Re: Perception as memory ...

[Kevin Austin <kevin.austin@xxxxxxxxxxxx>]

The memory part is remembering how to defocus the eyes.

This was brought up elsewhere in the question regarding  
transferability of the task of segregation in auditory (streaming)  
training.

In machine-modeling of perception and perceptual tasks, or perceptual  
development, are there any models that dispense with memory?


Best

Kevin



On 2009, Aug 24, at 3:49 PM, Eliot Handelman wrote:

> Kevin Austin wrote:
> > If 'perception' can be developed, I feel that this development is a  
> > matter of memory development, and the associated aspects of memory  
> > that relate to categorization.
> An interesting counterexample is perhaps the 3d cg stereograms  
> popular in the 90s. You have to "learn" to perceive these -- eg by  
> learning to relax your focus and trying other tricks. But once  
> you've figured it out,  your built-in stereoscopic processes kick in  
> and the rest is automatic. In what way does memory or categorization  
> figure in this learning? Or should this not be called perceptual  
> development?
>  -- eliot