Abstract
The neural representation of objects in the ventral visual pathway has been linked to high-level properties of the stimulus, such as semantic or categorical information. However, the extent to which patterns of neural response in these regions reflect more basic underlying principles is unclear. One problem is that existing studies generally employ stimulus conditions chosen by the experimenter, potentially obscuring the contribution of more basic stimulus dimensions. To address this issue, we used a data-driven analysis to describe a large database of objects in terms of their visual properties (spatial frequency, orientation, location). Clustering algorithms were then used to select images from distinct regions of this feature space. Images in each cluster did not clearly correspond to typical object categories. Nevertheless, they elicited distinct patterns of response in the ventral stream. Moreover, the similarity of the neural response across different clusters could be predicted by the similarity in image properties, but not by the similarity in semantic properties. These findings provide an image-based explanation for the emergence of higher-level representations of objects in the ventral visual pathway.
Meeting abstract presented at VSS 2017