Ambiguous Perceptions

There is a type of visual illusion that can be seen either of two ways – the vase or two profiles is one. (here) But it is very difficult to see it both ways at once. We seem to switch from one to the other. The two ways of seeing are in some sort of competition and there is always a winner and a loser but never a tie. Just thinking about one of the possible views is often enough to make a switch to that one.

I have taken this to be part of the workings of a MPOFBL system (massively parallel overlapping feed-back loops). It would stabilize on a ‘best fit scenario’ with a given set of inputs. But if the input is ambiguous, in this case designed intentionally to be as ambiguous as possible, there will not be a great deal of difference in the ‘fit’ of the two interpretations. The MPOFBL system can fall into either of the two scenarios with almost equal easy but cannot maintain both at the same time. So it flips between them.

There was a recent science daily report on work by Joel Pearson’s group at Vanderbilt, ‘Mind’s Eye’ Influences Visual Perception. They use a different method to force a choice between two perceptions.

“To test how imagery affects perception, Pearson, Tong and co-author Colin Clifford of the University of Sydney had subjects imagine simple patterns of vertical or horizontal stripes, which are strongly represented in the primary visual areas of the brain. They then presented a green horizontal grated pattern to one eye and a red vertical grated pattern to the other to induce what is called binocular rivalry. During binocular rivalry, an individual will often alternately perceive each stimulus, with the images appearing to switch back and forth before their eyes. The subjects generally reported they had seen the image they had been imagining, proving the researcher’s hypothesis that imagery would influence the binocular rivalry battle.”

It seems that in situations of ambiguity, we cannot maintain two differing perceptions, nor can we make one stable, and we flip between them. But adding some weight to one interpretation will make it the more stable one.

To me, probably because I cannot think of another mechanism, this switching between ambiguous perceptions points to an MPOFBL system.  

This entry was posted on 04/07/2008 at 12:27 pm and is filed under modeling. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response or trackback from your own site.