December 2001
Volume 1, Issue 3
Free
Vision Sciences Society Annual Meeting Abstract  |   December 2001
Revisiting snodgrass and Vanderwart's object database: Color and texture improve object recognition
Author Affiliations
  • Bruno Rossion
    Brown University, Road Island, USA
  • Gilles Pourtois
    Tilburg University, Tiburg, The Netherlands
Journal of Vision December 2001, Vol.1, 413. doi:10.1167/1.3.413
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Bruno Rossion, Gilles Pourtois; Revisiting snodgrass and Vanderwart's object database: Color and texture improve object recognition. Journal of Vision 2001;1(3):413. doi: 10.1167/1.3.413.

      Download citation file:


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

      ×
  • Supplements
Abstract

The Snodgrass and Vanderwart's object database contains a set of 260 black and white pictures with normative data on variables relevant to visual, mnesic and cognitive processing. This database has been widely used for 20 years in a large amount of behavioral studies with normal subjects and patients, as well as in neuroimaging studies aimed at studying object recognition. However, the original set is only available as line drawing slides, without any texture and color information, despite growing evidence that these visual variables play an important role in object perception and recognition. Here we created 2 new computerized sets from the original database, in which texture information and internal details have been added (set 1, gray level), as well as color information (set 2). We collected normative data on 240 subjects for the original and the 2 new sets for name agreement, image agreement, familiarity and visual complexity, as in the original study. Comparisons of the 3 sets clearly show that the addition of texture significantly improves naming agreement (p/s<0.01) with further improvement for the colorized set (p/s<0.01 as compared to gray level). Mean response times for the correct object label are also significantly faster for colorized items (903 ms) than gray level items (963 ms; p/s<0.001) and than original line drawings (1002 ms; p/s<0.0001; gray vs. line drawing: p/s<0.0001). In addition, texture and color improve familiarity agreement, as shown by significantly lower variances between subject responses in familiarity judgements (gray vs. line drawings: p/s<0.0001; color vs. Gray: p/s<0.0001), while complexity judgements and imagery agreement norms remained stable among the three sets of pictures. The stimuli and the corresponding normative data provide therefore valuable materials for a wide range of behavioral studies and further illustrate the importance of texture and color information on object processing.

Rossion, B., Pourtois, G.(2001). Revisiting snodgrass and Vanderwart's object database: Color and texture improve object recognition [Abstract]. Journal of Vision, 1( 3): 413, 413a, http://journalofvision.org/1/3/413/, doi:10.1167/1.3.413. [CrossRef]
×
×

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.

×