MIT Finds Clues to Visual Recognition

Posted 22 Aug 2005 at 20:17 UTC by steve Share This

How does the brain so easily recognize familiar objects when the actual images registered in the eye can change up to 200 times per second? An MIT press release describes one way in which the human brain achieves this feat of recognition. The researchers hypothesized that the hundreds of tiny movements made by the eye each second are somehow involved in the recognition process. By creating a virtual world for test subjects to look at, the researchers were able to make changes to the visual field during the brief intervals when the eyes are in motion. By changing the position of objects during these periods of blindness, they were able to create confusion in the subjects about where objects were, even though the subjects reported not being aware of the changes. The discovery could have important implications for machine vision. The research team was led by James DiCarlo of the DiCarlo Lab at MIT's McGovern Institute for Brain Research. This sounds very reminisent of saccadic eye motion studies in which changes to text could be made during eye saccades as the reader was reading but without them being aware of the changes. For previous related research see Susan Blackmore's paper on trans-saccadic memory.

See more of the latest robot news!

Recent blogs

30 Sep 2017 evilrobots (Observer)
10 Jun 2017 wedesoft (Master)
9 Jun 2017 mwaibel (Master)
25 May 2017 AI4U (Observer)
25 Feb 2017 steve (Master)
16 Aug 2016 Flanneltron (Journeyer)
27 Jun 2016 Petar.Kormushev (Master)
2 May 2016 motters (Master)
10 Sep 2015 svo (Master)
14 Nov 2014 Sergey Popov (Apprentice)
Share this page