News – AI Lab
News
NorwAI&NAIL AI Seminar - Opting out of facial recognition
Yesterday we had the pleasure of hearing Professor Gavin Taylor, from the US Naval Academy talk about our possibilities of opting out of facial recognition. The underlying conclusion was that LowKey can make facial recognition systems misbehave and still have the image look very similar to the human eye. However, the issue might lie in implementing this service in practise in an increasingly digital and photographed world.
You can read their paper “LowKey: Leveraging Adversarial Attacks to Protect Social Media Users from Facial Recognition” here and also test LowKey in their online interface here. The slides from his presentation is also available here for those interested.
Thanks to all participants for the great turnout and interesting discussions following his talk!
NorwAI&NAIL AI Seminar - Small particles, big data: Machine learning at the Large Hadron Collider
Thanks to Steffen Mæland, associate professor at the Western Norway University of Applied Sciences and contributor to the ATLAS experiment at CERN, who gave a very exciting talk about the advantages and challenges they face using machine learning at the Large Hadron Collider.
He has kindly shared his slides from the talk with us here!
NorwAI&NAIL AI Seminar - Medical machine learning: from basic research to startups
Thank you so much to Lars Ailo Bongo, from UiT - Norges arktiske universitet, who gave a very interesting talk about the rollercoaster of failure and success that is medical machine learning!
Among other things, he stressed the importance of data availability and open access research. In the spirit of accessibility, he has kindly provided us with a link to view his presentation slides for those who are interested.
Also, a big thanks to Kerstin Bach for organising this edition of our AI seminar series with NorwAI.
NTNU to invest in more powerful infrastructure for AI
The AI and machine learning research community at NTNU is growing at a fast rate. This calls for more powerful computing resources, with secure and efficient handling of data shared between NTNU researchers and external partners as a key element. We are therefore happy that NTNU will invest in updating the infrastructure.
Learn more about the investment in this article.
New project "RICO" will make intelligent control systems more robust
We recently learned that the project "RICO - Robust Intelligent Control" has been awarded funding from the Norwegian Research Council. The aim of the project is to develop new methods to improve the robustness of intelligent control systems. The project will include partners TrønderEnergi, SINTEF Digital, Solution Seeker and Ohmia charging.
Learn more about the RICO-project in this article.
Summer reading list
Summer is the time for reading, and perhaps also the time for learning something new. If you'd like to gain some insight on AI this summer, you should check out our AI summer reading list.
Whether you are a beginner to AI or an AI enthusiast interested in deepening your knowledge, we have selected a few books to widen your AI horizon.
Successful second Nordic Probabilistic AI School 2021
The second Nordic Probabilistic AI School (also known as ProbAI school) was held virtually on June 14-18. The school brought together more than 200 students and a great team of lecturers and speakers from all over the world.
The mission of ProbAI is to create an inclusive environment serving a state-of-the-art expertise in probabilistic machine learning. Learn more about ProbAI 2021 in this article.
Zeabuz is a new NAIL partner
We're happy to announce that Zeabuz has joined us a partner!
If you're curious about sustainable mobility solutions, you should definitely take notice of Zeabuz. The Trondheim-based start-up develops technology and services for a new type of mobility system poised to revitalize our waterways.
The Zeabuz mobility system is a flexible network of electric, autonomous passenger ferries. Their ambition as a NAIL partner is to develop safe, secure and trustworthy AI, as well as utilizing computer vision and digital twins for development, testing and verification. We look forward to collaboration with Zeabuz on their mission to make waterborne mobility more efficient, flexible and sustainable.
Are you a student looking for an AI master's thesis project?
Will you write your master's thesis next year and need some inspiration for a thesis topic? Would you like to work with real-life challenges and data? In that case, we recommend that you take a look at the thesis projects submitted by Norwegian companies and organisations during our master's thesis event March 9.
The recording from the event is available for NTNU students and staff here.
In this folder you'll find descriptions of all the thesis projects presented during the pitch session, as well as a few more!
Se opptak fra webinar om kunstig intelligens i Norge
14. august samlet vi KI-eksperter fra akademia, næringsliv og offentlig sektor til webinar for å diskutere status for KI-utviklingen i Norge.
