Personal

My name is Gertjan Burghouts. I am a researcher in deep learning and computer vision. My passion is to create new algorithms and models that learn to understand the visual world around us.

I am part of a great team at TNO: the Intelligent Imaging group. Our research is about image & video understanding and robotics. My personal ambition is to enable learning with few labels, to leverage prior knowledge in the learning process. Moreover, to put the models in practice: robots, forensic search and decision-making.

Our team is part of DARPA’s program Learning with Less Labels. I am a visiting researcher at the University of Amsterdam, at the VISlab led by prof. Cees Snoek.

We collaborate within the OpenBots Lab with the University of Amsterdam, TU Delft, Marechaussee, as part of the NWO Open Technology Program. We pursue open-world AI models, that are not limited to closed sets of concepts. Cees and me supervise two PhD researchers on Embodied Foundation Models.

In EDF FaRADAI we study Frugal AI, learning to efficiently change domains and learning to learn new concepts without forgetting. In the NWO Perspectief program FIND, we study foundation models, their adaptation, and uncertainty quantification.

More details about my research can be found under Research.