Thomas Kreutz

Thomas Kreutz

ML for Urban Scene Understanding

Hi, I’m Thomas. 🤗

I have been a Ph.D. candidate (02/2022 - 07/2025) at the Technical University of Darmstadt and recently submitted and successfully defended my dissertation. My work mainly evolves around LiDAR-based scene understanding tasks and (multi-modal) representation learning. More specifically, I have worked on various interesting problems and areas, including LiDAR motion segmentation, LiDAR semantic segmentation, crowd simulation, and multi-modal contrastive learning. Please find my peer-reviewed publications below if you are interested in my work.

Recent highlights

  • [Jun, 2025] 🔥 Paper accepted to ICCV 2025
  • [Jan, 2025] 🤘 Paper accepted to ICRA 2025
  • [Jul, 2024] 🤘 Paper accepted to IROS 2024 [Oral]
  • [April, 2024] 🤘 Paper accepted to the SynData4CV Workshop@ CVPR 2024
  • [Feb, 2024] 🤘 Paper accepted to the Second Tiny Papers Track at ICLR 2024 [Invite to present]
  • [Oct, 2022] 🤘 Paper accepted to WACV 2023
  • [Jul, 2022] 🔥 Got accepted into the Software Campus (2022) program. Update Dec, 2023: Finished all 6 trainings on methods, social competences, and leadership.
  • [Jun, 2022] 🤘 Paper accepted to ITSC 2022

Peer-reviewed Publications

a
DeSPITE: Exploring Contrastive Deep Skeleton-Pointcloud-IMU-Text Embeddings for Advanced Point Cloud Human Activity Understanding
Thomas Kreutz, Max Mühlhäuser, Alejandro Sanchez Guinea
ICCV 2025
Arxiv | Project Page | Code
a
Whenever, Wherever: Towards Orchestrating Crowd Simulations with Spatio-Temporal Spawn Dynamics
Thomas Kreutz, Max Mühlhäuser, Alejandro Sanchez Guinea
ICRA 2025
Arxiv | Project Page | Code
a
LiOn-XA: Unsupervised Domain Adaptation via LiDAR-Only Cross-Modal Adversarial Training
Thomas Kreutz, Jens Lemke, Max Mühlhäuser, Alejandro Sanchez Guinea
IROS 2024 [Oral]
PDF | Code
b
Implicit Neural Clustering
Thomas Kreutz, Max Mühlhäuser, Alejandro Sanchez Guinea
Synthetic Data for Computer Vision Workshop@ CVPR 2024
PDF
b
Common Sense Initialization of Mixture Density Networks for Motion Planning with Overestimated Number of Components
Thomas Kreutz, Max Mühlhäuser, Alejandro Sanchez Guinea
The Second Tiny Papers Track at ICLR 2024 [Invite to present]
PDF
b
Unsupervised 4D LiDAR Moving Object Segmentation in Stationary Settings With Multivariate Occupancy Time Series
Thomas Kreutz, Max Mühlhäuser, Alejandro Sanchez Guinea
WACV 2023
PDF | Code
b
Unsupervised Driving Event Discovery based on Vehicle CAN-Data
Thomas Kreutz, Ousama Esbel, Max Mühlhäuser, Alejandro Sanchez Guinea
ITSC 2022
PDF