AI Master's Thesis Awards 2022 and 2023
NTNU AI Master's Thesis Awards 2022 and 2023
Every year several students at NTNU write their master’s thesis in which AI plays a substantial part. We at the Norwegian Open AI Lab seek to encourage and award the hard work from our top tier students. This year we awarded the Annual AI Master’s Thesis Awards for both 2022 and 2023, as the event was not held in 2022. In the end, the two winners were Ellen Zhang Chang for 2022 with her thesis "Surrounding Dialogue Generation using Deep Learning with Adapters", and Michael Staff Larsen for 2023 for his thesis "Segmentation of Coronary Arteries using Transformers".
-
Photo: Kai T. Dragland/NTNU -
Photo: Kai T. Dragland/NTNU -
Photo: Kai T. Dragland/NTNU -
Photo: Kai T. Dragland/NTNU
This evaluation process
In order to identify the best AI master’s theses for these two years, we implored the wide AI research community at NTNU to nominate their students. This year, the committee only considered theses that had recieved the grade A, so the overall quality of the nominations was very high. The committee consisted of researchers from the NAIL core team, as well representatives from the Department of Computer Science
(IDI) - both in Trondheim and in Gjøvik, and from the Department of Electronic Systems (IES), the Department of Information Security and Communication Technology (IIK), and the Department of ICT and Natural Sciences (IIR) at campus Ålesund.
The varied group of researchers was tasked with selecting the top three theses from each year, within the two categories "AI Applications" and "Theory & Methods". Within these categories the nominated theses had topics ranging all the way from large language models and gender bias, to salmon health and cannibalistic lobster extermination. We were once again happy to see several female students nominated for the award, as tech and AI in particular continues to be dominated by mostly men. In the end the committee chose these six winners from the two years:
Winners 2022
- 2nd place within 'AI Applications': "You Shall Know a Female Word by the Company It Does Not Keep: Detecting and Mitigating Gender Bias in Norwegian Language Models” by Andrine Lossius and Regine Pösche Ruud (Department of Computer Science)
- 2nd place within 'Theory & Methods': "Applied Behavioral Phenotyping for the Selective Breeding of a Non-Cannibalistic European Lobster" by Eivind Hovdegård Furdal and Ole Johan Miøen (Department of Computer Science)
- First place Best overall AI Master’s Thesis Award: "Surrounding Dialogue Generation using Deep Learning with Adapters" by Ellen Zhang Chang (Department of Computer Science)
Winners 2023
- 2nd place within 'AI Applications': "Developing a deep learning pipeline for automated salmon welfare analysis by respiration frequency" by Espen Høgstedt (Department of Engineering Cybernetics)
- 2nd place within 'Theory & Methods': "Safe Reinforcement Learning in Marine Navigation and Control" by Aksel Vaaler and Svein Jostein Husa (Department of Engineering Cybernetics)
- First place Best overall AI Master’s Thesis Award: "Segmentation of Coronary Arteries using Transformers" by Michael Staff Larsen (Department of Computer Science)
Combined award ceremony and workshop
This year we hosted the award ceremony in Gruva at campus Gløshaugen. We were led through the day by the excellent host Jørgen Jore, from BRAIN NTNU. Not only did we award the winners for their work, but we also combined it with a workshop for our partners. The winning students all presented their work, and some of them partook in a panel, discussing what made them choose their particular theses. This was very insightful for the audience - partners and supervisors alike - as it can be hard to know exactly what drives a student to want to work with industry problems. Moreover, with the participation on stage from NAIL partners DNB and ANEO, we also gained very valuable insight from the other side of student-partner-collaborations.
Since submitting their AI master's theses, the winning students have gone off into the 'real world', working within both consulting and big tech departments, both in Oslo and in Trondheim. Furthermore, three of the winners are now employed as PhD-students here at NTNU, showing that their drive for AI research has not been dulled after finishing their master's projects.
Towards the end of the day, the prizes were handed out by the chair of the committee, Torbjørn Svendsen, who also read the committee's reasonings for all winners. The two top winners from each year, Ellen Zhang Chang and Michael Staff Larsen, were praised for their innovative use of AI and for their big possible impact in their respective fields.
Congratulations to all the winners and a big thank you to all the partners, students, and supervisors who partcipated to the program!
Nominations 2022
- Eivind Hovdegård Furdal and Ole Johan Miøen - Applied Behavioral Phenotyping for the Selective Breeding of a Non-Cannibalistic European
Lobster - Elias Mohammed Elfarri - Digital Twin of a Building Powered by Artificial Intelligence and Demonstrated in Virtual Reality
- Ellen Zhang Chang - Surrounding Dialogue Generation using Deep Learning with Adapters
- Magnus Eide Schjølberg and Nicklas Imanuel Paus Bekkevold - Simulation and Optimization of Emergency Medical Services in Oslo and Akershus
- Andrine Lossius and Regine Pösche Ruud - You Shall Know a Female Word by the Company It Does Not Keep:
Detecting and Mitigating Gender Bias in Norwegian Language Models - Lars-Magnus Underhaug - From Traits to Threats: Identification of Personality Traits for Individuals at Risk of Radicalisation on Social Media
Nominations 2023
- Aksel Vaaler and Svein Jostein Husa - Safe Reinforcement Learning in Marine Navigation and Control
- Eivind Dogger - Multi-agent reinforcement learning with graph neural networks for optimizing an industrial sorting system
- Michael Staff Larsen - Segmentation of Coronary Arteries using Transformers
- Marcus Tiedemann Økland Henriksbø - Prompting generative models for named entity recognition using language and visuals
- Erling Olweus - Deep Neural Network Architectures for Detection and Segmentation of Solar Farms in Satellite Imagery
- Thomas Løkkeborg - Deep Reinforcement Learning for International Diplomacy
- Espen Høgstedt - Developing a deep learning pipeline for automated salmon welfare analysis by respiration frequency
- Christopher Strøm - Towards robust and flexible point-object multi-target tracking using transformer neural networks
- Magnus Kristiansen and Magnus Morud Vågen - Aligning Diffusion-Based Text-to-Image Models using Reinforcement Learning from Human Feedback
- Erlend Stegavik Rygg, Hjalmar Jacob Vinje, and Cassandra Wu - Enhanced Option Pricing Using Deep Learning: A Time-Series Approach with a Combined LSTM-MLP Model
- Andreas Rønnestad - Evaluation of Safety-Oriented Metrics For Object Detectors
