Martina Gassen
Papers
1
Total Citations
2
H-Index
1
About
Martina Gassen is a rising researcher in computer vision and human-robot interaction, with a focus on few-shot action segmentation and interactive learning from demonstration. Her most-cited work, "I³: Interactive Iterative Improvement for Few-Shot Action Segmentation" (2023), tackles the critical challenge of extracting modular action segments from raw video demonstrations with minimal labeled data. Gassen's key contribution is a novel interactive framework that iteratively refines segmentation using human feedback, enabling robots to understand high-level task structures from just a few examples—a breakthrough for data-efficient learning in robotics. This approach directly addresses the limitations of supervised methods that require extensive annotation, making it practical for real-world human-robot collaboration. With 2 citations to date, her work is gaining traction in the few-shot learning community. Gassen's research sits at the intersection of computer vision, interactive machine learning, and robotics, promising to make robot learning more accessible and adaptable. Her innovative use of iterative human-in-the-loop refinement marks her as a promising young scientist shaping the future of intuitive robot teaching.
Research Focus
Key Achievements
Top Papers
- 1I³: Interactive Iterative Improvement for Few-Shot Action Segmentation2 citations · 2023