Awesome IT 2022
April 8th 2022
Het Sieraad, Postjesweg 1, 1057 DT Amsterdam
On the 8th of April 2022, the tenth edition of Awesome IT will be held. Join us for a day of informative and inspiring talks from experts in a wide range of IT-related fields.
Click on a name on the left for more information about a speaker
I'm an assistant professor for computer vision and machine learning at the QUVA lab at the University of Amsterdam, where I work with Cees Snoek, Max Welling and Efstratios Gavves. My PhD was at the Visual Geometry Group (VGG) at the University of Oxford where I worked with Andrea Vedaldi and Christian Rupprecht. Prior to this I studied physics at the University of Munich (LMU) and Economics in Hagen as well as a MSc in Mathematical Modelling and Scientific Computing at the Mathematical Institute in Oxford.
Talk: Self-supervised Learning: What do we need for learning from raw data?
in this brief talk I will talk about recent developments that show what can be achieved in deep learning for computer vision without using any human annotations, the field known as "self-supervised learning". After a first overview of what self-supervised learning is, we will first dive into how clustering can be combined with representation learning using a single loss and analysize how semantic the resulting clusters are (paper  at ICLR'20). Second, I will present a novel large-scale dataset without any pictures of humans as a more privacy-secure replacement for ImageNet for the purpose of developing self-supervised pretraining methods (paper  at NeurIPS-Data'21). Finally, I will show how the principles from self-supervised learning can be extended to the multi-modal (ie video + audio) domain. In this paper  (that will be presented at CVPR'22), we show how self-supervised clustering and contrastive learning can be combined for detecting objects fully unsupervisedly solely by utilizing multi-modal signals from a video dataset.
Domain: AI, computer vision, ethics in AI
Frank van Harmelen
Frank van Harmelen (1960) is a professor in Knowledge Representation & Reasoning at the Vrije Universiteit Amsterdam, with a PhD from the University of Edinburgh. He co-defined the Web ontology language OWL, which has become a worldwide standard, is in wide commercial use, and has become the basis for an entire research community. He co-authored the Semantic Web Primer, the first academic textbook of the field. He received the 10-year impact award of the International Semantic Web Conference, and he is one of the editors of the Handbook of Knowledge Representation, the standard reference in his field. He is principal investigator of the Hybrid Intelligence Centre, a 20m€, 10 year research project into AI systems that collaborate with people instead of replacing them. His current interests focus on methods to combine knowledge representation and machine learning in both directions. He is a fellow of the European AI Society ECCAI and the Asia-Pacific AI Association, he is a member of the Academia Europaea, and the Royal Netherlands Society of Sciences and Humanities (KNAW). He is a guest professor at the Wuhan University of Science and Technology (WUST) in Wuhan, China.
Talk: A research agenda for hybrid intelligence
Much of current AI research is implicitly aimed at building systems that replace humans: self-driving cars to replace Uber drivers, translation software replacing interpreters, image analysis software replacing radiologists. But it's becoming increasingly clear that machine intelligence will be rather different from human intelligence. It is therefore more interesting to build AI systems that collaborate in hybrid teams of people and machine, in order to combine their complementary skills. This will require that we start asking a whole set of new research questions. How to equip AI systems with a "theory of mind" to make them collaborative? How to make AI systems adaptive to changes in the team and the environment? How to instill moral values into these systems? And of course how to make them explainable? We will outline a research agenda for hybrid intelligence and present some early results from researchers worldwide into hybrid intelligence.
Domain: Theory of mind, Hybrid Intelligence, AI
Chris probeert het leven van mensen makkelijker, slimmer en vooral leuker maken. Als vader van twee, als werkgever, en als maker. Hij is mede-eigenaar van Q42, Hack The Planet, Micrio, TurtleBlaze en 42Puzzles, en investeerder in LessonUp en Mibo. Bij Q42 is hij in verschillende rollen betrokken bij het maken van digitale producten voor de Efteling, Rijksmuseum, Philips Hue, HEMA, PostNL, Madurodam, het Kinderboekenmuseum, en meer. Zijn vrije tijd besteedt hij graag aan reizen met zijn gezin, motorrijden, gitaarspelen, metal luisteren - en voornemen weer te gaan sporten.
