September 2024
Volume 24, Issue 10
Open Access
Vision Sciences Society Annual Meeting Abstract  |   September 2024
Temporal characteristics of perceived motion flow of naturalistic movies
Author Affiliations & Notes
  • Yung-Hao Yang
    Cognitive Informatics Lab, Graduate School of Informatics, Kyoto University
  • Taiki Fukiage
    NTT Communication Science Labs, Nippon Telegraph and Telephone Corporation
  • Zitang Sun
    Cognitive Informatics Lab, Graduate School of Informatics, Kyoto University
  • Shin’ya Nishida
    Cognitive Informatics Lab, Graduate School of Informatics, Kyoto University
  • Footnotes
    Acknowledgements  This work has been supported by JSPS Kakenhi 20H00603 and JP20H05957.
Journal of Vision September 2024, Vol.24, 724. doi:https://doi.org/10.1167/jov.24.10.724
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Yung-Hao Yang, Taiki Fukiage, Zitang Sun, Shin’ya Nishida; Temporal characteristics of perceived motion flow of naturalistic movies. Journal of Vision 2024;24(10):724. https://doi.org/10.1167/jov.24.10.724.

      Download citation file:


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

      ×
  • Supplements
Abstract

Visual motion perception involves the computational processing of signals evolving in space and time. Prior research using simple artificial stimuli suggested that human motion processing has sluggish temporal characteristics, integrating instantaneous motion over several tens of milliseconds or longer. Nevertheless, the dynamics of perceived motion in naturalistic scenes remain unexplored. Drawing on a recently proposed method (Yang et al., iScience, 2023), we explored temporal characteristics of the perceived optical flow in naturalistic scenes. Five movie clips sourced from a slow-flow version of the MPI Sintel Dataset, each featuring a large transient change in the middle of the clip, were presented at 60FPS within a circular aperture. Amidst the movie presentation, a tiny dot was momentarily flashed at the center of the aperture to indicate the spatiotemporal location of the target that the observers had to report by matching the speed and direction with a subsequently presented brown-noise field. Critically, to assess the temporal dynamics of perceived flow, the probed dot flashed -66.7ms, -33.3ms, 0ms, +33.3ms, or +66.7ms from the physical transient change. To control stimuli bias, we played each clip in eight ways: forward or backward in time, and four spatial flip conditions. The perceived vector, averaged across trials, shows gradual changes over time around the physically abrupt motion transition. This mainly reflects trial-by-trial temporal uncertainty of perceived direction transition. The temporal pattern of the averaged perceived vector can be well described by Gaussian temporal blurring of the ground-truth vector sequence, with [center, FWHM (Full width at half maximum) of Gaussian (ms)] = [+3.5, 119.7] for group data, and [+12.7, 78.1], [-40.2, 151.5], [-17.7, 117.4], [+30.7, 84.4] for individual data. The results indicate that the temporal window of motion processing, within which perceived motion cannot be accurately aligned in time with physical motion, is approximately 100 ms for naturalistic movies.

×
×

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.

×