Available Master topics: Machine Learning
Memory training exercises are known to have a positive effect on improving the memorability of humans. A memory training task can be as simple as trying to remember a password of increasing complexity or a number of varying length.
This topic focuses on developing automated testing methods for cancer registration and support system at Cancer Registry of Norway.
Developing automated testing tools for telehealth services at Oslo City.
We are developing an AI child avatar for the interview training of the police officers to protect vulnerable children from abuse.
Robots are playing key roles in many social and industrial applications such as in hospitals, automation in automotive domains, etc. Digital twins, i.e, live and digital representations of a physical twin such as robots, promise to significantly improve their reliability, helping them in making optimized decisions in real-time and predict future activities such as predictive maintenance. While the concept of digital twins is currently attracting a lot of attention, there is little research done on the principles to design and implement them.
The new advancement into Virtual Reality (VR) technologies increased the availability and affordability of head-mounted displays (HMD), which in turn has yielded more accessible and versatile game content. In this project, we would like to assist the analysis of QoE for VR applications, by integrating subjective evaluation directly into the headset.
Use machine learning and video processing techniques to automatically find events in sports videos. For example, in football, some "easy" ones are goals, but others like tackles are harder.
Develop and evaluate data-driven techniques and prototypes for automatically repairing bugs in source code.
Investigate, develop and evaluate data-driven techniques and prototypes that help software engineers build software systems that are autonomously self-healing. These are systems that can understand when they are not operating correctly and, without human intervention, make the necessary adjustments to restore themselves to normal operation.
Interested in sports? Technology used by real athletes? We have for a long time researched sports technology for performing performance monitoring for better results and injury avoidance. We developed running systems and tested them on real users. Want to be part of our research group and improve current systems?
In this project, students will be assigned to develop deep-learning solutions to analyze sperm videos.
Develop data-driven techniques and prototypes for automatically assessing the security of a software system by analyzing the system's source code for potential security vulnerabilities during the development stage.
Students will get experience in applying deep-learning solutions to analyze ECG signals.
This project involves developing new methods to design, develop, and test Autonomous Rovers, similar to those used on Mars.
Over the recent years Generative Adversarial Models (GANs) have revolutionized the field of artificial intelligence (AI). Given a set of training data, the models can generate outcomes with striking similarities in properties to the training data. E.g. if the training data is a set of images, the models are able to generate new images with the same properties and look stunningly realistic. However the theoretical foundation of GANs is still immature and further understanding is needed. For example, GANs have been criticised to possibly only replicate a limited set of the properties in the training data.
Implementing software tools for designing, developing, and testing of self-driving cars.
This topic is about developing digital twins for various types of cyber-physical systems.
Develop and evaluate techniques that use deep learning-based natural language processing and reasoning over knowledge graphs for the detection of online misinformation and fake news.
In building and deploying various government subsidy and support programs, government agencies used different approaches. The innovations should be studied to understand applicability in other situations, and what conditions determine design and scope.
For industries like healthcare and finance having the capacity to create high quality synthetic data that does not have the privacy constraints of normal data is extremely valuable. However, a key issue is to evaluate to what extent the synthetic data preserve privacy. For example, if a generated data point is identical to a normal data point, this obviously violates privacy.
Explain the prediction of a deep neural network analyzing microscopic videos of human semen.
In this project, different deep generative models will be implemented to generate synthetic medical data for deepsynthbody.org. Student want to learn and implement state-of-the-art deep generative modles for a selected medical dataset.
Survey experiments in studies of political and electoral behavior using profiles of potential candidates are standard practice in political science. With this type of experiments, the researchers' goal is to test which characteristics voters value most in political candidates. One of the main challenges, however, is to generate realistic profiles to be tested. One of the important components of these profiles are the faces of candidates. A potential option is to use faces of real candidates, but this involves complex legal and ethical issues. Another option is to hire models to represent candidates, but this can be expensive and time-consuming.
In this thesis, the research question is whether regular close-range images with low-cost sensors (tablet camera) of certain field sections are suitable for growth modeling of soybean. Specifically, the effects of different land use intensities on soybean growth patterns shall be investigated using current methods of image analysis and plant phenotyping. Within the scope of the work, a comprehensive image data set along growth period will be generated. The development of a suitable experimental setup for stable and repeatable image acquisition at constant height and perspective with suitable georeferencing, as well as the execution of the photo campaign are essential parts of this work.
Evaluation of the performance of machine learning method is an indispensable part of modern machine learning. In project we will explore techniques to improve the evaluation.
This master project will investigate novel use of emerging programmable network adaptors for accelerating Big data processing.
This master project will investigate novel use of emerging programmable network cards for improving the performance of the distributed in-memory machine learning.
In this project we will explore approaches to obtain general insight about the properties in a dataset, or in more insight about the environment in which the data was collected from.
Construction, evaluation and reasoning using knowledge graphs for software security assessments.
Missing data occurs a lot in practice: images that lack a piece, people don't fill in all values in a survey, and broken sensors lead to unavailable data,... You will try to fill in the missing entries with reasonable values or make a new model that can handle missing data directly
Interested in sport technology for videos used by elite leagues and live broadcasters? We have for long researched sports technology, both for distributing the video of live broadcasts and managing the archives in Norway and Sweden. We develop systems that are actively used and tested on real users, where the users can see highlights and summaries, get recommendations, collect favorite events, make personalized playlists, etc.
In this project, we are going to investigate the impact of different network parameters in VR as well as cloud gaming, and model the user experience based on that.
The goal is to apply existing quantum search algorithms or develop new quantum search and optimization algorithms to solve classical optimization problems.
Are you up for building real-world quantum software?
Automatic classification of sport news into different categories using state of the art Natural Language Processing (NLP). The project is building upon an existing dataset of Norwegian soccer news.
Predict uncertainty of superpixels for image segmentation.
Time series analysis using Residual neural networks. The main task is to explore how different ResNets can be combined to solve complex time series problems.
Developing methods based on AI techniques to discover unforeseen situations in Elevators or Train.
“Huldra” is a React-based framework for conducting custom online surveys, which can be used for collecting crowdsourced feedback on various types of multimedia content. We would like to extend this framework with added functionality, improve the implementation, and study the influence of UI design on user experience.