Neural network models for
color and brightness perception
and
binocular rivalry

Ph.D. Dissertation, Boston University, 1993
University Microfilms International, Ann Arbor, Michigan

Karl Frederick Arrington

Advisor: Stephen Grossberg

The thesis is on neural network models of the early visual system. It describes research in four separate areas.

1. A study of the temporal dynamics of brightness filling-in under visual masking conditions.
See: Arrington, K. F. (1994). The temporal dynamics of brightness filling-in. Vision Research, 34(24), p. 3371-3387.

2. A theoretical development of directional filling-in theory with simulation experiments.
See: Arrington, K. F. (1996). Directional filling-in. Neural Computation, 8(2), p. 300-318.

3. A model of binocular rivalry using nonlinear feedback and slow habituating chemical transmitters.
This chapter has yet to be submitted for publication elsewhere, but the chapter may be viewed as: Ch. 6

4. A model for preserving invariant contrast-ratio information across hierarchical processing layers.
The research presented in this chapter was done in collaboration with Stephen Grossberg.It has not been published elsewhere.


Return to CV.