A Self-organizing Neural Network for Learning A Body-centered Invariant Representation of 3-D Target Position
This paper describes a self-organizing neural network that learns a body-centered representation of 3-D target positions. This representation remains invariant under head and eye movements, and is a key component of sensory-motor systems for producing motor equivalent reaches to targets . Learning requires no teacher, instead utilizing information gained from an action-perception cycle in which head movements are made while a stationary target is foveated. Because the spatial representations used relate closely to neck anatomy, the network learns very rapidly, converging after foveating only 200 targets.
Unable to display preview. Download preview PDF.
- Bullock, D., Grossberg, S., and Guenther, F. H. (1993). A self-organizing neural model of motor equivalent reaching and tool use by a multijoint arm. Journal of Cognitive Neuroscience. In press.Google Scholar
- Guenther, F. H., Bullock, D., Greve, D., and Grossberg, S. (1993). Neural representations for sensory-motor control, III: Learning a body-centered representation of 3-D target position. Technical Report, Boston University Center for Adaptive Systems.Google Scholar
- Vidal, P. P., de Waele, C., Graf, W., and Berthoz, A. (1988). Skeletal geometry underlying head movements. In Cohen, B. and Henn, V., (eds.): Representation of Three-Dimensional Space in the Vestibular, Oculomotor, and Visual Systems, pp. 228–238. New York: New York Academy of Sciences.Google Scholar