenges across multiple domains.
KEY RESEARCH AREAS
- Medical image analysis and clinical AI
- Extended reality (XR) for education and medicine
- Generative AI and synthetic data
- AI and language
- Intelligent autonomous and interactive systems
RESEARCH AREAS
IG03 focuses its research on several interrelated themes spanning medical AI, immersive technologies, generative modelling, natural language processing, and intelligent interactive systems. The group develops advanced deep learning methods for cardiac and anatomical segmentation from CT and MRI, investigates data efficiency and bias in medical image analysis, and builds XR-based simulation and navigation systems for surgical training and education. It also conducts research on generative models for synthetic data creation, including diffusion models and GANs applied to medical imaging, historical photo restoration, and data augmentation for challenging detection scenarios. Another core area involves large language models adapted for Croatian and low-resource languages, with applications in education, glottodidactics, legal text processing, machine translation, and retrieval-augmented generation. The group further develops multi-agent systems for autonomous orchestration and intelligent autonomous and interactive systems that combine computer vision, speech, and adaptive behaviour to support users in real and virtual environments.
PROFESSIONAL EXPERTISE
IG03 possesses deep professional expertise that bridges advanced research with practical applications across healthcare, education, and assistive technology. Its researchers are experienced in developing end-to-end AI pipelines — from synthetic data generation and model training through to deployment in interactive real-time systems. The group combines strong competencies in medical image processing and computer vision with expertise in immersive application development, language model adaptation, and the design of intelligent multi-agent architectures. This breadth enables IG03 to deliver interdisciplinary solutions that integrate AI with domain-specific needs across clinical, educational, and accessibility contexts.
KEY PROFESSIONAL EXPERTISE
- Deep learning for segmentation, detection, and classification
- Medical image analysis across CT, MRI, ultrasound, and dermatoscopy
- Real-time computer vision, scene segmentation, and image restoration
- Synthetic data generation using GANs, diffusion models, and image-to-image translation
- VR/AR application development, 3D modelling, haptic integration, and spatial registration
- Large language model adaptation, quantisation, retrieval-augmented generation, and domain-specific NLP for legal and low-resource languages
- Multi-agent system design, intent-based orchestration, and autonomous task decomposition
- User experience evaluation, adaptive interaction design, and feedback-driven system optimisation
