October 2020
Volume 20, Issue 11
Open Access
Vision Sciences Society Annual Meeting Abstract  |   October 2020
Object recognition based on visual shape can facilitate material recognition
Author Affiliations & Notes
  • Eiji Kimura
    Chiba University
  • Footnotes
    Acknowledgements  Supported by JSPS KAKENHI (26285162 & 18K18686)
Journal of Vision October 2020, Vol.20, 960. doi:https://doi.org/10.1167/jov.20.11.960
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Eiji Kimura; Object recognition based on visual shape can facilitate material recognition. Journal of Vision 2020;20(11):960. https://doi.org/10.1167/jov.20.11.960.

      Download citation file:

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

  • Supplements

Some everyday objects are made of a particular material in most cases (e.g., cotton T-shirts). This high association between objects and material may mediate mutual interaction between object and material recognition. This study investigated using a priming paradigm whether object recognition based on visual shape can affect material recognition. In Experiment 1, the prime was a line drawing of an object without any color and texture information. Twenty prime objects were selected that are highly associated with one of five material categories (leather, metal, stone, wood, and cloth); e.g., a baseball glove (leather), wrench (metal), and gravestone (stone). The target was a close-up material image without any object information. The relationship between the prime and target was systematically manipulated in three conditions. In the congruent-typical condition, the target image had typical color and texture of the prime object. In the congruent-atypical condition, the material represented in the target image belonged to the same category as the one associated with the prime object but its surface properties were atypical. In the incongruent condition, the image was selected from one of the other four categories than the one associated with the prime. On experimental trials, a prime was presented for 200 msec and, after a blank of 500 msec, a target was presented for 60 msec. The target was followed by a mask of 300 msec. Observers were asked to categorize the target material in a 5AFC task. Results showed that material categorization was most accurate in the congruent-typical condition and significantly better than that in the incongruent condition. The performance in the congruent-atypical condition was intermediate. This facilitative effect disappeared, when an object name was used as the prime in Experiment 2. These findings suggested that, at least for some objects, material properties are visually, not conceptually, encoded in their representations.


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.