The Arctic Ocean is turning blue. Abrupt Arctic warming and amplification is driving rapid sea ice decline and irreversible deglaciation of Greenland. The already emerging, substantial consequences for the planet and society are intensifying and yet, model-based projections lack validatory consensus. To date, we cannot anticipate how a blue Arctic will respond to and amplify an increasingly warmer ...

Uncertainty estimation is an important component in the safe deployment of deep learning (DL) models, particularly in high-stakes domain such as medical imaging. While DL models often achieve impressive predictive performance, they typically provide no indication of how reliable their outputs are. In this thesis, we investigate the use of test-time augmentation (TTA) to estimate predictive uncertainty ...

As society moves towards increasing levels of digitalisation, the risk of cyber-attacks having potentially devastating impacts is increasing. In the maritime industry, the result of digitalisation is showing itself in the ongoing introduction of maritime autonomous surface ships (MASS) which can operate without crew on board. The increased reliance on technology introduces concerns regarding more ...

Computational pathology continues to advance through deep learning. However, interactive analysis of gigapixel whole slide images remains challenging due to high computational demands and long processing times. Variability in tissue morphology and staining quality further complicates automated cell detection and quantification in clinical workflows. In this study, we present an integrated and ...

Denne masteroppgaven undersøker hvordan resiliens fremstilles som et individuelt ansvar i norske regjeringsdokumenter, og hvordan dette reflekterer og påvirker forholdet mellom individ og stat innenfor samfunnssikkerhets- og beredskapspolitikken. Studien bygger på en kvalitativ dokumentanalyse av fire sentrale offentlige dokumenter: NOU 2000: 24 Et sårbart samfunn, Meld. St. 10 (2016–2017), NOU 2023: ...

Deep crustal shear zones, fundamental to the dynamics of terrestrial plate tectonics, exhibit complex processes of initiation and evolution that are yet to be comprehensively quantified across both long and short temporal scales. Conventionally, thermo–mechanical models posit that crustal rock behaviour is dominated by monomineralic aggregates undergoing processes like intracrystalline plastic ...

Modern day systems increasingly rely on data-driven insights derived from IoT networks. However, the resource-constrained nature of IoT devices creates significant challenges when outsourcing computations to third-party services, particularly regarding data integrity. This thesis investigates the feasibility of combining a homomorphic signature protocol with a pairing delegation scheme to provide ...

Denne masteroppgaven undersøker hvordan samhandling mellom politiet, brann- og redningsvesenet og ambulansetjenesten i Tromsø kommune påvirker organisasjons-læringsprosessen på tvers. Oppgaven kartlegger hvilke læringsarenaer og samhandlingsformer som preger samarbeide, hvordan læringsprosessene fungerer, og hvilke forhold som påvirker disse. Studien bygger på et kvalitativt forskningsdesign med ...

Alongside the rapid development of deep learning, image recognition has achieved significant advances, with particularly promise within the supervised learning paradigm. While current models demonstrate strong predictive performance on benchmark datasets, deep neural networks often lack robustness, exhibiting unpredictable failures in operation. Challenges such as out-of-distribution data from unseen ...

Cationic antimicrobial peptides (cAMPs) kill bacteria in solution by membrane lysis; however, translating cAMPs into a covalently attached antibacterial coating is challenging since it remains unclear how the specifics of the conjugation impact the antifouling efficacy. Furthermore, studies have typically assessed cAMP coatings with a high and homogeneous surface coverage, although this may be ...