Institut für Informationsverarbeitung (TNT)

Tom Wehrbein studied computer science at Leibniz University Hannover. In 2017, he was awarded the "Preis des Präsidiums" for his excellent academic achievements. For his bachelor's thesis "Representation of Human Motion using Neural Networks" he received the "KI Talente 2019" award. His master's thesis "Deep Learning Based Probabilistic 3D Human Pose Estimation" dealt with generating multiple plausible 3D poses consistent with an image. He received his master's degree in November 2020.

Since December 2020, he has been working towards his PhD at the Institut für Informationsverarbeitung (TNT) at Leibniz University Hannover. His main research interests are 3D human motion capture and normalizing flows. In 2022, he interned at Epic Games in Pittsburgh, USA, where he worked on 3D human pose and shape estimation.

Publications

Show selected publications only

Tom Wehrbein, Marco Rudolph, Bodo Rosenhahn, Bastian Wandt
Utilizing Uncertainty in 2D Pose Detectors for Probabilistic 3D Human Mesh Recovery
IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), IEEE, February 2025
(pdfGitHub) BibTeX
Tom Wehrbein, Bodo Rosenhahn, Iain Matthews, Carsten Stoll
Personalized 3D Human Pose and Shape Refinement
International Conference on Computer Vision Workshops (ICCVW), IEEE, Paris, France, October 2023
(pdf, pdfLink) BibTeX
Marco Rudolph, Tom Wehrbein, Bodo Rosenhahn, Bastian Wandt
Asymmetric Student-Teacher Networks for Industrial Anomaly Detection
Winter Conference on Applications of Computer Vision (WACV), IEEE, Hawaii, USA, January 2023
(pdfarXiv.org) BibTeX
Ilona Rosenboom, Tobias Scheithauer, Fabian C. Friedrich, Sophia Pörtner, Lisa Hollstein, Marie‑Madlen Pust, Konstantinos Sifakis, Tom Wehrbein, Bodo Rosenhahn, Lutz Wiehlmann, Patrick Chhatwal, Burkhard Tümmler, Colin F Davenport
Wochenende - modular and flexible alignment-based shotgun metagenome analysis
BMC Genomics, Springer Nature, November 2022
(pdfLink) BibTeX
Marco Rudolph, Tom Wehrbein, Bodo Rosenhahn, Bastian Wandt
Fully Convolutional Cross-Scale-Flows for Image-based Defect Detection
Winter Conference on Applications of Computer Vision (WACV), IEEE, Hawaii, USA, January 2022
(pdfarXiv.org, GitHub) BibTeX
Tom Wehrbein, Marco Rudolph, Bodo Rosenhahn, Bastian Wandt
Probabilistic Monocular 3D Human Pose Estimation with Normalizing Flows
International Conference on Computer Vision (ICCV), IEEE, October 2021
(pdfarXiv.org, GitHub) BibTeX