About me

I am a Lead Research Scientist at Bosch Corporate Research, specializing in robot learning—the intersection of robotics and machine learning. My research focuses on robot skill learning for dexterous manipulation, with broader interests in sequential decision making (reinforcement learning, imitation learning, optimal control), geometric learning and control, and computer vision for object understanding.

I spend my free time with my family and astronomy, check out my astrobin page.

Publications

For the full puclication list check my google scholar page.

Some notable publications:

• Kupcsik A., Deisenroth M., Neumann G., Peters J., Data-efficient generalization of robot skills with contextual policy search, AAAI, 2013
• Kupcsik A., Lee W.S., Hsu D., Learning dynamic robot-to-human object handover from human feedback, Robotics Research, 2018
• Kupcsik A., Spies M., Klein A., Todescato M., Waniek N., Schillinger P., Bürger M., Supervised training of dense object nets using optimal descriptors for industrial robotic applications, AAAI, 2021

Previously

• I’m with Bosch since 2018
• In 2016 I joined Sylvain Calinon’s lab at Idiap research institute in Switzerland to work on the DexRov project
• In 2014 I joined the lab of David Hsu at NUS and worked on human-robot interaction
• I did my PhD in robot skill learning at the National University of Singapore from 2010-2014. During my studies I spent 8 month as a PhD sabbatical at TU Darmstadt, with Jan Peters’ lab