Vi fikk høre om den nasjonale KI-strategien og hvordan noen av landets største virksomheter arbeider med KI-teknologi. Vi diskuterte fremtiden til og med KI, og stilte det store spørsmålet kommer robotene egentlig til å ta over?
Gikk du glipp av webinaret? Da kan du se det i opptak her.
New project SupportPrim will provide recommendations to physiotherapists
The online magazine for Norwegian physiotherapists "Fysioterapeuten" covers the new project SupportPrim, which uses AI to give recommendations about possible treatment for patients. The SupportPrim systems takes into account individual differences between patients.
The project is a collaboration between NTNU's Musculoskeletal research group and the AI Community, with associate professor Kerstin Bach as main contact person from our side.
The project has received support from the Norwegian research council.
New SFI center, NorwAI, will develop novel ways to utilize data-driven AI in innovative solutions
We were thrilled to learn that the Norwegian Center for Research-based Artificial Intelligence innovation (NorwAI) will be granted status as an SFI center (center for research-based innovation) by the Norwegian Research Council.
NorwAI will develop breakthrough theories, methods and technology for efficient and trustable utilization of data-driven AI in innovative, industrial solutions, together with some of the most competent research environments on AI and big data, and companies with high technological ambitions
You can read about all the new SFI centers at NTNU in this article.
Trym Holter i Digitaliseringspådden
Nysgjerrig på kunstig intelligens (AI)? Hør Trym Holter, direktør for Norwegian Open AI Lab, gjeste Digitaliseringspådden, hvor han forteller om flere spennende AI-relaterte temaer, som talegjenkjenning, motlyd, "smart" teknologi, og gir deg AI-ordlista!
Du kan høre episoden her.
Visit by the Minister of Public Security
On May 13, 2019 we were delighted to welcome the Minister of Public Security, Ingvil Smines Tybring-Gjedde (Progress Party) to NTNU and our AI Lab.
Public security is an important aspect of the work we do at NTNU. Together with the excellent research community on cyber security at NTNU in Gjøvik, we presented some of this work to the Minister of Public Security.
Nemnoor EDIH (European Digital Innovation Hub) lanseres i mars!
Bli med på lansering av Norges nye senter for kunstig intelligens!
Kunstig intelligens har aldri vært mer omtalt, diskutert eller benyttet enn nå. For norske virksomheter betyr det at tilgang på riktig kompetanse, testsentre og nettverk vil være avgjørende i tiden fremover.
Nemonoor er et nytt nasjonalt senter for kunstig intelligens som skal hjelpe bedrifter med å ta i bruk teknologien. Senteret består av Norges fremste eksperter innen kunstig intelligens som skal veilede og tilby tjenester innen fire domener; produksjon, helse, energi og smarte byer/samfunn.
Meld deg på den digitale lanseringen 23. mars og få et innblikk i hva Nemonoor kan tilby din virksomhet!
Catch IDI 2023
A big thanks to our amazing students that took on the challenge to present their work at Catch IDI on February 24th 2023. We heard about four AI master-projects, all in only 15 minutes!
First, William Dalheim told us more about diffusion models. Then, Patrik Hammersborg talked on explainable AI and chess models such as AlphaZero. We heard about developing videogame agents using NeuroEvolution was presented by Martin Sondov Hallan, who works alongside Kristian Tveiten. Lastly, we heard about making self-driving cars using Imitation Learning and computer vision, as told by Aksel Lunde Aase, who is working with Mathias Wold.
We are positive that their glowing enthusiasm has inspired other students to delve into the world of AI as well!
Visit from the Ministry of Education and Research
On February 2nd 2023, the Faculty of Information Technology and Electrical Engineering (IE) had a visit from the Ministry of Education and Research (Kunnskapsdepartementet). As part of the visit, they guested us at the Norwegian Open AI Lab. Here they heard from different representatives at the Department of Computer Science (IDI) as well as the Department of Information Security and Communication Technology (IIK) at campus Gjøvik.
AI lab-leader and head of department, Heri Ramampiaro, spoke on AI’s role in politics and everyday life. He talked about our ongoing research projects and initiatives, stressing their societal relevance in an increasingly digital world.
