Department of Applied Informatics - Comenius University in Bratislava

🎉 Veľká gratulácia našim kolegom Zuzana Berger Haladová & Viktor Kocur k ich úspešnej habilitácii! 🎓 Prajeme veľa ďalších pracovných a vedeckých úspechov!👏

📣 už ste videli prezentáciu nášho kolegu Petra Lacka na tému Use of AI in Modern Software Development z Innovaite Slovakia konferencie? 👇👀

The Use of AI in Modern Software Development | Peter Lacko (KAI FMFI UK) Enjoy the videos and music you love, upload original content, and share it all with friends, family, and the world on YouTube.

📣 Naši kolegovia sa minulý týždeň zúčastnili 24. ročníka konferencie Kognícia a umelý život 2026 🤖

👉Konferencia vytvára priestor na prezentáciu odborných príspevkov, ich diskusiu v transdisciplinárnom kontexte a nadväzovanie spolupráce s expertmi z rôznych oblastí. Teší nás, že sme mohli byť súčasťou podujatia, ktoré prepája vedu, výskum a inovácie a podporuje výmenu poznatkov naprieč disciplínami 🦾

🤝Ďakujeme organizátorom za príležitosť a všetkým účastníkom za podnetné stretnutia a inšpiratívne diskusie 👇

✨Another week we are starting with the doctoral colloquium✨

Sabeer Saeed with the topic Application of AI in Software Engineering: Advanced Methods in the Development of Large-Scale System Software

Abstract: This doctoral research lies at the intersection of empirical software engineering, requirements engineering, software traceability, Artificial Intelligence for Software Engineering (AI4SE), and software design. The primary objective is to develop advanced methods that bridge industrial needs and academic theory to streamline the development of large-scale system software. On the industrial front, this research focuses on Cindy, a comprehensive toolkit suite developed by Markuz at Gratex, which analyzes entire software repositories by mapping not only source code but also surrounding environmental artifacts. To optimize this suite, we are establishing a robust, normalized taxonomy classification for Cindy’s internal detectors. This framework enhances data standardization, reduces metadata redundancy, and ensures the strict reproducibility of software analysis reports and diagnostics. On the academic front, we address the inherent limitations of rigid, rule-based Abstract Syntax Trees (ASTs) in architectural modeling by leveraging Large Language Models (LLMs) and Agentic AI. This approach automates the generation and intelligent verification of critical system representations, including UML, Use Case, Activity, Flow Chart, and generic System Model diagrams compiled through code-based rendering engines like PlantUML and Mermaid. Furthermore, we are exploring the comparison of UML models and ArchiMate, as well as applying Agentic AI pipelines to generate complex ArchiMate enterprise architectures, mirroring advanced human reasoning to improve validation and verification strategies. Ultimately, this work lowers the barrier to understanding complex software design. It provides a scalable framework to ease the development of enterprise systems while creating accessible instructional pathways to teach fundamental programming logic and software design concepts to early educators and young learners.

More about all colloquia here 👉 https://dai.fmph.uniba.sk/w/Doctoral_Colloquia/en

📣 pozrite si prednášku našej kolegyne Kristíny Malinovskej na tému AI a kognitívna robotika 🤖 z konferencie projektu InnovAIte Slovakia

AI a kognitívna robotika | Kristina Malinovská (FMFI UK) Ako môžu roboty lepšie rozumieť ľuďom, učiť sa zo svojho prostredia...

📣 our colleagues from Innovaite Slovakia project presented this year at VISAPP 2026 conference on their latest research in pattern recognition, image and video analysis, 3D reconstruction, and deep learning 👇

FMFI UK Team Showcases InnovAIte Research at VISAPP 2026 From February 26-28, 2026, the 21st edition of VISAPP (International Conference on Computer Vision Theory and Applications) brought together global experts to discuss the latest theoretical and applied advances in the field.

The 2nd InnovAIte Slovakia Conference – AI Synergy | Žilina, April 29, 2026 Researchers, industry experts, and innovators came together in Žilina for the second InnovAIte Slovakia Conference. Under the theme AI Synergy: Connecting Re...

