August 2023
Volume 23, Issue 9
Open Access
Vision Sciences Society Annual Meeting Abstract  |   August 2023
Gestalt grouping vs. ensemble perception when following a crowd
Author Affiliations & Notes
  • William Warren
    Brown University, Providence, RI USA
  • Meghan Willcoxon
    Brown University, Providence, RI USA
  • Footnotes
    Acknowledgements  Supported by NIH R01 EY029745
Journal of Vision August 2023, Vol.23, 5795. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      William Warren, Meghan Willcoxon; Gestalt grouping vs. ensemble perception when following a crowd. Journal of Vision 2023;23(9):5795.

      Download citation file:

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

  • Supplements

Most models of collective motion in animal groups and human crowds assume ensemble averaging of one’s neighbors (e.g. Rio, Dachner & Warren, PRSB 2018). Last year we reported that attending to a subset of neighbors did not spontaneously increase their influence when following a crowd, consistent with ensemble perception (V-VSS 2022). Here we ask whether the Gestalt grouping principles of common fate and proximity spontaneously influence following behavior, as previously observed in multiple object tracking (Yantis, 1992; Erlikhman, et al., 2013) and fish schooling (Lemasson, et al. 2018). Participants were instructed to “walk with” a virtual crowd of 9 avatars presented in a mobile HMD, and the time series of heading was recorded. During each trial, a subset of avatars (targets) was perturbed while the remainder (distractors) continued walking straight ahead. Experiment 1 tested grouping by common fate. In the Test condition, 6 randomly positioned targets simultaneously changed speed (-0.2, 0, +0.2 m/s) and turned (±20˚), while 3 distractors changed speed and walked straight; in the Control condition, all 9 avatars were perturbed. Final heading was greater with 9 than 6 perturbed avatars (p<0.05), and slightly larger when Test targets slowed down than sped up (BF10=1.76, anecdotal), as predicted by our weighted averaging model. There was thus no effect of common fate. Experiment 2 tested grouping by proximity. In the Test condition, targets (N=2, 4, or 5) were spatially clustered and turned (±20˚), while the distractors walked straight ahead; in the Control condition, the targets were randomly positioned amid the distractors. There were no differences between the Test condition and the Control condition (BF01 > 6.3, substantial-strong), and thus no effect of proximity. We conclude that walking with a crowd is normally characterized by ensemble perception of neighbors, with no influence of grouping by common fate or proximity.


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.