Marina Ivasic-Kos | computer vision | Research and Innovation Citation Excellence Award

Prof Dr. Marina Ivasic-Kos, computer vision, Research and Innovation Citation Excellence Award

Professor  at the University of Rijeka, Croatia

Dr. Marina Ivašić-Kos is a renowned computer scientist and professor at the University of Rijeka, Croatia. With expertise in computer vision and deep learning, she specializes in applications such as person detection in search and rescue, thermal object detection, and activity-based player recognition. Dr. Ivašić-Kos has led various research projects, published extensively in reputable journals, and holds key positions in academic committees. Her impactful contributions include innovative models for image annotation, showcasing a multi-faceted approach integrating machine learning and fuzzy-knowledge representation schemes. As a distinguished figure in the field, she continues to advance the frontiers of intelligent systems and pattern analysis.

Professional Profiles:

Scopus Profile

Orcid Profile

GoogleScholar profile

Researchgate profile

LinkedIn profile

📚 Education:

Ph.D. degree in Computer Science (2002-2012): Faculty of Electrical Engineering and Computing, University of Zagreb, Zagreb M.Sc. degree in Information Science (1997-2001): Faculty of Philosophy, University of Zagreb, Department of Information Science, Zagreb.

Work Experience:

Present Employer: Faculty of Informatics and Digital Technologies, University of Rijeka, and Centre for Artificial Intelligence and Cybersecurity, University of Rijeka Position: Professor Head of Laboratory for Pattern Recognition and Soft Computing Head of Laboratory for Computer Vision, Virtual and Augmented Reality at the Centre for Artificial Intelligence, University of Rijeka Responsibilities: Teaching on postgraduate, graduate, and Doctoral Studies in various subjects Project leader and researcher in computer vision scientific projects President of the Publishing Committee at the Department of Informatics Member of various university councils and committees.

Training:

Year 22.-26.7.2019: Place of training: Warsaw, Poland Training on Deep Learning by IRDTA Brussels, London Year 4.-6. 9.2017: Place of training: Ionic Center, Athens, Greece Training on Integration in Cognitive Robotics by IV&L Net Training School Year 2016: Place of training: Las Palmas De Gran Canaria, Spain Training on Biometry and de-identification by COST IC1206 2nd Training School Year 21.-24. 3. 2016: Place of training: University of Malta, Msida, Malta Training on Computer Vision and Deep Learning by The 2nd iV&L Net Summer School.

Technical skills and competencie

Computer vision, machine and deep learning, recognition and detection of objects, recognition of actions (Matlab, Keras, TesorFlow) – fuzzy systems, knowledge representation – programming, work on developing and implementing applications (Java, C++, Oracle) – design of IS projects (Agile, Scrum, UML).

Research Focus

Dr. Marina Ivašić-Kos specializes in the research areas of computer vision, image processing, and pattern analysis. Her work encompasses the development and application of deep learning techniques, particularly Convolutional Neural Networks (CNN), for tasks such as automatic person detection in search and rescue operations. Additionally, she has expertise in thermal object detection under challenging weather conditions using YOLO, and active player detection in handball scenes based on activity measures. Dr. Ivašić-Kos has contributed significantly to the field of image annotation, proposing innovative models such as a two-tier image annotation model with a multi-label classifier and fuzzy-knowledge representation scheme. Her research also extends to knowledge-based multi-layered image annotation systems, Bayesian classifiers, and fuzzy Petri net knowledge representation schemes.

Publications (TOP NOTES):

 

Deep Learning Approach for Objects Detection in Underwater Pipeline Images, Cited by 12, Publication date: 2022/12/31.

 

Review and analysis of synthetic dataset generation methods and techniques for application in computer vision, Cited by 12, Publication date: 2023/1/30.

 

Thermal object detection in difficult weather conditions using YOLO, Cited by 197, Publication date: 2020/7/6.

 

Automatic person detection in search and rescue operations using deep CNN detectors, Cited by 93, Publication date: 2021/3/4.

 

An overview of Human Action Recognition in sports based on Computer Vision, Cited by 41, Publication date: 2022/6/1.

 

A new approach to dominant motion pattern recognition at the macroscopic crowd level, Cited by 31, Publication date: 2022/11/1.

 

Person Detection in thermal videos using YOLO, Cited by 25, Publication date: 2020

 

Tracking Handball Players with the DeepSORT Algorithm.,Cited by 24, Publication date: 2020/2.

 

Active player detection in handball scenes based on activity measures,. Cited by 22, Publication date: 2020/3/8

 

.