Papers
12
Total Citations
167
H-Index
8
About
Pedro Vicente is a robotics researcher specializing in autonomous manipulation, learning from demonstration, and adaptive body perception for humanoid robots. His work focuses on enabling robots to perform complex household tasks, such as table-cleaning and grasping, through deep learning and kinesthetic teaching. Vicente’s major contributions include developing methods for online body schema adaptation—where robots use internal mental simulation combined with multisensory feedback to continuously update their own kinematic models—and markerless visual servoing for hand-eye coordination during reaching and grasping. His papers, including the highly cited “iCub, clean the table!” (29 citations), demonstrate practical applications of deep neural networks for service robotics. Vicente also explores advanced topics like haptic exploration for safe 3D grasping using Bayesian optimization and reinforcement learning for dry-stacking irregular rocks, a task with potential extraterrestrial applications. His work on GPU-enabled particle optimization for hand pose estimation and self-calibration further highlights his impact in real-time robotic perception. With over a decade of research, Vicente’s contributions are foundational for developing autonomous robots that can adapt and learn in unstructured environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4Towards markerless visual servoing of grasping tasks for humanoid robots21 citations · 2017
- 5
- 6Incremental adaptation of a robot body schema based on touch events14 citations · 2018
- 7
- 8
- 9
- 10Eye-hand online adaptation during reaching tasks in a humanoid robot5 citations · 2014