Fatemeh Ziaeetabar

Bernstein Center for Computational Neuroscience Göttingen, University of Göttingen, Amirkabir University of Technology

Papers

Total Citations

H-Index

About

Fatemeh Ziaeetabar is a leading researcher in cognitive robotics and human-robot interaction, with a core focus on enabling machines to understand, recognize, and predict human manipulation actions. Her work bridges computer vision, artificial intelligence, and cognitive science to create algorithms that can interpret the semantic meaning behind physical interactions with objects. Ziaeetabar’s major contributions include the development of **Enriched Semantic Event Chains (ESEC)** and **Semantic Spatial Reasoning** frameworks, which encode manipulation actions through a series of spatial relation changes between objects. These methods allow robots to not only recognize actions post-hoc but also **predict future actions** in real-time, a critical capability for fluent human-robot collaboration. Her most cited work, “Semantic analysis of manipulation actions using spatial relations” (28 citations), established the foundation for this approach, while her 2020 study “Humans Predict Action using Grammar-like Structures” (11 citations) demonstrated that human action prediction follows structured, grammar-like patterns, inspiring biologically plausible AI models. With over 75 total citations across her key publications, Ziaeetabar’s research has direct implications for assistive robotics, autonomous systems, and video understanding. Her latest work extends these ideas to generating multi-sentence descriptions of complex manipulation videos, pushing toward more natural and detailed human-robot communication.