Alina Kloss

Max Planck Institute for Intelligent Systems, University of Tübingen

Papers

Total Citations

116

H-Index

About

Alina Kloss is a researcher at the intersection of computer vision and robotics, whose work focuses on enabling robots to perceive and interact intelligently with their environments. Her primary research areas include visual attention for object search and the integration of learned and analytical models for physical reasoning. In her highly cited 2016 paper (56 citations), Kloss tackled the fundamental challenge of household robotics: searching for specific objects when they are not immediately visible. She developed a visual attention mechanism that allows robots to identify promising search locations, effectively teaching them "where to look" rather than relying on exhaustive scene scanning. Her 2017 work (55 citations) addresses another core robotic skill—predicting the outcomes of physical actions. Kloss pioneered a hybrid approach that combines the interpretability of physics-based analytical models with the flexibility of learned models, creating more accurate and robust predictions for action effects. This work has significant implications for manipulation and grasping tasks. Additionally, Kloss contributed to the robotics community through her participation in the SICK Robot Day 2014 competition, where she demonstrated the importance of system robustness in real-world, time-constrained scenarios. Her research continues to bridge the gap between controlled laboratory settings and practical, deployable robotic systems.