NOTE TIME CHANGE: The Contribution of Top-Down Predictions to Visual Recognition
Speaker: Moshe Bar , Martinos Center for Biomedical Imaging, MGH, Harvard Medical School
Date: July 20 2006
Time: 2:00PM to 3:00PM
Location: 32-D507
Host: Polina Golland, CSAIL
Contact: Polina Golland, x8005, polina@csail.mit.edu
Relevant URL: Cortical analysis related to object recognition is traditionally thought to propagate
serially along a bottom-up hierarchy of visual areas. Recent proposals gradually
promote the role of top-down processing in recognition, but how such facilitation is
triggered remains a puzzle. We tested a specific model (Bar, 2003), proposing that
visual object recognition is facilitated by top-down processes originating in the
prefrontal cortex. The gist of this proposal is that a partially analyzed version of
the input image, comprised of the low spatial frequency components (i.e., a blurred
image), is projected rapidly from early visual areas directly to the prefrontal
cortex, possibly using the dorsal magnocellular pathway. This coarse representation is
subsequently used to activate predictions about the most likely interpretations of the
input image in recognition-related regions within the temporal cortex. Combining this
top-down initial guess with the bottom-up systematic analysis facilitates
recognition by substantially limiting the number of object representations that need
to be considered. The present study combined magnetoencephalography (MEG), which has
superior temporal resolution, functional magnetic resonance imaging (fMRI), and a
behavioral task that gradually yields successful recognition with repetitions. Object
recognition elicited differential activity that developed in the left orbital
prefrontal cortex 50 ms earlier than it did in recognition-related areas in temporal
cortex. In addition, this early activity was directly modulated by the presence of low
spatial frequencies in the image. We propose that this early orbitofrontal activity
reflects the origin of top-down facilitation in object recognition, and further
present the dynamics and functional connectivity that mediate successful recognition.
I will outline the logic and discuss behavioral and neuroimaging data that support
various aspects of the proposed model.
http://barlab.mgh.harvard.edu
Supported by NINDS RO1 Grants NS50615 and NS44319, the M.I.N.D. Institute, and NIH
National Center for Research Resources Grant 5P41RR014075
See other events that are part of Biomedical Imaging and Analysis 2005/2006
See other events happening in July 2006
