E. Chinellato, Y. Demiris (UK), and A.P. del Pobil (Spain)
Although other primates have grasping skills, human be ings evolved theirs to the extent that a large fraction of our brain is involved in grasping actions. Recent neuroscience findings allow us to depict the outline of a model of vision based grasp planning that differentiates from the previous ones in that it is the first to rest mainly, if not exclusively, on human physiology. The main theory on which our proposal is based is that of the two streams of the human visual cor tex [1]. Although they are evolved for different purposes, being the ventral stream dedicated to perceptual vision, and the dorsal stream to action-oriented vision, they need to collaborate in order to allow proper interaction of human beings with the world. Our framework has been conceived to be applied on a robotic setup, and the design of the dif ferent brain areas has been performed taking into account not only biological plausibility, but also practical issues re lated to engineering constraints. Connectionist Models, Cognitive Processes, Computa tional Neuroscience, Robotic Grasping, Visual Streams
Important Links:
Go Back
