August 2023
Volume 23, Issue 9
Open Access
Vision Sciences Society Annual Meeting Abstract  |   August 2023
Determinants of Canonical Forms in Memory Storage and Object Recognition
Author Affiliations
  • Dongcheng He
    University of Denver
  • Haluk Ogmen
    University of Denver
Journal of Vision August 2023, Vol.23, 4695. doi:https://doi.org/10.1167/jov.23.9.4695
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Dongcheng He, Haluk Ogmen; Determinants of Canonical Forms in Memory Storage and Object Recognition. Journal of Vision 2023;23(9):4695. https://doi.org/10.1167/jov.23.9.4695.

      Download citation file:


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

      ×
  • Supplements
Abstract

Theories of invariant object recognition include those based on invariant features, on view-variant storage with interpolation/generalization mechanisms, as well as on canonical representations with active mental transformation operations. One crucial evidence for canonical storage theories is the association between object recognition and mental rotation, reflected by the incremental reaction times during object recognition with respect to orientation. However, how a “good canonical form” is determined for a given novel stimulus remains largely unknown. The goal of this study was to examine systematically the roles of (i) elongation and symmetry (two ubiquitous aspects of natural stimuli), as well as their combination, and (ii) boundaries and surface-texture (two fundamental aspects of natural objects) in the choice of canonical forms, while eliminating extraneous environmental (by using a VR headset that displays the stimulus without any other visible structures that could serve as reference-frames) and pre-learned (by using reinforcement learning) reference-frames. We used algorithmically generated textures with different orientations having specified levels of symmetry and aspect ratios as the stimuli. In each trial, we presented only one stimulus and asked observers (N=5) to judge handedness. Feedback was provided at the end of each trial. We hypothesized that canonical forms were used and could be triggered by symmetry and aspect ratio if the reaction times were linearly dependent on the angular difference between the observed orientation and either the most symmetrical or the most elongated orientation. Our results are consistent with this canonical representation hypothesis. Boundary information was more efficient than texture information, but this difference was not statistically significant. We found that subjects used the most symmetrical or elongated orientation as the canonical representation when only one factor was present, and they used a “winner-take-all” strategy when both factors were present, with elongation being more dominant than symmetry.

×
×

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.

×