✨Another week we are starting with the doctoral colloquium✨

Thakar Mayur Pramod with the topic Literature Review and Research Directions in Trustworthy AI for Medical Imaging

Abstract: Artificial intelligence has become increasingly important in medical imaging applications such as image segmentation, classification, localization, and computer-assisted diagnosis. Although recent deep learning approaches have achieved high performance across multiple medical imaging tasks, several challenges still limit their reliability and broader clinical adoption. Current research highlights important issues related to explainability, robustness, annotation quality, model generalization, and trustworthiness in AI-assisted medical systems. This presentation summarizes the initial literature review conducted in the area of trustworthy AI for medical imaging. The review focused on understanding common deep learning workflows used in medical image analysis, including preprocessing, region of interest localization, segmentation, classification, and evaluation methodologies. Particular attention was given to explainable AI approaches, segmentation architectures such as U-Net and nnU-Net, transformer-based models, and hybrid deep learning frameworks applied in CT and CBCT imaging studies. The reviewed literature indicates that while many existing models report high quantitative performance using metrics such as Dice score, IoU, and classification accuracy, several open research problems remain insufficiently addressed. These include limited dataset availability, annotation inconsistency, lack of robustness across different imaging conditions, and limited validation of explainability methods for real clinical use. Based on these observations, the presentation outlines possible future research directions related to trustworthy and explainable AI systems in medical imaging, with a particular interest in model reliability, preprocessing workflows, segmentation robustness, and interpretable deep learning methods for clinical applications.

More about all colloquia here 👉https://dai.fmph.uniba.sk/w/Doctoral_Colloquia/en

✨Another week we are starting with the doctoral colloquium✨

Aisha Suleiman Umar with the topic Formal Verification of the Consistency of Explainable AI in Cyber-Physical Systems

Abstract: As Deep Learning (DL) models are increasingly integrated into the control loops of safety-critical Cyber-Physical Systems (CPS), such as smart grids, autonomous vehicles and industrial robotics, the black-box nature of these models poses a significant barrier to trust and regulatory compliance. While Explainable AI (XAI) aims to provide transparency, current post-hoc explanation methods are often heuristic, computationally unstable, and lack formal guarantees. Specifically, the consistency problem where small, non-semantic perturbations in input lead to radically different explanations, remains a critical vulnerability that can mislead human operators or mask adversarial attacks. A crucial but underexplored aspect is the formal verification of the consistency of these explanations: ensuring that explanations are robust, reliable, and faithfully reflect model behavior across different scenarios. This research review synthesizes recent literature on the intersection of formal verification, explanation consistency, and XAI in CPS, highlighting current approaches, evaluation strategies, limitations, and open research directions.

More about all colloquia here 👉https://dai.fmph.uniba.sk/w/Doctoral_Colloquia/en

✨Another week we are starting with the doctoral colloquium✨

Milica Kiš with the topic Computational modeling of the effects of psychedelics on cognitive functions

Abstract: Psychedelics are a class of psychoactive substances known to induce alterations in perception, mood, and cognitive processes. Recent studies have reported promising therapeutic outcomes for several psychiatric conditions, including anxiety, treatment-resistant depression (TRD), and posttraumatic stress disorder (PTSD). However, the impact of psychedelics on cognitive functions remains insufficiently explored. Computational modeling provides a valuable approach for elucidating the mechanisms underlying psychedelic effects. This project aims to integrate methodologies from computer science (spiking neural networks, entropy measures), physics (criticality theory), mathematics (graph theory), statistics (Bayesian framework), and neuroscience to model, analyze, and investigate the complex dynamics of brain activity under the influence of psychedelics. The research project will proceed as follows: (1) select a specific cognitive function and replicate its typical operation in a computational model, such as human vision; (2) manipulate the model, informed by existing literature, to reproduce psychedelic effects, such as visual hallucinations; (3) hypothesize emergent properties within the system, for example, an increase in entropy; (4) quantify, analyze, and validate findings using appropriate measures and metrics grounded in established literature, such as entropy measures; (5) compare and refine the models based on the results. If possible, the final step will involve validating the model against externally obtained experimental data or open-source datasets. We are currently focusing on reproducing psychedelic effects in the domain of human vision, specifically visual hallucinations, by manipulating precision in a hierarchical predictive coding model.

More about all colloquia here 👉https://dai.fmph.uniba.sk/w/Doctoral_Colloquia/en