Professor Georgios Leontidis and Athinoulla Konstantinou presenting at ICCV.
Image:
Private

Professor Georgios Leontidis and Athinoulla Konstantinou presenting at ICCV.

My Research Stay at Visual Intelligence: Athinoulla Konstantinou

Athinoulla Konstantinou is a PhD student at the University of Aberdeen. She visited Visual Intelligence's hub in Tromsø to work with equivariant modelling.

My Research Stay at Visual Intelligence: Athinoulla Konstantinou

Athinoulla Konstantinou is a PhD student at the University of Aberdeen and a member of the first cohort of the UKRI AI Centre for Doctoral Training in Sustainable Understandable agri-food Systems Transformed by Artificial INtelligence (SUSTAIN). She visited UiT The Arctic University of Norway in Tromsø to work with Professor Georgios Leontidis and researchers from UiT’s Machine Learning group on equivariant modelling.

Published by Petter Bjørklund, Communications Officer at SFI Visual Intelligence

Hi! My name is Athinoulla Konstantinou, and I am a PhD student at the University of Aberdeen, where I am part of the SUSTAIN Centre for Doctoral Training. My PhD research focuses on developing machine learning approaches for crop breeding in precision-controlled vertical farming, with the broader aim of supporting more sustainable and resilient food production.

My research draws particularly on deep learning and computer vision. One of the challenges when working with plant data is that the same biological structure can be observed from different viewpoints, orientations and configurations. Ideally, a machine learning model should be able to account for these transformations in a principled way rather than having to learn every possible variation independently.

This was one of the main motivations for my research visit to Tromsø. I visited Professor Georgios Leontidis at UiT and worked with researchers in the Machine Learning group to explore equivariant machine learning and how these ideas could be incorporated into my PhD research.

During the visit, I had the opportunity to investigate models that explicitly encode symmetries and geometric structure into the learning process. We explored how equivariant representations can allow neural networks to respond predictably to transformations of their input, and how this could be useful when analysing visual and spatial characteristics of plants. This gave me a different perspective on model design: rather than relying solely on increasingly large models and extensive data augmentation, we can sometimes incorporate useful prior knowledge directly into the architecture.

An important part of the visit was being able to discuss these ideas with researchers working on different aspects of machine learning. These conversations helped me think more carefully about which symmetries are meaningful for my own problem, how they should be represented, and when equivariance is genuinely beneficial rather than simply an additional architectural constraint.

The research stay has therefore opened up a new direction within my PhD. I am now looking at how equivariant modelling can complement the computer vision and representation-learning approaches I am developing for crop breeding and controlled-environment agriculture. More broadly, the experience reinforced the value of bringing methodological developments in machine learning together with challenging real-world problems such as sustainable food production.

Spending time in Tromsø was also a great opportunity to experience a different research environment and become part of the UiT machine learning community for a while. I returned to Aberdeen with new ideas, new technical tools, and several research questions that I am looking forward to developing further during the remainder of my PhD.

Latest news

Visual Intelligence well represented at Arendalsuka 2026

August 13, 2026

Visual Intelligence gathered experts from Norwegian politics, research council, industry, the public sector, and academia to discuss the value of Norwegian AI competence and Norwegian-developed models.

Two Visual Intelligence-authored papers accepted for MICCAI 2026

June 17, 2026

Visual Intelligence will be well represented at MICCAI 2026, one of the leading AI conferences on medical imaging and computer assisted intervention, with two accepted research papers.

Visual Intelligence represented at the Datacloud Global Congress' Talent in Tech programme

June 8, 2026

PhD Candidates Solveig Thrun and Christian Salomonsen attended the Datacloud Global Congress' Talent in Tech programme, which invited emerging talent to a unique programme designed to inspire, educate, and connect young professionals with senior leaders from global tech giants.

Kyst og Fjord: Råfisklaget tester AI-teknologi for å knipe syndere

June 2, 2026

En ny AI-løsning utviklet ved UiT kan gjøre det enklere å avdekke feilrapportering og ulovlig fiske. Nå skal Norges Råfisklag teste teknologien i praktisk bruk (News article on kystogfjord.no)

Geo365.no: Norges nye seismikkekspert

May 28, 2026

En ny, åpent tilgjengelig KI-modell kan endre hvordan geologer tolker seismikk. Den norske grunnmodellen lover raskere analyser, lavere terskel for innovasjon og nye måter å forstå undergrunnen på.

NRK Radio interview on AI for sustainable fisheries

May 27, 2026

Centre Director Robert Jenssen was interviewed by Norsk rikskringkasting (NRK) about AI in fisheries and this year's UArctic (University of the Arctic) Congress on the Faroe Islands.

Centre Director Robert Jenssen in Japan to discuss Norwegian-Japanese AI collaborations‍

May 26, 2026

Professor and AI expert Robert Jenssen is attending the Japan-Norway AI Innovation Forum and Japan-Norway Research Symposium, two high-level meetings with Norwegian and Japanese government leaders, business actors and researchers.

Call for Papers and Abstracts: NLDL 2027

April 22, 2026

The Call for Papers and Abstracts for the Northern Lights Deep Learning (NLDL) Conference 2027 is officially announced – with submission deadlines on August 7th and Mid-September 2026 respectively.

Trends in Visual Intelligence 2026

April 17, 2026

The field of Visual Intelligence is continuously transforming. Chief Research Scientist Arnt-Børre Salberg dives deeper into the current trends in the field of visual intelligence as of early 2026.

Centre-developed seismic foundation model is now open source!

April 6, 2026

The NCS model, a seismic foundation model trained on data from the Norwegian data repository for subsurface data, is now available as an open-source model, allowing anyone to download, utilize, and further develop the model.