To år med nasjonal KI-strategi – hvordan ligger vi an?
14. januar 2022 var det to år siden regjeringen lanserte Norges KI-strategi. Sammen med gode samarbeidspartnere har vi i anledning årsdagen gjort oss noen tanker rundt Norges KI-strategi og hvordan vi ligger an.
Language technology projects in NAIL
The interest in language technology, and particularly natural language processing (NLP), has increased rapidly in recent times, as a consequence of its many potential application areas and the technological advances that have been made in the field. However, the most impressive breakthroughs within NLP have been made in the English language. For Norwegian and the Scandinavian languages, there is still a job to be done.
Luckily, our Norwegian Open AI Lab partners and researchers are working on it. In this article you can learn more about our language technology projects.
Successful final review for SELFBACK
On May 27th, the SELFBACK EU project had its final review. The goal of the project has been to develop a decision support system to facilitate self-management for low back pain patients. The multidisciplinary team, lead by Paul Jarle Mork (ISM) and Kerstin Bach (IDI), has used Case-based Reasoning to personalize self-management recommendations delivered via a smartphone app. Partners in the project have been Robert Gordon University in Aberdeen, University of Glasgow, University of Southern Denmark, the National Research Center for the Working Environment in Denmark and SMEs Trade Expansion and Health Leads.
In the last two years, the effectiveness of SELFBACK has been evaluated in an international multi-center randomized controlled trial (RCT) using pain-related disability as primary outcome. A company (SelfBack APS) has been established in Denmark, projected to license, and commercialize the SELFBACK technology. While the project is officially closed, the work on analyzing the data is ongoing. In the months to come, more publications on the details of the technology and the RCT’s outcome will be published.
The top AI Master's Theses in 2020 awarded
Artificial Intelligence is a popular topic among students at NTNU. In fact, more than 200 master's students who graduated from the Faculty of Information Technology and Electrical Engineering last year submitted a thesis in which AI plays a substantial part.
On December 15th, we celebrated the top AI theses of 2020 in a digital event. Eivind Meyer, a graduate from Department of Engineering Cybernetics, was the lucky winner of the awards.
You can read more about the event and see all nominated theses in this article.
New projects to start up in 2021
When 2020 was coming to an end, several NAIL researchers got the great news that their project applications had been approved for funding. These are some of the successful projects to start up in 2021:
- In the project "TEFLON", to be financed by NordForsk, several Nordic partners will design digital language-learning games for Finnish, Swedish and Norwegian.
- "New Machine Learning Algorithms for the Calculation of Water Values" is a new research-based innovation project together with Statnett and the Norwegian start up, Optimeering.
- "SmaRTWork" will develop a digital system for facilitating the return to work for people suffering from muscle and skeletal diseases. The project is a collaboration between the AI group at the department of Computer Science and the department of public health and nursing.
You can find a list of all ongoing NAIL projects in our project overview.
New project will use AI methods to improve time series analysis
The new project "ML4ITS - Machine Learning for Irregular Time Series analysis" was recently approved for funding (16MNOK) by the Norwegian Research Council. The project aims to improve the analysis of irregular time series, by exploring ways to overcome the issue of limited available or labelled data for (multivariate) time series modelling. These are cases where heterogeneity of the data (e.g. non-stationarity, multi-resolution, irregular sampling) as well as noise, pose further challenges.
ML4ITS’s main objective is to advance the state-of-the-art in time series analysis for “irregular” time series by using modern AI methods. The methods will be applied and tested on industrial different use cases in partnership with Refinitiv, Telenor, and Exabel.
At the core of ML4ITS there is a cross-disciplinary environment, including Computer Science, Mathematical Sciences and Electronic Systems departments at NTNU. This is also the first collaborative platform between NAIL and the newly established strategic IoT initiative at NTNU to receive external funding.
EXAIGON: New project on explainable artificial intelligence
New project will analyse physical behaviour and sleep
NAIL researcher Kerstin Bach, along with colleagues at NTNU Department of Public Health and Nursing, will receive funding for an innovative project on physical behaviour and sleep. The project will explore the unique data in the HUNT study (Nord-Trøndelag Health Study) to examine how biological, socioeconomic, individual, and demographic determinants affect physical behaviour and sleep.
