Re: [OM] topaz sharpen AI

[Mike Gordon via olympus]

Is there a link to Ctein's comments? 

 Intriguing software that might be just the ticket in some situations.  Back in 
early 90's when back propagation neural networks were relatively new, had used 
them
to model some biological systems. (had to purchase a special board  optimized 
for back-propagation neural networks or new 386 would have to run forever)  It 
works well but the mechanism is always rather mysterious.  The training set 
must have images similar to the one being worked on or the results can be 
really off.
There can be really odd results as it creates pixels out of thin air when the 
training set is not right. 
One might predict a quite fuzzy image might be saved by it where nothing else 
would be suitable.  One would not expect ringing artifacts as when employing 
overly aggressive R--L deconvolution.
When it is better and how it can be used in conjunction with FM requires 
experimentation.  I hope Moose provides an update.

Intrigued Mike
-- 
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/