作者: J. Kubilius , J. Wagemans , H. P. Op de Beeck
DOI: 10.1167/14.9.11
关键词: Mathematics 、 Multivoxel pattern analysis 、 Functional magnetic resonance imaging 、 Visual processing 、 Artificial intelligence 、 Visual information processing 、 Pattern recognition 、 Visual memory 、 Signal strength 、 Neuroscience 、 Human visual system model 、 Stimulus (physiology)
摘要: The visual system is very efficient in encoding stimulus properties by utilizing available regularities the inputs. To explore underlying strategies during information processing, we presented participants with two-line configurations that varied amount of configural regularity (or degrees freedom relative positioning two lines) a fMRI experiment. Configural ranged from generic configuration to stimuli resembling an "L" (i.e., right-angle L-junction), "T" midpoint T-junction), or "+",-the latter being most regular stimulus. We found response strength shape-selective lateral occipital area was consistently lower for higher degree stimuli. In second experiment, using multivoxel pattern analysis, further show encoded terms signal but not distributed responses. Finally, results these experiments could be accounted low-level and are distinct norm-based encoding. Our suggest plays important role ventral processing stream.