Biography

I am a Research Associate at the Max Planck Institute for Intelligent Systems in Tübingen, working at the frontier of generative AI in Michael J. Black's group. My work is multimodal by nature, spanning video diffusion models and vision-language models applied to digital humans. I first-authored InterDyn (CVPR 2025), which showed that video diffusion models can generate physically plausible interactions from a single image and a driving motion without an explicit simulator, and co-authored a second CVPR 2025 paper on reconstructing 3D faces from heavily occluded images. Underneath the research I am hands-on in PyTorch, training models on multi-GPU A100/H100 clusters and building the evaluation tooling to measure them.

Before joining Max Planck, I completed my MSc in Artificial Intelligence at the University of Amsterdam and a BSc in Technology, Liberal Arts & Sciences at the University of Twente. Along the way, I worked on legal tech at Loyens & Loeff, interned at Perron038, where I built a thermal-imaging system to detect leaks inside sealed detergent-pod boxes, and at Tembo Group, where I mapped the quality trade-offs of federated learning.