Franziska Herbert

I’m a 2nd year PhD student at the Interactive Learning and Perception Lab (PEARL) lab at TU Darmstadt supervised by Professor Georgia Chalvatzaki. Before starting my PhD, I obtained M.Sc. Degrees in Autonomous Systems and Computer Science from TU Darmstadt, where I wrote my master thesis on Learning Semantic-Geometric Task Graphs from Bimanual Human Demonstrations.

I work at the intersection of robotics, machine learning, and human-robot interaction. My research focuses on enabling robots to learn meaningful structured representations (e.g. graphs) from human demonstrations. I’m particularly interested in extracting humans’ natural ability to interact with their environment and learn novel concepts easily, and leveraging these abilities to make robot learning more efficient. This includes exploring how humans and robots can collaborate effectively and developing interactive AI algorithms that enable this interaction.

news

May 20, 2026	Our paper Semantic-Geometric Task Representations for Bimanual Manipulation from Human Demonstrations to Robot Action Planning has been accepted to IEEE Robotics and Automation Letters (RA-L).

selected publications

Semantic-Geometric Task Representations for Bimanual Manipulation from Human Demonstrations to Robot Action Planning

Franziska Herbert, Vignesh Prasad, Han Liu, Dorothea Koert, and Georgia Chalvatzaki

IEEE Robotics and Automation Letters (RA-L), 2026

Abs PDF Website

Learning structured task representations from human demonstrations is essential for understanding long-horizon manipulation behaviors, particularly in bimanual settings where action ordering, object involvement, and interaction geometry can vary significantly. A key challenge lies in jointly capturing the discrete semantic structure of tasks and the temporal evolution of object-centric geometric relations in a form that supports reasoning over task progression. In this work, we introduce a semantic-geometric task graph-representation that encodes object identities, inter-object relations, and their temporal geometric evolution from human demonstrations. Building on this formulation, we propose a learning framework that combines a Message Passing Neural Network (MPNN) encoder with a Transformer-based decoder, decoupling scene representation learning from action-conditioned reasoning about task progression. The encoder operates solely on temporal scene graphs to learn structured representations, while the decoder conditions on action-context to predict future action sequences, associated objects, and object motions over extended time horizons. Through extensive evaluation on human demonstration datasets, we show that semantic-geometric task graph-representations are particularly beneficial for tasks with high action and object variability, where simpler sequence-based models struggle to capture task progression. Finally, we demonstrate that task graph representations can be transferred to a physical bimanual robot and used for online action selection, highlighting their potential as reusable task abstractions for downstream decision-making in manipulation systems.