Gorjan Radevski
I am a Research Associate (June 2024 – present) at the VISICS lab within ESAT-PSI working together with Prof. Tinne Tuytelaars. My current area of research ordinates around (but it is not limited to) Deep Learning and its applications in Natural Language Processing and Computer Vision. Among other things, I am interested in the applications of LLMs on multimodal problems spanning images, text, videos, audio, knowledge graphs; weakly-supervised learning; (egocentric) video understanding; etc.
Additionally, I am working as a (remote) Research Consultant at NEC Labs in the Human-Centric AI group (November 2023 – present). The projects I work on involve Large Language Models (LLMs), Retrieval Augmented Generation (RAG), and enhancing the interpretability of LLMs.
During my PhD and Master’s studies I completed Machine Learning (research) internships at:
-
NEC Labs (worked with Kiril Gashteovski, Carolin Lawrence & Goran Glavas)
- Published at EMNLP (2023): Linking Surface Facts to Large-Scale Knowledge Graphs
-
IBM Research (worked with Jasmina Bogojeska)
- Published in the Journal of Infections diseases (2021): Cohort-derived machine learning models for individual prediction of chronic kidney disease in people living with HIV
-
Luminovo
- Worked on Computer Vision industry solutions
In the past, I finished a PhD (2024) and graduated Cum Laude from the Master in Artificial Intelligence (2019) at KU Leuven, and did a Bachelor in Computer Science at University of Ss. Cyril and Methodius in North Macedonia (2017).
selected publications
-
PhD ThesisBridging Modalities and Transferring Knowledge: Enhanced Multimodal Understanding and Recognition2024
-
ICCVMultimodal Distillation for Egocentric Action RecognitionIn Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) Oct 2023*Equal contribution
-
BMVCRevisiting Spatio-Temporal Layouts for Compositional Action RecognitionProceedings BMVC 2021 Oct 2021Oral Presentation