The HUNT data will provide insights to why there are large differences in the level of physical activity and the quality of sleep among the population. The funding is granted by NTNU Helse, one of NTNU’s strategic areas, and the project will run for three years.
Symposium of the Norwegian AI Society
The 2019 Symposium of the Norwegian AI Society took place from May 27-28 at NTNU, hosted by the Norwegian Open AI Lab. The symposium brought together researchers and practitioners in the field of artificial intelligence in Norway and Scandinavia to present ongoing work and future directions of AI.
The symposium provides a forum for networking among researchers as well as building and strengthening connections with related research fields, practitioners and businesses.
The new Nordic AI Network
The leading technical universities in Denmark, Finland, Norway and Sweden form the Nordic Five Tech alliance. In April 2019 the alliance announced the creation of the Nordic Artificial Intelligence Network.
The intention of the network is to bring together leading expertise in the field of AI and aim to turning the region into a global hub of AI research, education and innovation.
Learn more about the new AI network by reading this article.
AI for a broad audience
Our researcher Inga Strümke was recently at Lindmo speaking about AI, you can rewatch the program here (Norwegian).
Her book about AI, Maskiner som tenker - algoritmenes hemmelighet og veien til kunstig intelligens (2023), is also available for preorder now. The popular scientific book can be enjoyed by a broad audience as well as experts.
And if you are in Trondheim, you can attend the book launch at Litteraturhuset on the 8th of May!
NorwAI&NAIL AI Seminar - The AI Privacy Problem
On Friday 27th January our AI seminar series with NorwAI was guested by Professor Marija Slavkovik from the University of Bergen. She gave a talk about AI and privacy, shedding light on some important issues and distinctions to consider when building algorithms that use sensitive information.
Marija has kindly agreed to share her PowerPoint slides with us. We would like to thank Marija so much for the visit and the great talk!
Applications open for ProbAI 2023
You can now apply to join this year's Nordic Probabilistic AI School! The event will be in Trondheim from 12-16 June.
Read more and apply on the ProbAI website!
The latest celebrations at NAIL
This fall we have celebrated both the 5th anniversary of the AI-lab, as well as the 50th anniversary of the Department of Computer Science. In this article you can read more about the latest celebrations, challenges and optimism for the future.
Third annual Nordic Probabilistic AI School 2022
The third annual ProbAI summer school event was held in person in Helsinki Finland on June 13-17 of this year. From over 750 applicants, 150 participants were chosen to attend the event.
The program included inspiring talks, poster sessions and time to network and socialize. You can read more about ProbAI 2022 here.
The top AI theses of 2021 awarded
Artificial Intelligence is a popular topic among students at NTNU. In recent years, around 200 master’s theses in which AI plays a substantial part have been submitted yearly. On December 3rd, we celebrated the top AI theses of 2021 in an awards ceremony at Gløshaugen.
Sindre Stenen Blakseth, a 2021 graduate from the Department of Physics, was the lucky winner of the 2021 AI Master’s thesis awards.
Learn more about this year's AI Master's thesis awards by reading this article.
DeepInMotion to continue the discovery of new ways to detect CP in infants
Cerebral Palsy (CP) is the most common cause of physical disability in childhood which results in life-long needs for special services and is a huge burden for the child and family. Being able to diagnose infants at an early stage is crucial in order to prevent the disease from deteriorating.
Over the last years, NTNU researchers have worked on developing and improving Deep Learning methods to enable early detection of CP. They recently learned that they will be able to continue this work with the support of the Norwegian Research Council in the project DeepInMotion.
The project will focus on innovative explainable AI techniques and develop a smartphone-based clinical service implementation for early detection of CP. The XAI approach is particularly important when developing the smartphone app to be used by medical personnel, in order to give meaningful explanations for end-users and stakeholders. The techniques will therefore be implemented in close collaboration with hospitals. The NAIL contact person for this project is Heri Ramampiaro.
NAIL research featured in Wall Street Journal article
NAIL research on AI in aquaculture was recently featured in this article by Wall Street Journal.
