next up previous contents
Next: 4.1 Training inputs Up: Tilt Aftereffects in a Previous: 3.8 Conclusion

4 Training the Orientation Map

  As discussed in chapter 2, neurons in the primary visual cortex are selective for the orientation and position of stimuli. Furthermore, the neurons are arranged retinotopically across the cortex, forming an orientation map (Blasdel and Salama, 1986; Blasdel, 1992a; Grinvald et al., 1994; Ts'o et al., 1990; Weliky and Kandler, 1995). Each local area in the map contains neurons with each possible orientation preference, all responding to the same location on the retina. The neurons tuned to a particular orientation are activated when an input of that orientation is present at their preferred location on the retina.

It has been shown previously that, given oriented inputs, the RF-LISSOM model develops orientation maps similar to those seen in the primary visual cortex (Miikkulainen et al. 1997; Sirosh and Miikkulainen 1997; Sirosh et al. 1996). In this thesis, the functional aspects of such self-organized orientation maps will be studied. This chapter describes the process of organizing and characterizing an orientation map, and chapter 5 will examine how the behavior of the map leads to tilt aftereffects.



 

James A. Bednar
9/19/1997