December 2022
Volume 22, Issue 14
Open Access
Vision Sciences Society Annual Meeting Abstract  |   December 2022
Two Approaches to Visual Relations: Deep Learning versus Structural Models
Author Affiliations
  • Hongjing Lu
    University of California, Los Angeles
  • Phil Kellman
    University of California, Los Angeles
Journal of Vision December 2022, Vol.22, 3086. doi:https://doi.org/10.1167/jov.22.14.3086
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Hongjing Lu, Phil Kellman; Two Approaches to Visual Relations: Deep Learning versus Structural Models. Journal of Vision 2022;22(14):3086. https://doi.org/10.1167/jov.22.14.3086.

      Download citation file:


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

      ×
  • Supplements
Abstract

Humans are remarkably adept at seeing in ways that go well beyond pattern classification. We represent bounded objects and their shapes from visual input, and also extract meaningful relations among object parts and among objects. It remains unclear what representations are deployed to achieve these feats of relation processing in vision. Can human perception of relations be best emulated by applying deep learning models to massive numbers of problems, or should learning instead focus on acquiring structural representations, coupled with the ability to compute similarities based on such representations? To address this question, we will present two modeling projects, one on abstract relations in shape perception, and one on visual analogy based on part-whole relations. In both projects we compare human performance to predictions derived from various deep learning models and from models based on structural representations. We argue that structural representations at an abstract level play an essential role in facilitating relation perception in vision.

×
×

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.

×