Computer Vision

Since March 2020, I have been working at the Centre for the Advanced Study of Collective Behaviour. Biologists, psychologists, computer scientists and many more disciplines interdisciplinary work together and study collective behaviour across a wide range of species and across scales of organization.

I am a computer scientist within the Computer Vision and Image Analysis group. We focus on multi-object tracking and pose estimation methods for animals. We acquire data in our imaging facilities – the Imaging Barn and the Imaging Hangar – which are located in Möggingen at the Max Planck Institute of Animal Behavior and in the Centre for Visual Computing of Collectives at the University of Konstanz respectively.

Contact

e-mail: urs.waldmann@uni-konstanz.de
github: urs-waldmann
mastodon: @ursNwaldmann@sigmoid.social
X: @ursNwaldmann

News

2024/07/04 – uni’kon features our latest work „3D-MuPPET“ in #79 Freiheit, pp. 24-27. [German]

2024/05/27 – Press release about our latest work „3D-MuPPET“. [English, German]
2024/03/28 – 3D-MuPPET accepted for publication in the International Journal of Computer Vision. [paper]
2024/01/18 – We are organizing the 4th CV4Animals workshop at CVPR 2024.
News Archive

Publications

Urs Waldmann*, Alex Chan*, Hemal Naik, Máté Nagy, Iain D. Couzin, Oliver Deussen, Bastian Goldluecke, and Fumihiro Kano: 3D-MuPPET: 3D Multi-Pigeon Pose Estimation and Tracking. International Journal of Computer Vision, 2024. (* Authors contributed equally.) [project page, press release]

Urs Waldmann, Ole Johannsen, and Bastian Goldluecke: Neural Texture Puppeteer: A Framework for Neural Geometry and Texture Rendering of Articulated Shapes, Enabling Re-Identification at Interactive Speed. In Winter Conference on Applications of Computer Vision (WACV) Workshops, 2024 (oral presentation). [arXiv, code, video]

Simon Giebenhain*, Urs Waldmann*, Ole Johannsen, and Bastian Goldluecke: Neural Puppeteer: Keypoint-Based Neural Rendering of Dynamic Shapes. In Asian Conference on Computer Vision (ACCV), 2022. (* Authors contributed equally.) [project page, press release]

Urs Waldmann, Hemal Naik, Nagy Máté, Fumihiro Kano, Iain D. Couzin, Oliver Deussen, and Bastian Goldlücke: I-MuPPET: Interactive Multi-Pigeon Pose Estimation and Tracking. In Pattern Recognition (Proc. of DAGM GCPR), 2022 (oral presentation). [project page]

Urs Waldmann, Jannik Bamberger, Ole Johannsen, Oliver Deussen, and Bastian Goldlücke: Improving Unsupervised Label Propagation for Pose Tracking and Video Object Segmentation. In Pattern Recognition (Proc. of DAGM GCPR), 2022. [project page]

Peer-Reviewed Workshops

Urs Waldmann*, Alex Chan*, Hemal Naik, Máté Nagy, Iain D. Couzin, Oliver Deussen, Bastian Goldluecke, and Fumihiro Kano: 3D-MuPPET: 3D Multi-Pigeon Pose Estimation and Tracking. Accepted in track 1 (unpublished work) in CVPRW „CV4Animals“, 2023. (* Authors contributed equally.) [workshop page]

Urs Waldmann, Hemal Naik, Nagy Máté, Fumihiro Kano, Iain D. Couzin, Oliver Deussen, and Bastian Goldlücke: I-MuPPET: Interactive Multi-Pigeon Pose Estimation and Tracking. Accepted in track 2 (published work) in CVPRW „CV4Animals“, 2023. [workshop page]

Urs Waldmann, Jannik Bamberger, Ole Johannsen, Oliver Deussen, and Bastian Goldlücke: Improving Unsupervised Label Propagation for Pose Tracking and Video Object Segmentation. Accepted in track 2 (published work) in CVPRW „CV4Animals“, 2023. [workshop page]