Meet Sigurd and Adrian, our newest Doctoral Research Fellows

We happily welcome Sigurd Almli Hanssen and Adrian Duric as new PhD Candidates at SFI Visual Intelligence's hubs in Tromsø and Oslo respectively.

By Petter Bjørklund, Communications Officer at SFI Visual Intelligence

Both Hanssen and Duric recently started their doctoral research positions at SFI Visual Intelligence. Hanssen's project is tied to the centre's "Earth observation" innovation area, while Duric's project is connected to the "Marine science" innovation area.

Machine learning for sustainable oil spill monitoring

Hanssen is originally from Evenskjer, a town near Harstad, Norway. He graduated with a master's degree in Applied Physics and Mathematics from UiT The Arctic University of Norway in Spring 2025. His thesis focused on how to apply diffusion models for oil spill detection, in collaboration with industry partner Kongsberg Satellite Services (KSAT). Building upon his master's thesis, Hanssen's PhD project will focus on using satellite imagery and machine learning for sustainable oil spill monitoring.

Satellite images used for environmental monitoring are often huge in size. Current methods, which slice the images into smaller, more manageable patches, tend to remove the context in which the patches exist in. A core idea of the project is to research methods which allow us to reincorporate the large-scale context of these images.

"This could hopefully reduce some sources of false alarms – that is, when a detection system reports an oil spill even though no actual spill is present. This is critical as false alarms, especially far from shore, can be quite expensive and resource-intensive to investigate," Hanssen explains.

Hanssen says it's inspiring to work with what he deems as an interesting and accomplished group of researchers, both at KSAT and UiT.

"Having worked for years with manual image analysis, I really appreciate the importance of and effect I can have through my research, which I find really exciting," says Hanssen.

Automatically estimating fish populations with echosounder data

Duric is from Oslo, Norway and has a master's degree in Robotics and Intelligent Systems from the University of Oslo, specializing in AI and machine learning. His thesis, written in collaboration with the Simula Research Laboratory, focused on using explanations generated by explainable AI techniques as additional data features to enhance the deep learning models' prediction performance.

His PhD project focuses on developing and utilizing the latest deep learning advances to improve performance on various image recognition tasks within marine imaging. Echosounder data captured from Norwegian waters, which is used to estimate different fish populations at specific locations, will have an important focus.

"From a more technical aspect, I will work on several different research topics that contribute in different ways towards the project: foundation models, distrubuted ML and compression, representation learning and explainable AI being some of them," Duric explains.

Duric says he's eager to produce high-impact contributions that can help society, both on a scientific and applied level.

"This is something that drives me and continues to motivate me every day, and a main reason I decided to pursue a PhD in the first place. I really look forward to collaborating with my colleagues and other fellow researchers," Duric says.