Contributed Symposium

The mental and neuronal representation of material properties

[Speaker] Gegenfurtner, Karl R:1
1:Giessen University (Germany)

The past 15 years have seen an increased interest in the perception of materials and their properties. I will present findings showing that the classification of images into different material categories is as fast as an analogous task with object classes. Using linear classifiers, material categories could be differentiated based on the ERP pattern in single trials as early as 100 ms after stimulus onset. In a series of experiments with sighted and blind observers, we found that there is a systematic perceptual representation of materials that transcends the visual and haptic modalities. Neuroimaging experiments showed that perceptual judgments on material properties coincided quite well with an analysis of low level image features and with the activity in early visual cortical areas. Overall, there has been a remarkable advance in our knowledge of the perception of materials and material categories, at the neuronal, perceptual and computational levels.
Advanced Search