26.10.2025
CoProU-VO Wins GCPR 2025 Best Paper Award
Award-Winning Work by MCML Director Daniel Cremers and His Team for Advances in Unsupervised Visual Odometry
The paper "CoProU-VO: Combining Projected Uncertainty for End-to-End Unsupervised Monocular Visual Odometry" by MCML Director Daniel Cremers and Junior Members Weirong Chen and Johannes Meier, together with Jingchao Xie, Oussema Dhaouadi, and Jacques Kaiser received the Best Paper Award at GCPR 2025.
Their work presents CoProU-VO, an end-to-end unsupervised visual odometry framework that propagates and combines uncertainty across temporal frames to improve robustness in dynamic scenes. Built on vision transformer backbones, it jointly learns depth, uncertainty, and camera poses, achieving state-of-the-art performance on KITTI and nuScenes datasets.
Congratulations to the team on this outstanding achievement!
Check out the full paper:
CoProU-VO: Combining Projected Uncertainty for End-to-End Unsupervised Monocular Visual Odometry.
GCPR 2025 - German Conference on Pattern Recognition. Freiburg, Germany, Oct 23-26, 2025. Best Paper Award. To be published. Preprint available. arXiv
Related
©Juli Eberle / TUM / ediundsepp Gestaltungsgesellschaft
04.12.2025
World’s First Complete 3D Model of All Buildings Released
Xiaoxiang Zhu’s team releases GlobalBuildingAtlas, a high-res 3D map of 2.75B buildings for advanced urban and climate analysis.
04.12.2025
When to Say "I’m Not Sure": Making Language Models More Self-Aware
ICLR 2025 research by the groups of David Rügamer, and Bernd Bischl introduces methods to make LLMs more reliable by expressing uncertainty.
28.11.2025
MCML at NeurIPS 2025
MCML researchers are represented with 46 papers at NeurIPS 2025 (37 Main, and 9 Workshops).
27.11.2025
Seeing the Bigger Picture – One Detail at a Time
FLAIR, introduced by Zeynep Akata’s group at CVPR 2025, brings fine-grained, text-guided detail recognition to vision-language models.
