September 2021
Volume 21, Issue 9
Open Access
Vision Sciences Society Annual Meeting Abstract  |   September 2021
Image- and Task-Based Contributions to Human Object Localization in Natural Scenes
Author Affiliations & Notes
  • Colin S. Flowers
    University of Arizona
    University of Minnesota
  • Mary A. Peterson
    University of Arizona
  • Footnotes
    Acknowledgements  ONR N00014-16-1-2127
Journal of Vision September 2021, Vol.21, 2851. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Colin S. Flowers, Mary A. Peterson; Image- and Task-Based Contributions to Human Object Localization in Natural Scenes. Journal of Vision 2021;21(9):2851.

      Download citation file:

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

  • Supplements

To investigate object localization in natural scenes, we presented a flickering dot probe in 384 color photographs (100 ms; masked). Participants reported whether the probe was “on” or “off” the object bounded by the nearest border. Object localization sensitivity (d') was > 1.0 (VSS 2019). Now, we compared performance with these masked photographs under three conditions. Participants (1) made a localization judgment only (N = 39), (2) also made an object categorization judgment (N = 48), or (3) made a localization judgment on photographs preceded by one of three intermixed primes: a neutral letter string or a word denoting either the object near the probe or a different object (N = 122; ~40/prime condition). Photograph-based analyses of d' and criterion (c) were conducted for each task. Sensitivity was higher when participants made localization responses only (d' = 1.26) than when they also made categorization judgments or when any prime preceded the photographs, [F(4,1532) = 12.385, p <0.001]; d's in the other conditions did not differ (mean = 1.08). Criterion was biased significantly towards ‘on’ when the prime denoted the nearest object and in condition 2 compared to the other conditions, [F(4,1532) = 66.554, p <0.001]. Thus, task factors affect object localization sensitivity and criterion assessed with briefly exposed photographs of natural scenes. Next, local complexity was indexed as the percentage of pixels on borders within 2° of the border nearest the probe (using canny filter output). Complexity never correlated with d', but in condition (1) and all priming conditions, it correlated significantly with criterion (r’s > 0.165; p’s <0.01): As complexity increased, the bias to respond ‘on’ increased. We consider how attention, object borders, and distinct task requirements contribute to these effects of task and complexity on both object localization sensitivity and participants’ criterion.


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.