SPARC: Reliable Spatial Annotations from Robot Demonstrations at Scale
Turning robot demonstrations into reliable spatial supervision, with explicit control over annotation quality at scale.
Best Paper Award @ Lab2Production Workshop RSS 2026
I’m a PhD researcher at KIT’s Intuitive Robots Lab, advised by Rudolf Lioutikov.
My research focuses on foundation models for robotics, spatial grounding, and learning from robot demonstrations. I study how vision-language models can annotate and curate robot datasets and provide supervision for robot learning.
I received my master’s degree in Computer Science from KIT. During my studies, I interned at SAP SE and IONOS.
News
- SPARC accepted to CoRL 2026.
- SPARC received the Best Paper Award at the Lab2Production Workshop at RSS.
- Announcing our RSS 2026 workshop: Trustworthy Embodied Foundation Models: Safety by Design versus Safety by Practice
- SIR published at CVPR 2026.
- We are hosting the first Foundation Model Bootcamp at KIT.
Publications
SIR: Structured Image Representations for Explainable Robot Learning
Sparse, task-relevant scene graphs make robot policies interpretable and more robust to visual distractors.
BEAST: Efficient Tokenization of B-Splines Encoded Action Sequences for Imitation Learning
Compact B-spline action tokens enable fast parallel decoding and smooth robot trajectories, without separate tokenizer training.
Scaling Robot Policy Learning via Zero-Shot Labeling with Foundation Models
Automatically segmenting and labeling over 115,000 trajectories from 430+ hours of robot demonstrations with pretrained foundation models.