December 2022
Volume 22, Issue 14
Open Access
Vision Sciences Society Annual Meeting Abstract  |   December 2022
Tracking accuracy predicted using object-level properties
Author Affiliations & Notes
  • Jiří Lukavský
    Czech Academy of Sciences
  • Lauri Oksama
    Finnish Defence Research Agency, Human Perfomance Division
  • Filip Děchtěrenko
    Czech Academy of Sciences
  • Footnotes
    Acknowledgements  The research has been supported by Czech Science Foundation (GA19-07690S).
Journal of Vision December 2022, Vol.22, 3892. doi:https://doi.org/10.1167/jov.22.14.3892
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Jiří Lukavský, Lauri Oksama, Filip Děchtěrenko; Tracking accuracy predicted using object-level properties. Journal of Vision 2022;22(14):3892. https://doi.org/10.1167/jov.22.14.3892.

      Download citation file:


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

      ×
  • Supplements
Abstract

Multiple Object Tracking task is a popular method to investigate human attention. Many factors contribute to the successful tracking, and consequently, the difficulty varies for individual items. However, the tracking capacity is estimated from the trial's overall performance, and the information about individual items is neglected. Here, we reanalyzed the data from three MOT experiments featuring prolonged object occlusions (N=3x50, a total of 14450 trials). The boundaries of the tracking area were partly covered and both targets and distractors could disappear behind the cover for hundreds of milliseconds to several seconds. We designed an exponential decay model, where we model the tracking performance for each object based on its occlusion history (namely the total time visible vs occluded). In the model, we assume there is a fixed probability of successful tracking of a visible object (per time unit), and another probability of successful tracking for an occluded object. Thus, it is possible to test whether experimental manipulations generally affect the tracking performance or, specifically, the tracking of occluded objects. In the presented experiments, we had independently manipulated the attentional workload (2 to 4 targets) or the memory workload (MIT vs MOT). Both manipulations affected the probabilities of tracking when the object was occluded and yielded only negligible effects for the probability of tracking a visible object. These results suggest that the available resources do not enhance the tracking capacity directly (people could span more objects) but instead help overcome the critical situations in tracking (cover, crowding, collisions). Although the model is still very simple and there are many more parameters one can derive from the object history, the model is easily extensible to other types of events (e.g., number of collisions or number of occlusions).

×
×

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.

×