About me

I am an AI Research Scientist at Odyssey, specializing in spatial intelligence, generative world models, and 3D dynamics. My research focuses on bridging the gap between static scene representation and temporally dynamic, physical simulations. Currently, my work focuses on pushing the frontiers of multi-modal autoregressive generation, diffusion models, and (multi-agent) reinforcement learning frameworks.

I completed my PhD in Computer Vision at the Technical University of Munich (TUM) under the supervision of Prof. Dr. Laura Leal-Taixé, where I also earned a Bachelor's in Mechanical Engineering and a Master's in Robotics, Cognition, and Intelligence. Throughout my research, I have had the privilege of collaborating with leading groups across industry and academia, including a 10-month research internship at NVIDIA focusing on 3D detection, a 5-month stay at Carnegie Mellon University working on 3D point tracking and reconstruction, and a subsequent 4-month return to NVIDIA focusing on RGB-D diffusion for 3D reconstruction.

When I’m not thinking about latent rollouts and physical simulators, I am usually chasing down green spaces. I’m an experienced hiker, a distance runner (always down for a 9K to clear the mind), and currently embracing the beginner’s mindset in surfing, kite-surfing, and motorcycling. Otherwise, you can find me sitting down somewhere quiet with a good book!



Updates

  • 2026/05/26: Checkout Starchild-1, a autoregressive audio-video model for joint audio-video generation!
  • 2026/05/12: Check out our new paper PROWL, a reinforcement Learning framework to make the improve action conditioned world models!
  • 2025/06/02: I joined the research staff at Odyssey in London pushing research boundaries towards interactive generative models in a more product focused way!
  • 2024/11/05: Our paper “DynOMo: Online Point Tracking by Dynamic Online Monocular Gaussian Reconstruction” got accepted to 3DV 2025!
  • 2024/08/29: Our paper “DeformGS: Scene Flow in Highly Deformable Scenes for Deformable Object Manipulation” got accepted to WAFR 2024!
  • 2024/08/12: I started another internship with NVIDIA in Santa Clara, CA!
  • 2024/02/26: Our paper “SeMoLi: What Moves Together Belongs Together” got accepted to CVPR 2024!
  • 2023/12/15: I will join the group of Prof. Ramanan at CMU for a visiting PhD from Feburary!
  • 2023/12/05: After submitting the results fo my internship to CVPR I’m back at the University!
  • 2023/02/25: Our paper “Simple Cues Lead to a Strong Multi-Object Tracker” got accepted to CVPR 2023!
  • 2023/02/06: I joined NVIDIA for a 10 months intership working on LiDAR point clouds!
  • 2022/07/06: I attended the International Computer Vision Summer School in Sicily!
  • 2022/06/18: We gave our tutorial on Deep Visual Similarity Learning at CVPR 2022!
  • 2022/04/04: Our paper “GroupLoss++: A Deeper Look into Group Loss for Deep Metric Learning” got accepted to PAMI!
  • 2021/05/10: Our paper “Learning Intra-Batch Connections for Deep Metric Learning” got accepted to ICML 2021!