Postdoctoral Research Associate in Dexterous Manipulation and Robot Learning

This postdoctoral position at Imperial College London’s DexLab, led by Dr. Rolandos Potamias and Stefanos Zafeiriou, focuses on pioneering dexterous manipulation and robot learning. The successful candidate will develop novel algorithms and policies for high-degree-of-freedom (high-DoF) robotic hands and arms, leveraging vision-language-action models (VLAs) and human demonstration data. The role integrates expertise in robot learning, computer vision, generative modeling, and control, with opportunities to deploy learned policies on state-of-the-art robotic platforms, including Sharpa Wave dexterous hands and UR7e arms. Collaboration with a multidisciplinary team and publication in top-tier conferences like CVPR, NeurIPS, and ICRA are expected.

• Conduct original research at the intersection of dexterous manipulation, robot learning, and foundation models, focusing on transforming large-scale human demonstration data into capable, general-purpose robot behavior.
• Develop novel algorithms and policies for high-degree-of-freedom (high-DoF) robotic hands and validate them on real hardware platforms, including Sharpa Wave dexterous hands and UR7e arms.
• Collaborate with a multidisciplinary team of experts in robotics, computer vision, and machine learning to advance state-of-the-art methodologies.
• Publish research findings in top-tier conferences such as CVPR, NeurIPS, ICLR, ECCV, ICCV, ICRA, and RSS.
• Deploy and test learned policies on real robotic systems, ensuring seamless integration of theoretical advancements with practical applications.
• Contribute to the design and implementation of vision-language-action models (VLAs) and other robot foundation models.