In the article, our partners in Telenor Research shares the story of how AI is changing aquaculture in Norway, by contributing to saving costs and improving efficiency in raising salmon.
If you are interested in the research behind this story, take a look at these two academic articles:
A spatio-temporal recurrent network for salmon feeding action recognition from underwater videos in aquaculture by Håkon Måløy, Agnar Aamodt ang Mismi Ekrem.
FishNet: A Unified Embedding for Salmon Recognition by Bjørn Magnus Matthiesen, Kerstin Bach, Espen Meidell, Håkon Måløy and Edvard S. Sjøblom.
Gratis kurs om kunstig intelligens på norsk
Lurer på du på hva kunstig intelligens (KI) er? NTNU og Norwegian Open AI Lab tilbyr et online introduksjonskurs om kunstig intelligens. Kurset kombinerer teori med praktiske oppgaver, og du kan fullføre det i ditt eget tempo. Og det er helt gratis!
Students developed app to monitor air quality
Master students at NTNU developed the app “Lufta” to keep the inhabitants of Trondheim better informed about the air quality in the city.
By utilizing data from small sensors located in both permanent locations and on cars and buses, they are able to provide improved measurements of the air quality in Trondheim. The students have worked with IoT technology, as the sensors are connected to the 4G network, which enables wireless and immediate collection of the data. They have also worked with machine learning and prediction of how the air quality will be in the future.
The project was initiated by Telenor and has been carried out in close cooperation between NTNU students, AI Lab researchers and Telenor. You can learn more about the project by reading this article (in Norwegian).
AI Challenge - the Future of Artificial Intelligence
On June 18, the AI Challenge conference took place - a full-day conference about the future of artificial intelligence and a side event to NTNU´s science festival "Big Challenge". The conference was co-organized by NTNU Digital, Sintef Digital and the Norwegian Open AI Lab.
We had several interesting speakers: Kamalika Das (RIACS/USRA, NASA ARC), Nicola Palmarini (MIT-IBM Watson AI Lab), Teemu Roos (University of Helsinki), Silvija Seres, Astrid Undheim (Telenor Research), Roger Moore (University of Sheffield), Abhishek Thakur (Boost.ai), and our own Rudolf Mester.
Have a look at this article by Universitetsavisa (in Norwegian) to learn more about the conference.
Machine learning used to recognise salmon
Meet former NTNU students, Espen Meidell and Edvard Schreiner Sjøblom, who took a deep dive into the ocean while writing their master's thesis during spring 2019. They focused on how machine learning can be used to recognise individual salmon.
In this video (in Norwegian) they explain how it was done, and why recognizing salmon can be useful for the salmon industry. This work was done in cooperation with Sealab Ocean Group and Norwegian Open AI Lab.
Ole Jakob Mengshoel on Lørn.tech
In this podcast episode of Lørn.tech, Ole Jakob Mengshoel is interviewed by Silvija Seres. They talk about how AI has developed and how it can be used to solve global challenges, while improving the newest technology (in Norwegian) https://lorn.tech/lorn-pod/157-ai-ole-jakob-mengshoel-ai-in-theory-and-practice/
There you can also find several other interesting interviews, among others with our partners Telenor, Sintef and Norsk Regnesentral.
Nordic Probabilistic AI School
The first edition of the Nordic Probabilistic AI School (ProbAI) was organised June 3-7 2019. The school offered state-of-the-art expertise in machine learning and artificial intelligence to the public, students, academia and industry.
This was an intermediate to advanced level school with a particular focus on probabilistic models and deep generative models, covering the topics of latent variable models, inference with sampling and variational approximations, probabilistic programming and tools.
The event was a big success, gathering more than 150 participants from all over the world. In addition to an interesting and challenging academic program, the participants attended various social events, making it a great place to network and meet new people. We are already excited about repeating the summer school in the coming years.
Successful AI Hackathon
More than 50 NTNU students spent the weekend March 1-3 at NTNU attending the 2019 AI Hackathon. The event was organized by Telenor and our student organization, Brain.
The students worked with air quality data from Trondheim kommune, gathered by Telenor sensors. They showed enthusiasm and lots of creativity while working on the data. In the end, the winner team was Tannhjulsridderne - congratulations!
