• Top page
  • Timetable
  • Per session
  • Per presentation
  • How to
  • Meeting Planner



Visual System

開催日 2014/9/12
時間 14:00 - 15:00
会場 Poster / Exhibition(Event Hall B)

Neural response characteristics that are invariant to face inversion in monkey area TE

  • P2-146
  • 菅生-宮本 康子 / Yasuko Sugase-Miyamoto:1 松本 有央 / Narihisa Matsumoto:1 河野 憲二 / Kenji Kawano:2 
  • 1:産総研・ヒューマンライフ / Human Tech Res Inst, AIST, Ibaraki, Japan 2:京都大院・医・認知行動脳科学 / Dept Integrative Brain Sci, Kyoto Univ, Kyoto, Japan 

We have reported that face-responsive neurons in monkey area TE represent information about a global category of faces, human vs. monkey vs. simple shapes, earlier in their firing rates than information about fine categories, facial expression and identity. We recently found that when faces were presented with inversion, information about the fine categories decreased, whereas information about the global category remained. We studied whether the responses to the inverted faces were similar to those for the upright faces on a face-by-face basis. We analyzed activities of 119 face-responsive neurons in TE of two rhesus monkeys (Macaca mulatta), performing a fixation task. Test stimuli were colored pictures of monkey faces (4 models with 4 expressions), human faces (3 models with 4 expressions), 8 geometric shapes, and inverted picture of the faces. Each was presented for 400 ms. Population activity vectors consisting of responses to upright and inverted faces were computed in a window 115-165 ms after stimulus onset; this is the time that the global categorization is observed. The Euclidean distances between 28 upright faces and 28 inverted faces were smaller than those of the 192 combinations of 16 monkey faces and 12 human faces (Wilcoxon test, p < 0.0001), showing that inverted monkey/human faces evoked population responses similar to those evoked by upright monkey/human faces, respectively. Principal component analysis was applied individually to the upright monkey vs. human faces, to the upright vs. inverted human faces, and to the upright vs. inverted monkey faces. The distribution of eigenvector values of the first principal component showed that members of the neuronal population gave different contributions to the global categorization and to the upright vs. inverted categorization, but they contributed in a similar manner to both the upright vs. inverted categorization of the monkey faces, and to the upright vs. inverted categorization of the human faces. Thus, the face inversion was coded in a different aspect of the neuronal response than that which participates in the global categorization.

Copyright © Neuroscience2014. All Right Reserved.