September 2019
Volume 19, Issue 10
Open Access
Vision Sciences Society Annual Meeting Abstract  |   September 2019
High-fidelity visual features form complex objects in memory
Author Affiliations & Notes
  • Aedan Y Li
    Department of Psychology, University of Toronto
  • Keisuke Fukuda
    University of Toronto Mississauga
  • Morgan D Barense
    Department of Psychology, University of Toronto
    Rotman Research Institute, Baycrest
Journal of Vision September 2019, Vol.19, 76b. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Aedan Y Li, Keisuke Fukuda, Morgan D Barense; High-fidelity visual features form complex objects in memory. Journal of Vision 2019;19(10):76b.

      Download citation file:

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

  • Supplements

Though neuroanatomical and computational models predict feature integration, behavioral evidence for object representations in memory have been mixed. Here, we introduce a novel “conjunction” task to simultaneously measure the fidelity of shape and color memory, building on a recently described perceptually uniform shape space (VCS space: available on the Open Science Framework at In a first experiment, we extend a classic object-based memory study by Luck and Vogel (1997) using our conjunction task. Memory fidelity was reduced only slightly as the number of features within an object increased. In contrast, memory fidelity was drastically reduced when the number of objects increased, regardless of the number of features within each object. These results suggest that visual representations can be both object- and feature-based, though features tend to be stored as bound objects. In a second experiment, an interference paradigm examined whether object-based memory was driven by direct integration between shape and color, or driven by indirect integration through spatial location. This second experiment found an interaction between spatial location and the type of interference on memory fidelity, suggesting that shapes and colors can be bound together as well as bound to spatial location. Across two experiments, we show that although objects rather than features are preferentially stored in memory, trial-by-trial memory can flexibly contain both object- and feature-based representations. Moreover, we establish at least two levels of integration for complex objects: direct shape and color binding as well as indirect binding through shared spatial location.


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.