June 2007
Volume 7, Issue 9
Free
Vision Sciences Society Annual Meeting Abstract  |   June 2007
Optimal searcher, saccadic targeting model, and human eye movements during search: Effects of target visibility maps
Author Affiliations
  • Wade Schoonveld
    University of California — Santa Barbara
  • Miguel Eckstein
    University of California — Santa Barbara
Journal of Vision June 2007, Vol.7, 635. doi:10.1167/7.9.635
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Wade Schoonveld, Miguel Eckstein; Optimal searcher, saccadic targeting model, and human eye movements during search: Effects of target visibility maps. Journal of Vision 2007;7(9):635. doi: 10.1167/7.9.635.

      Download citation file:


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

      ×
  • Supplements
Abstract

Many models of saccadic eye movements during search direct saccades toward the location that is most likely to contain the target (saccadic targeting models; Beutter et al., 2003; Torralba, 2003, JOSA-A). Alternatively, a recently proposed model of human eye movements during search, the optimal searcher uses knowledge of the differential detectability of the target across the retina (i.e., visibility map) to direct saccades to areas that maximize the probability of correctly localizing the target (Najemnik & Geisler, 2005, Nature). The optimal searcher occasionally executes “center-of-gravity” saccades between potential target locations. There has been little work comparing these models to human saccades across different target types. Here, we compare the ability of saccadic targeting and optimal search models to predict saccades during search for three different target types (a Gaussian blob and two Gabors, one with high and one with low contrast; spatial frequency=9.7 cycles/degree) that each had different associated visibility maps. Three observers each performed three search tasks localizing the high contrast target among four dimmer distractors within a 600 ms presentation time. The contrast of the target and distractors varied over time, every 25ms, by adding independent samples of random Gaussian contrast noise. Display elements were placed equidistant around the circumference of a circle with a radius of 5 degrees. The first two saccades were recorded from each trial and compared to the predictions made by the two models. “Center-of-gravity” saccades were predicted by the optimal search model for the Gaussian and high contrast Gabor, but are uncommon in the human data. The ideal saccadic targeting model accounts for human performance better than the optimal search model. For the low contrast Gabor, model predictions were essentially the same for both models with no “center-of-gravity” saccades and agreed with human behavior.

Schoonveld, W. Eckstein, M. (2007). Optimal searcher, saccadic targeting model, and human eye movements during search: Effects of target visibility maps [Abstract]. Journal of Vision, 7(9):635, 635a, http://journalofvision.org/7/9/635/, doi:10.1167/7.9.635. [CrossRef]
Footnotes
 Grant Support: NSF 0135118
×
×

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.

×