Job description

Jeg er direktør for Visual Intelligence. Dette er et Senter for Forskningsdrevet Innovasjon (SFI) innen kunstig intelligens finansiert for 8 år av Norges forskningsråd og et konsortium av private og offentlige partnere (prosjektnr. 309439). Vi er i den internasjonale frontlinjen innen forskning på dyp læring for bildeanalyse spesielt og multimodal læring generelt. 

Min hovedstilling er som Professor i Forskningsgruppa for maskinlæring ved UiT.

Jeg er også ansatt som Adjunct Professor ved: Pioneer Centre for AI, University of Copenhagen & Norwegian Computing Center.

For en kort CV, se "Vedlegg".

Fremme vår felles framtid gjennom forskning

Min motivasjon er å bidra til løsninger for de store samfunnsutfordringene i vår tid innen helse, marin kartegging, bedre utnyttelse av energiressurser og presis observasjon av jorda. Jeg har omfattende samarbeid med industri og offentlige aktører. Min metodologiske forskning har fokusert på emner som nevrale nettverk, informasjonsteoretisk læring, "kjernemetoder", ikke-styrt læring, selv-læring og forklarbar AI (XAI). Min forskning publiseres jevnlig innen de mest sentrale konferanser og journaler innen feltet (ICLR, ICML, NeurIPS, etc). Jeg har vært heldig å jobbe med mange dyktige kolleger, og sammen har vår forskning blitt anerkjent i feltet:   

Priser for forskning (og undervisning)

• Beste artikkel, Pattern Recognition Letters (2024)
• Dissertation Award, Norwegian Artificial Intelligence Society (Trosten, biveileder, 2023)
• Beste artikkel, Colour and Visual Computing Symposium (2022)
• Beste artikkel Int’l Medical Informatics Association (2018)
• Undervisningsprisen, Fakultet for naturvitenskap og teknologi, UiT (2018)
• Beste studentartikkel, Scandinavian Conference on Image Analysis (veileder) (2017)
• Vinner av IEEE GRS Society Letters Prize Paper Award (2013)
• Framhevet artikkel, IEEE Transactions on Pattern Analysis and Machine Intelligence (2010)
• Pris for Yngre Forsker Universitetet i Tromsø (deles ut annethvert år) (2007)
• Vinner av ICASSP Outstanding Student Paper Award (2005)
• Beste artikkel, Pattern Recognition Journal, Honourable Mention (2003)

Internasjonalt lederskap (nåværende)

• Vitenskapelig rådgivningsstyre (Scientific Advisory Board - SAB) for Max Planck Institute for Intelligent Systems https://is.mpg.de
• I SAB for et av Frankrikes nye "AI Excellence Clusters" SequoIA, finansiert av France 2030 og under administrasjon fra French National Agency for Research (ANR)
• I SAB for DIREC - Digital Research Centre Denmark https://direc.dk 

Utvalgte artikler (Engelsk):

Google Scholar

Aggregation of Dependent Expert Distributions in Multimodal Variational Autoencoders. ICML 2025. https://openreview.net/forum?id=jYmGi1175R

The Conditional Cauchy-Schwarz Divergence with Applications to Time-Series Data and Sequential Decision Making. IEEE TPAMI 2025.

REPEAT: Improving Uncertainty Estimation in Representation Learning Explainability. AAAI 2025. https://arxiv.org/abs/2412.08513

Editor for Special Issue on Information Theoretic Methods for the Generalization, Robustness, and Interpretability of Machine Learning. IEEE TNNLS 2025. https://doi.org/10.1109/TNNLS.2025.3525991

Finding NEM-U: Explaining unsupervised representation learning through neural network generated explanation masks. ICML 2024. https://proceedings.mlr.press/v235/moller24a.html

MAP IT to visualize representations. ICLR 2024. https://openreview.net/pdf?id=OKf6JtXtoy

Cauchy-Schwarz divergence information bottleneck for regression. ICLR, 2024. https://openreview.net/pdf?id=7wY67ZDQTE

ADNet++: A few-shot learning framework for multi-class medical image volume segmentation with uncertainty-guided feature refinement. Medical Image Analysis, 2023. https://doi.org/10.1016/j.media.2023.102870

Hubs and Hyperspheres: Reducing Hubness and Improving Transductive Few-Shot Learning With Hyperspherical Embeddings. CVPR, 2023. https://openaccess.thecvf.com/content/CVPR2023/html/Trosten_Hubs_and_Hyperspheres_Reducing_Hubness_and_Improving_Transductive_Few-Shot_Learning_CVPR_2023_paper.html

On the Effects of Self-Supervision and Contrastive Alignment in Deep Multi-View Clustering. CVPR, 2023. https://openaccess.thecvf.com/content/CVPR2023/html/Trosten_On_the_Effects_of_Self-Supervision_and_Contrastive_Alignment_in_Deep_CVPR_2023_paper.html

RELAX: Representation Learning Explainability. International Journal of Computer Vision, 2023. https://doi.org/10.1007/s11263-023-01773-2

ProtoVAE: A Trustworthy Self-Explainable Prototypical Variational Model. NeurIPS, 2022. https://openreview.net/forum?id=L8pZq2eRWvX

Principle of Relevant Information for Graph Sparsification. UAI, 2022. https://proceedings.mlr.press/v180/yu22c.html

Anomaly Detection-inspired Few-shot Medical Image Segmentation through Self-supervision with Supervoxels. Medical Image Analysis, 2022. https://doi.org/10.1016/j.media.2022.102385

Clinically Relevant Features for Predicting the Severity of Surgical Site Infections. IEEE Journal of Biomedical and Health Informatics, 2021. https://doi.org/10.1109/JBHI.2021.3121038

Measuring Dependence with Matrix-based Entropy Functional. AAAI, 2021. https://doi.org/10.1609/aaai.v35i12.17288

Reconsidering Representation Alignment for Multi-view Clustering. CVPR, 2021. https://openaccess.thecvf.com/content/CVPR2021/papers/Trosten_Reconsidering_Representation_Alignment_for_Multi-View_Clustering_CVPR_2021_paper.pdf

Joint Optimization of an Autoencoder for Clustering and Embedding. Machine Learning, 2021. https://doi.org/10.1007/s10994-021-06015-5

Uncertainty-aware Deep Ensembles for Reliable and Explainable Predictions of Clinical Time Series. IEEE Journal of Biomedical and Health Informatics, 2020. https://doi.org/10.1109/JBHI.2020.3042637

SEN: A Novel Feature Normalization Dissimilarity Measure for Prototypical Few-Shot Learning Networks. ECCV, 2020. https://link.springer.com/chapter/10.1007/978-3-030-58592-1_8