For example, it is dim light and an object is in partial shade, hiding some features of an object. It is a rounded, dark brownish object. The object is rather ambiguous - it could be a rock, a potato, a turtle, a grenade, a ball, etc.

When looking at an ambiguous object like that described above, do different neural networks get activated with different intensity? Does the brain pick the highest probability and the person label the object with that name? For example, does the brain get input like:

  • 63% of potato detecting neurons get activated, saying it looks like a potato
  • 50% ball
  • 34% turtle
  • 28% grenade

    Or is it the feature recognition that helps distinguish objects? For example:

  • potato has skin texture and dimples

  • ball is perfectly round
  • turtle has irregular edge and rectangular pattern
  • grenade has a handle and pin

If an object has one of these features, it is unlikely to be another object?

enter image description here

enter image description here

  • 2
    $\begingroup$ How is feature recognition any different from voting of different networks? What do you think recognizes the features? Or do you literally believe that there is a "potato" and a "grenade" neural network in your brain that is somehow distinct from the "networks for other objects"? Finally, is it possible to remove the images? They don't add anything to the question except length. $\endgroup$ – Artem Kaznatcheev Mar 3 '14 at 17:36
  • 1
    $\begingroup$ To put it simply, our recognition of things is based on our past experiences. If something looks like something you've used a lot, your brain will likely recognize it as that item. If it's something vague, you will probably have to consciously think about what the item may be. $\endgroup$ – NoLongerContributingToSE Mar 6 '14 at 1:52
  • $\begingroup$ I voted to close as too broad. This is a technically good question, but the only reasonable answer is tantamount to a chapter or book-length review of perception. $\endgroup$ – Christian Hummeluhr Mar 28 '15 at 12:30
  • $\begingroup$ @ChristianHummeluhr I agree. If you'd like to whittle it down to focusing on the inferior temporal cortex which is where much of this processing is thought to go on, I think it would be answerable. $\endgroup$ – Chuck Sherrington Apr 5 '15 at 0:35

Browse other questions tagged or ask your own question.