Talk: Technologie die niet te onderscheiden is van magie (DUTCH ONLY!))
“Any sufficiently advanced technology is indistinguishable from magic,” is een befaamde quote van futurist en science-fiction schrijver Arthur C. Clarke. Bij Q42 gebruiken we het graag als inspiratie voor ons werk. Namelijk het bouwen van digitale producten die het leven van mensen makkelijker, slimmer en leuker maken.
De overtreffende trap daarvan is magische belevingen creëren voor kinderen. Dat vinden we het mooiste dat er is. Misschien omdat we zelf stiekem nooit volwassen zijn geworden. En omdat we graag geloven dat programmeren super powers geeft.
Door de jaren heen hebben we allerlei magische digitale producten ontwikkeld. Van een pratende toverspiegel voor de Efteling en toiletten die je vertellen waar je de minste kans op onprettige geurtjes hebt, tot een prototype van een magisch voorleesboek. Dat laatste idee leidde uiteindelijk tot de Voorleestent van het Kinderboekenmuseum. Die reis, van concept tot Voorleestent, daar zal ik alles over vertellen. Over hoe de eerste demo faliekant mislukte, over wat we daarvan geleerd hebben, en over hoe we uiteindelijk iets neer hebben kunnen zetten dat simpel, robuust en magisch is.
En over misschien wel het belangrijkste dat we geleerd hebben: dat Arthur C. Clarke het mis had.
Domain: Immersive experiences
Joris Mollinga is the co-founder of DuckDuckGoose. He has experience in high performance machine learning, computer vision projects and AI applications.
Talk: Designing Deepfake detection software: Challenges, Solutions and Lessons Learnt
In this talk Joris Mollinga will outline the technical difficulties they faced when designing deepfake detection software, and how they overcame these challenges. He will share his insights and best practices of converting machine learning models from research to production.
Domain: AI, Computer Vision, deep learning,
Jaap Kaandorp received his MSc in biology and a PhD in computer science and mathematics, both from the University of Amsterdam. He did research as a postdoctoral fellow at the Department of Computer Science of the University of Calgary in Canada. Currently he has a permanent position as an associate professor at the Computational Science Lab of the Faculty of Science of the University of Amsterdam. He runs a group on computational biology of 5 MSc and 2 Phd students at the Computational Science Lab of the Faculty of Science of the University of Amsterdam. The group is doing research at a range of different levels of organisation in computational biology: modelling gene regulatory networks controlling morphogenesis, cell-based modelling of embryogenesis, physiology of biomineralization, modelling the emergence of growth and form of biomineralized structures and organisms and quantification of complex-shaped three-dimensional structures.
Talk: Modelling gene regulation and cells in embryogenesis Jaap Kaandorp
Rather than being directed by a central control mechanism, embryogenesis can be viewed as an emergent behavior resulting from a complex system in which several sub-processes on very different temporal and spatial scales (ranging from nanometer and nanoseconds to cm and days) are connected into a multi-scale system. In our research we have been focusing on the embryogenesis of basal organisms like sea anemones, jelly fish and stony corals. During the talk I will introduce some methods for spatio-temporal modelling of biological processes We have developed methods for analysing spatio-temporal gene expression patterns, methods for spatio-temporal modelling and inferring gene regulatory networks from gene expression data (qPCR data and in–situ hybridizations) and a cell-based mechanical model of early embryogenesis.
Domain: Computational Biology, Gene modelling,
Mattijs Ghijsen obtained a MSc in computer science at the University of Twente and a PhD at the University of Amsterdam. He has worked as researcher at VITO (Flemish Institute for Technological Research) and as a post-doc at University of Amsterdam. After his career in research he started working as a senior technical consultant at the Software Improvement Group and he currently the head of Product Engineering at Navara Consulting Services. Mattijs has a passion for building software development teams and agile product development.
Talk: Building High Performance Software Development Teams
In 2018, a groundbreaking piece of research by Nicole Forsgren was published in the book Accelerate and in the past years this research has been transforming the IT industry. It guides software developers to make design choices, it helps teams to organise their way of working and it shows entire IT organisations how to manage software development. In this talk I will take you through this research. I'll show some remarkable trends in team performance and discuss examples that show how the ideas from this research can be applied in practice. At the end of this talk you will understand that in order to deliver high quality software you have to move fast, and that in order to move fast you have to deliver high quality software. And as a bonus you will know the four questions you should definitely ask during any job interview in the IT industry.
