A Norwegian centre for research-based innovation

Through long-term research in close collaboration between industry, public institutions and prominent research partners, we enable novel innovations, technology transfer, internationalization and researcher training.

Learn More

We research the next generation of deep learning methodology for visual data and produce solutions for our consortium partners across innovation areas in medicine and health, marine science, energy, and earth observation.

Did you miss this?

VI Seminar #77: Addressing Label Shift in Distributed Learning via Entropy Regularization

June 5, 2025

Latest news

Three Visual Intelligence-authored papers accepted for leading AI conference on medical imaging

June 24, 2025

Visual Intelligence will be well represented at MICCAI 2025—one of the leading AI conferences on medical imaging and computer assisted intervention—with three recently accepted research papers.

2025 Norwegian AI Society Symposium: An insightful and collaborative event

June 23, 2025

More than 50 attendees from the Norwegian AI research community gathered in Tromsø, Norway for two days of insightful presentations, interactive technical sessions, and scientific and social interactions.

When:
August 12, 2025
,
14:00
August 12, 2025
,
14:45
@
Kunstig intelligens-teltet (Kirkebakken 19)

Vi spør; Hva bør vi gjøre for å styrke forskningsdrevet innovasjon innen KI i Norge? Hvilke muligheter ligger i denne styrkingen? Hva er utfordringene, og hva må vi gjøres for å løse de? KI-senteret SFI Visual Intelligence ved UiT Norges arktiske universitet inviterer til en samtale mellom akademia, næringsliv og politikere for å diskutere disse spørsmålene.

twitterfacebookYoutubeGithub logoSign up for the Visual Intelligence newsletter.

Visual Data Diagnosis and Debiasing with Concept Graphs

March 6, 2025

We propose ConBias, a bias diagnosis and debiasing pipeline for visual datasets.

Read More

Modular Superpixel Tokenization in Vision Transformers

March 6, 2025

ViTs partition images into square patches to extract tokenized features. But is this necessarily an optimal way of partitioning images?

Recent publications

Leveraging Foundation Model Adapters to Enable Robust and Semantic Underwater Exploration

By authors:

Changkyu Choi, Arangan Subramaniam, Nils Olav Handegard, Ali Ramezani-Kebrya and Robert Jenssen

Published in:

Proceedings of the Symposium of the Norwegian AI Society 2025, CEUR Workshop Proceedings ( ISSN 1613-0073)

on

June 17, 2025

Pixel-Level Predictions with Embedded Lookup Tables

By authors:

Marius Aasan, Adín Ramírez Rivera

Published in:

Proceedings of the Symposium of the Norwegian AI Society 2025, CEUR Workshop Proceedings ( ISSN 1613-0073)

on

June 17, 2025

Assessing the Efficacy of Multi-task Learning in Mammographic Density Classification: A Study on Class Imbalance and Model Performance

By authors:

Suaiba A. Salahuddin, Elisabeth Wetzer, Kristoffer Wickstrøm, Solveig Thrun, Michael Kampffmeyer and Robert Jenssen

Published in:

Lecture Notes in Computer Science (LNCS) 2025 ;Volum 15726.

on

June 16, 2025

ProxyDR: Deep Hyperspherical Metric Learning with Distance Ratio-Based Formulation

By authors:

Hyeongji Kim, Changkyu Choi, Michael Christian Kampffmeyer, Terje Berge, Pekka Parviainen, Ketil Malde

Published in:

Lecture Notes in Computer Science (LNCS) 2025

on

May 12, 2025

Addressing Label Shift in Distributed Learning via Entropy Regularization​

By authors:

Zhiyuan Wu, Changkyu Choi, Volkan Cevher, Ali Ramezani-Kebrya

Published in:

International Conference on Learning Representations 2025

on

April 29, 2025

Fast TILs—A pipeline for efficient TILs estimation in non-small cell lung cancer

By authors:

Nikita Shvetsov, Anders Sildnes, Masoud Tafavvoghi, Lill-Tove Rasmussen Busund, Stig Manfred Dalen, Kajsa Møllersen, Lars Ailo Bongo, Thomas K. Kilvaer,

Published in:

Journal of Pathology Informatics

on

March 12, 2025

Research challenges

Visual Intelligence address the research challenges of deep learning and computer vision that limit our user partners in utilizing their complex visual data in their applications.

Read more
Go to our research challenges
Go to our research challenges

Innovation areas

We contribute to reliable use of AI to detect heart disease, monitor the environment and potential natural disasters as well as detecting natural resources. Read more about our work in the different innovation areas.

Read more

Our partners

Visual Intelligence is a consortium headed by UiT The Arctic University of Norway with research partners at the University of Oslo and the Norwegian Computing Center. Together with our consortium of high-profile user partners, we create cutting-edge solutions that will be implemented in the applications of the user partners.

UiT The Arctic University of Norway logoUiO: University of Oslo logoUniversity hospital of north norway logoHelse nord ikt logoInstitute of marine research logoKongsberg satellite services logoGE Healthcare logoEquinor logoCancer Registry of Norwat logo