Steve Hovland shovland at mindspring.com
Mon Dec 6 05:24:31 UTC 2004

1. Introduction
1.1 Traditional Methods of Visual Image Processing. Shortcomings.
1.2 Neuromorphic Methods. Advantages.
1.3 Visual Image segmentation Task: Posing of the Problem.
2. The Brain Visual System: Brief Neurophysiological Data.
2.1 The Visual Pathway: Retina, LGN, the Visual Cortex (VC).
2.2 Main Propetrties of Primary Visual Cortex: RFs, Columns, Connections.
2.3 The Role of Synchronized Oscillations in VC and in the Other Brain 
Structures. Experiments. Associative Binding.
2.4 Some Other Features of the Brain Visual System.
3. Models of VC
3.1 Different Levels of Modelling. Neural Oscillator.
3.2 Models, Based on Networks of Spiking Neurons.
3.3 Oscillatory Network Models.
3.4 The Model under Development: Tunable Oscillatory Network with 
Self-organized Dynamical Connections and Synchronization-Based Performance.
4. Method of Visual Image Segmentation Based on Cluster Synchronization of 
Oscillatory Network.
4.1 Reduced Network, Obtained from the Oscillatory Model of VC.
4.2 Method of Adaptation of Connections. Successive Cluster Synchronization 
of the Network.
4.3 Computer Realization of the Method. Experiments.
4.4 Advantages of the Algorithm. Futher Perspectives.
5. Concluding Remarks. Abilities of Neuromorphic Models.

More information about the paleopsych mailing list