I am a professor in Artificial Intelligence at the Cognitive Science & AI department at Tilburg University and at the Jheronimus Academy of Data Science in ‘s-Hertogenbosch, a joint initiative of Eindhoven University of Technology and Tilburg University.
I perform AI research (see below), teach courses on AI, and give “down-to-earth” lectures for organisations and industry on the latest developments and impact of AI.
My research focuses on pattern recognition in humans and machines. Although the term “pattern recognition” is not in vogue anymore, it does capture the human capability to perceive patterns in images, signals, and data in general. Nowadays, deep learning and generative AI algorithms excel in pattern recognition on narrow domains. My research interests and activities are listed below.
Unidentified Anomalous Phenomena in Astronomy Data
Most of my scientific work addresses pattern-recognition tasks that can be performed by humans but also by machines. My current focus is on astronomy as an application domain. Astronomy provides a continuous stream of data that can be analysed for anomalous or “interesting” patterns (e.g., Unidentified Anomalous Phenomena) by means of AI. Together with Koko Visser and Bas Bosma, we have worked on AI for the detection of exoplanets from light curves. Essentially, the task comes down to the detection of repeating dips in time series. Humans perform this task using their pattern recognition skills and understanding of stellar and planetary dynamics, whereas AI algorithms perform brute-force pattern recognition. My current research is focussing on detecting anomalous or repeating patterns in astronomical time-series data.
AI based Art Attribution
The use of AI for art attribution, e.g., determining the authenticity of an alleged Van Gogh painting, has been a small (but very media savvy) part of my research since 2000. The task of attributing artworks to their artists is a skill mastered by human connaisseurs but can also be performed by AI algorithms. What fascinates me is how AI algorithms differ in their performances from humans. AI algorithms are not hindered by biases about the economic consequences of attributions, whereas humans have a much richer understanding of the meaning and historical context of artworks. I am an (unpaid) adviser to the Swiss company Art Recognition (led by Carina Popovici) that offers AI-based authenticity verification. With a leading role by Ludovica Schaerf, I participated in two studies, one on art authentication using Vision Transformers and one on AI-generated images to support forgery detection.
ILUSTRE
I am also involved in application-oriented projects that focus on the integration of AI techniques to support a more sustainable future. The ILUSTRE project (at JADS) was initiated together with Renato Calzone (now at TUe) and Rigo Selassa (LaNubia) and studies how AI can be used for water and energy management in Curaçao and the Caribbean region. The project is part of the nation-wide ROBUST program and hosts four PhD researchers and one postdoc researcher. Currently, together with Uzay Kaymak (JADS/TUe) and Çiçek Güven (TiU), I coordinate the project.
Other projects
Some other projects I am involved in: HAICu, SteadFast, Child Growth Monitor, CERTIF-AI (at JADS).
AI algorithms provide powerful tools that in many ways exceed the human pattern-recognition capabilities, but require careful guidance by human domain experts that are aware of the strengths and limitations of AI.
I am member of SIGAI, IPN, the Lorenz center Computational Science Board, Scientific Board member of Kennisnet, and board member of the NL Advisory Council of CLAIRE.