AI Master's Thesis Awards 2024
AI Master's Thesis Awards 2024
Each year, numerous students at NTNU delve into AI for their master’s theses. At the Norwegian Open AI Lab, we strive to recognize and reward the dedication of our exceptional students. This year, we presented the Annual AI Master’s Thesis Awards for 2024, and ultimately, the awards went to Eirik Runde Barlaug and Jørgen Lind Fløystad with their thesis «Reactive Quadrotor Guidance System Using Deep Reinforcement Learning, Autoencoders and Nonlinear Control», Henrik Friis and Håkon Nese for their work on «CT Neural Computed Tomography for Sparse-View CT and 4DCT», as well as «Stochastic Approximation with Contrastive Learning» by Erland Brandser Olsson.
Narrowing down the strong 14 candidates
To find the top AI master’s theses for this year, we called upon the extensive AI research community at NTNU to nominate their students. Once again, the committee only considered theses that had received an A grade, ensuring a high standard of quality among the nominations. The committee was composed of researchers from the NAIL core team, as well as representatives from the Department of Computer Science (IDI) in both Trondheim and Gjøvik, the Department of Electronic Systems (IES), and the Department of ICT and Natural Sciences (IIR) at the Ålesund campus.
This diverse group of researchers was tasked with selecting the top three theses from the 14 nominations, within the categories of "AI Applications" and "Theory & Methods." The nominated theses covered a wide range of topics, from optimizing the norwegian emergency medical service and detecting bias in LLMs, to computer vision both on the football field and for autonomous driving. Ultimately, the committee selected these three winners for the year 2024:
Winners 2024
- Best AI Master’s Thesis, 2nd place – AI Applications: «Reactive Quadrotor Guidance System Using Deep Reinforcement Learning, Autoencoders and Nonlinear Control» by Eirik Runde Barlaug and Jørgen Lind Fløystad. Supervised by Adil Rasheed (ITK). Co-supervisor has been Thomas Nakken Larsen (ITK)
- Best AI Master’s Thesis, 2nd place – Theory & Methods: «CT Neural Computed Tomography for Sparse-View CT and 4DCT» by Henrik Friis and Håkon Nese. Supervised by Dag W. Breiby (Physics Dpt.) og Ole Jakob Mengshoel (IDI). Co-supervisor has been Anders Kristoffersen, leader of Equinor ASA’s CT laboratory at Rotvoll.
- Best AI Master’s Thesis, 1st place – Best overall thesis: «Stochastic Approximation with Contrastive Learning» by Erland Brandser Olsson. Supervised by Zhirong Yang (IDI).
Try and try again – and you will succeed
This year, we once again hosted the award ceremony in Gruva at the Gløshaugen campus in Trondheim. The event was expertly guided by our host, Fredrik William Husemoen-Zhang from the student organisation BRAIN NTNU. In addition to celebrating the winners' achievements, we integrated the ceremony with a workshop for our partners.
The winning students presented their work and participated in a panel discussion, sharing insights into why they chose their specific theses. This was particularly enlightening for the audience, including both partners and supervisors, as it shed light on what motivates students to tackle industry-related problems. Additionally, with contributions from NAIL partner NINA – Norwegian Institute for Nature Research, we gained valuable perspectives on student-partner collaborations.
One of the crucial insights from the event was the significance of having dedicated and engaged supervisors, both from NTNU and the partner organizations. Additionally, it highlighted the necessity for partners to adopt a long-term perspective on their projects with the university. As one panel member aptly put it, industry partners should persistently engage and aim for a five-year horizon, focusing on building strong relationships with the AI community at NTNU, rather than seeking to solve short-term issues within a single semest
Since completing their AI master's theses, the winning students have ventured into the professional world, working in consulting and big tech departments in both Oslo and Trondheim. Notably, one of the winners has continued their academic journey as a PhD student here at NTNU, demonstrating their sustained passion for AI research.
Towards the end of the day, the prizes were awarded by the chair of the evaluation committee, Professor Helge Langseth, who also shared the committee's commendations for each winner. The winners were lauded for their innovative use of AI and the significant potential impact of their work in their respective fields.
Congratulations to all the winners, and a heartfelt thank you to all the partners, students, and supervisors who contributed to the program!
Nominations 2024
- Erlend Lokna og Johan Vik Mathisen – Blending Low and High-Level Semantics of Time Series for Better Masked Time Series Generation
- Erland Brandser Olsson – Stochastic Approximation with Contrastive Learning
- Torjus Åmellem Kallekleiv og Sindre Eiklid – Optimizing Emergency Medical Service Compliance in Norway: A Simulation-Based Study
- Henrik Friis og Håkon Nese – NeCT: Neural Computed Tomography for Sparse-View CT and 4DCT
- Selma Kristine Bergstrand – Detecting and Mitigating LGBTQIA+ Bias in Norwegian Large Language Models
- Espen Boman Fosseide og Birk Gustav Samson Stoveland – Novel Applications of Sampling Methods and Backbone Architectures for Histopathological DINOv2
- Simen Dymbe – Multi-channel speech recognition
- Eirik Runde Barlaug og Jørgen Lind Fløystad – Reactive Quadrotor Guidance System Using Deep Reinforcement Learning, Autoencoders and Nonlinear Control
- Shayan Tafrishi – The Future of Correctional Workforce Management: AI and Societal Cybernetics to the Rescue
- Jon Torgeir Grini – Bayesian Adaptive Segmentation of Water Surfaces with Reflection Correction
- Mathias Pettesen – PLANNER - Planning Layouts Autonomously using Novel, Nuanced Evolutionary Reasoning
- Matias Gran-Henriksen og Hans Andreas Lindgård – Assessing the Accuracy of Computer Vision in Football Player Tracking
- Jarl Sondre Bringslid Sæther og Bendik Gjermundrød Holter – Multi-Level Magnification with Transformer-Based Architectures for Enhanced Cell Detection and Classification in Computational Pathology
- Shaira Tabassum – From Roadside Rendering to Reality: Advancing NeRF Pipeline for Photorealistic 3D Views in Autonomous Driving
