A Model of Position-Invariant, Optic Flow Pattern-Selective Cells
Two apparently inconsistent proposals for the functionality of MSTd cells exist, based either on optic flow patterns being represented by motion components 2–4, such as translation, expansion/contraction and rotation, or by a continuum of motion patterns that includes spirals6. A model, consisting of excitatory and inhibitory subunits, has been proposed3 to support the component view. Here, we used a neural network to show that a model of this type can be selective to the continuum of patterns6. We extended this model by adding inhibitory connections between units to account for the reported6 position-invariant characteristics of MSTd cells.
KeywordsReceptive Field Optic Flow Local Motion Prefer Pattern Spiral Pattern
Unable to display preview. Download preview PDF.
- Beardsley, S. A., Vaina, L. M. and Poggio, T. The development of optic flow selectivity in MSTd neurons using backpropagation networks. Soc. Neurosci. Abstr 22, 1619 (1996).Google Scholar
- Duffy, C. J. and Wurtz, R. H. Personal communication.Google Scholar
- Graziano, M. S. A., Andersen, R. A., and Snowden, R. J. Tuning of MST neurons to spiral motion. J. Neu osci. 14, 54–67 (1994).Google Scholar
- Pitts, R. I. and Vaina, L. M. A computational model of MSTd neurons sensitive to optic flow patterns. (Submitted).Google Scholar