Free webinar – CV4DT: Computer vision for digital twins, where perception meets geometry

The Polish Society of Photogrammetry and Remote Sensing invites you to a free webinar entitled

“CV4DT: Computer Vision for Digital Twins, where perception meets geometry”,

which will be led by

Dr Olaf Wysocki from the University of Cambridge

The event will take place on 5 November 2025 at 12 noon on Microsoft Teams.

To take part in the webinar, please register via the link below:  https://forms.gle/EJ3KbBejRugBYPjGA

Abstract:
This presentation will provide an insight into the vision of the CV4DT group at the University of Cambridge, led by Olaf Wysocki.

The aim of the CV4DT group is to bring together researchers seeking to push the boundaries of machine learning, photogrammetry and computer vision. The CV4DT research programme focuses on the following areas:
a) three-dimensional understanding of semantics,
b) three-dimensional semantic reconstruction,
c) three-dimensional models acting as sensors,
d) quantification of uncertainty – encompassing all the aforementioned aspects.
The rationale for this approach lies in the interdependence of these issues, which are essential for creating digital twins based on diverse sensory data, thereby enabling digital simulations to be carried out before taking action in the real world.

In this context, a digital twin is understood not only as a three-dimensional geometric representation of reality, but as a three-dimensional model comprising:
a) a minimal, realistic and tight geometric representation,
b) hierarchical semantics,
c) estimated uncertainty regarding both the predicted semantics and the geometry,
enabling updates when new data becomes available.

The webinar presentation is available to download:  presentation (PDF)

Author about himself:

I conduct research as an Assistant Research Professor at the University of Cambridge and lead the CV4DT group. I was a research assistant specialising in 3D object reconstruction and semantic point cloud segmentation at the Department of Photogrammetry and Remote Sensing at the Technical University of Munich, where I obtained my PhD with distinction (summa cum laude) in photogrammetry and remote sensing under the supervision of Prof. Thomas H. Kolbe and the late Prof. Uwe Stilla. A few years ago, I obtained a Master’s degree in Geodesy and Geoinformatics, specialising in photogrammetry, at the Technical University of Munich. I had the opportunity to undertake an internship and complete my Master’s thesis on HD maps for automated driving at Audi AG, and prior to that I worked as a geomatics engineer at a company specialising in 3D city reconstruction. I am currently a member of Runder Tisch GIS, the TUM Data Innovation Lab and Get Kids into Survey. I value interdisciplinary collaboration with distinguished professors and their laboratories, including Yan Xia, Christoph Holst, Boris Jutzi, Filip Biljecki and Daniel Cremers.