Domain: Software engineering
Dr. Ir. Clara Maathuis is Assistant Professor in AI & Cyber Operations at Open University. She has a technical-military background, holds a PhD in Military Cyber Operations & AI, and conducts multidisciplinary research in Cyber/Information Operations, AI, military technologies, social manipulation, and digital innovation. At the same time, she is involved in developing and providing AI/ML and cyber security courses.
Talk: Intelligent Social Manipulation in the Ongoing Coronavirus Pandemic
The ongoing Coronavirus pandemic rewrote several dimensions of our lives and made us build, project, and take major steps for functioning and living in the new digitalized world. Nevertheless, this made us directly exposed to both known and just-developed unknown methods that have (depending on the case) on a specific personal degree influenced our perceptions and believes towards different aspects and perspectives regarding the origin, significance, and impact of Coronavirus itself as well as its corresponding vaccination campaigns. Such actions require intelligence through intelligent methods used, motivation, and dedication for i) attackers to build efficient and effective social manipulation campaigns, ii) defenders to build efficient and effective awareness, detection, and prevention social manipulation strategies and mechanisms, and iii) targets and victims to deal with and learn from social manipulation campaigns. In this presentation, we will adopt these three angles to understand what intelligent social manipulation is.
Domain: social media manipulation, cyber security, AI, Coronavirus
Gerrit Oomens graduated from the University of Amsterdam in 2012 with a MSc. in Mathematics. He is now employed at the Education Service Center of the Faculty of Science, where he works to simplify the administrative systems used by the university. During his studies, he developed the FNWI-specific DataNose system, which most of you have probably used at some point in your studies. Datanose has grown into a considerable system that is used for a broad variety of processes within the faculty.
Talk: IT at the UvA: DataNose and more
The University of Amsterdam is a large organisation that makes use of an impressive and ever growing number of information systems. Some of these are visible to most students and teachers, such as SIS and Canvas; others are hidden to most users. In this talk, a behind the scenes perspective on these systems will be given, including a demo of a number of less visible UvA systems. The overall architecture of these systems will be discussed and the rationale behind the faculty’s custom system DataNose. We will discuss the history, the current technical state and the possible future of this system and look into how and why the UvA makes certain decisions with regards to IT systems.
Click or tap on a talk for more information.
Plenary Welcome09:45 - 10:00
Frank van Harmelen10:00 - 11:00
A research agenda for hybrid intelligence
Yuki Asano10:00 - 11:00
Self-supervised Learning: What do we need for learning from raw data?
Clara Maathuis11:30 - 12:30
Intelligent Social Manipulation in the Ongoing Coronavirus Pandemic
Matthijs Ghijsen11:30 - 12:30
Building High Performance Software Development Teams
Lunch12:15 - 13:15
Joris Mollinga13:30 - 14:30
Designing Deepfake detection software: Challenges, Solutions and Lessons Learnt
Jaap Kaandorp13:30 - 14:30
Modelling gene regulation and cells in embryogenesis Jaap Kaandorp
Gerrit Oomens13:30 - 14:30
IT at the UvA: DataNose and more
Speeddates13:30 - 15:15
Chris Waalberg14:45 - 15:45
Technologie die niet te onderscheiden is van magie (DUTCH ONLY!))
Plenary Closing16:00 - 16:15
Drinks16:15 - 17:15
Next to all the exciting speeches, there will be a chance to be heard and see what the future has in stake for you. In just 30 minutes of your time, you will get to “date” with several interesting companies, and talk about what you can do for them and what they can do for you. Even if you do not know yet what you want or simply want to get some more information about the companies, it's well worth your time. So don't forget to register for speeddating and get the full experience of Awesome IT!
Awesome IT 2021 pictures
We have throughly enjoyed organizing AwesomeIT this year and hope that all of you did as well. We would like to give a big thank you for being with us and are are looking forward to seeing you again next year!
For the photos of Awesome IT 2022, look here.