December 2022
Volume 22, Issue 14
Open Access
Vision Sciences Society Annual Meeting Abstract  |   December 2022
An active naturalistic navigation task induces large attentional shifts in semantic representation
Author Affiliations
  • Tianjiao Zhang
    UC Berkeley
  • Jack Gallant
    UC Berkeley
Journal of Vision December 2022, Vol.22, 4475. doi:https://doi.org/10.1167/jov.22.14.4475
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Tianjiao Zhang, Jack Gallant; An active naturalistic navigation task induces large attentional shifts in semantic representation. Journal of Vision 2022;22(14):4475. https://doi.org/10.1167/jov.22.14.4475.

      Download citation file:


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

      ×
  • Supplements
Abstract

One important effect of visual selective attention is to shift perceptual representations toward attended semantic categories, increasing representation of attended categories at the cost of representation of other unattended categories (Çukur et al., 2013). However, prior work on this topic has only examined attentional tuning shifts under passive viewing conditions. Here we sought to determine whether similar shifts occur during active, naturalistic tasks. We used fMRI to record brain activity in three subjects while they performed an active navigation task in a large virtual world (110-180 minutes of data per subject). Data from the video game engine was used retrospectively to semantically segment the navigation video. Voxelwise encoding models were then estimated for each subject separately by using banded ridge regression to find a set of semantic weights that optimally predicted brain activity. A held-out dataset was used to test statistical significance, prediction accuracy and generalization. For comparison we also used fMRI to record brain activity while subjects passively watched naturalistic videos (210 minutes per subject; Huth et al., 2012). Videos were semantically labeled using WordNet. Voxelwise encoding models were estimated using the same procedure as in the active navigation task. Attentional shifts in semantic tuning between the active and passive tasks were identified by comparing the semantic maps obtained in the two tasks. This comparison shows that active navigation elicits semantic tuning shifts broadly across the cerebral cortex. Strong attentional tuning shifts in semantic representation are found in RSC, OPA, PPA, IPS, and FEF. Somewhat weaker shifts are found in EBA, FFA, and TPJ. Anterior visual regions appear to shift semantic tuning preferentially towards semantic categories that are particularly relevant for navigation, such as traffic signs. Thus, our results show that selective attention induces strong attentional tuning shifts between active and passive tasks.

×
×

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.

×