August 2014
Volume 14, Issue 10
Vision Sciences Society Annual Meeting Abstract  |   August 2014
Invariant representations of face identity in the ATL
Author Affiliations
  • Stefano Anzellotti
    Harvard University
  • Alfonso Caramazza
    Harvard University
Journal of Vision August 2014, Vol.14, 1451. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Stefano Anzellotti, Alfonso Caramazza; Invariant representations of face identity in the ATL. Journal of Vision 2014;14(10):1451.

      Download citation file:

      © ARVO (1962-2015); The Authors (2016-present)

  • Supplements

A large body of evidence has documented the involvement of occipitotemporal regions in face recognition. Neuropsychological studies found that damage to the anterior temporal lobes (ATL) can lead to face recognition deficits, and recent neuroimaging research has shown that the ATL contain regions that respond more strongly to faces than to other categories of objects. What are the different contributions of anterior temporal and occipitotemporal regions to face recognition? In a series of fMRI studies, we investigated representations of individual faces in the human ATL using computer generated face stimuli for which participants did not have individual-specific associated knowledge. Recognition of face identity from different viewpoints and from images of part of the face was tested, using an approach in which pattern classifiers are trained and tested on the responses to different stimuli depicting the same identities. The anterior temporal lobes were found to encode identity information about faces generalizing across changes in the stimuli. Invariant information about face identity was found to be lateralized to the right hemisphere. Some tolerance across image transformations was also detected in occipitotemporal regions, but it was limited to changes in viewpoint, suggesting a process of increasing generalization from posterior to anterior temporal areas. Consistent with this interpretation, information about identity-irrelevant details of the images was found to decline moving along the posterior-anterior axis, and was not detected in the ATL.

Meeting abstract presented at VSS 2014


This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.