Miguel Castelo‐Branco

Association for Innovation and Biomedical Research on Light and Image, University of Coimbra

Papers

Total Citations

H-Index

About

Miguel Castelo‐Branco is a leading researcher at the intersection of robotics, computational neuroscience, and active perception. His work is fundamentally driven by understanding how biological systems, particularly the human brain, process sensory information to guide action, and then translating those principles into artificial systems. His major contributions lie in developing Bayesian frameworks for active multimodal perception, inspired by the dorsal visual pathway's role in spatial awareness and motion processing. His most cited work, "A Bayesian framework for active artificial perception" (2012, 48 citations), establishes a computational architecture for robots to actively perceive 3-D structure and motion by integrating multiple sensory inputs, mirroring human egocentric spatial reasoning. This framework provides a principled approach for robots to decide where to look and how to move to gather the most informative data. Additionally, his research on "Integration of touch attention mechanisms to improve the robotic haptic exploration of surfaces" (2016) extends this active perception paradigm to the tactile domain, enabling robots to more efficiently and intelligently explore unknown surfaces. By bridging neuroscience and robotics, Castelo‐Branco’s work is paving the way for more autonomous, perceptually-aware machines that can interact with the world in a human-like manner.