Real life challenges, impact on society and solutions that are applicable for the Norwegian industry were among the criteria for winning
“All the nominees were exceptional in contributing to cutting-edge research on artificial intelligence and machine learning, but where the winners excelled was on the criteria of presenting work that is highly relevant for the Norwegian industry and make an impact on society. The winners each solve very different real life challenges and is a very valuable contribution to AI research,” says Kerstin Bach, Associate Professor at NTNU and head of the Evaluation Committee.
The winners were honoured at a ceremony at NTNU on Friday the 17th of November and received a monetary award sponsored by Telenor Group.
“We want to award talents who are contributing towards digitising the Norwegian industry and society, and I am very impressed with the theses. The Telenor-NTNU AI-Lab set out to be a powerhouse for artificial intelligence research and we are now seeing the fruits. I’d like to encourage those interested in discovering the endless possibilities within AI and machine learning to join the lab, where they get access to very relevant problems to solve from our industry partners and real data to experiment with,” says Ieva Martinkenaite, Head of Telenor-NTNU AI-Lab initiative and VP, Telenor Research.
The winning theses and students are:
Håkon Måløy for “A Dual-Stream Deep Learning Architecture for Action Recognition in Salmon from Underwater Video”.
The thesis looks at applying Deep Learning for automating the feeding process for salmon farms by using salmon motion behaviour in order to optimize operations and reduce potential food waste. Key deliverables from the research can be put into production fast and could have noticeable impact on the Norwegian fish farming industry.
Øyvind Kjerland for “Segmentation of Coronary Arteries from CT-scans of the heart using Deep Learning”.
The thesis proposes a Deep Learning method for fully automatic segmentation of the coronary arteries and was tested on a dataset provided by St. Olavs Hospital. The research contributes to improving efficiency and precision in diagnosing the coronary artery disease – one of the leading causes of death in Europe.
Isuf Deliu for “Extracting Cyber Threat Intelligence From Hacker Forums”.
The thesis explores machine learning methods to locate relevant threat intelligence from hacker forums, understand the methodology used by threat actors to launch their campaigns, and proactively adapt security controls to detect and prevent such activity. It’s an important step towards extracting actionable cyber-threat intelligence from a variety of big data sources that could also lead to new methods for enhancing cyber-security.