May 2008
Volume 8, Issue 6
Free
Vision Sciences Society Annual Meeting Abstract  |   May 2008
Neural dynamics of visually-based object segmentation and navigation in complex environments
Author Affiliations
  • Ennio Mingolla
    Department of Cognitive and Neural Systems, Center for Adaptive Systems and Center of Excellence for Learning in Education, Science and Technology, Boston University
  • N Andrew Browning
    Department of Cognitive and Neural Systems, Center for Adaptive Systems and Center of Excellence for Learning in Education, Science and Technology, Boston University
  • Stephen Grossberg
    Department of Cognitive and Neural Systems, Center for Adaptive Systems and Center of Excellence for Learning in Education, Science and Technology, Boston University
Journal of Vision May 2008, Vol.8, 1154. doi:10.1167/8.6.1154
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to Subscribers Only
      Sign In or Create an Account ×
    • Get Citation

      Ennio Mingolla, N Andrew Browning, Stephen Grossberg; Neural dynamics of visually-based object segmentation and navigation in complex environments. Journal of Vision 2008;8(6):1154. doi: 10.1167/8.6.1154.

      Download citation file:


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

      ×
  • Supplements
Abstract

Visually guided navigation through a cluttered natural scene is a challenging problem that animals and humans accomplish with ease. A dynamical neural model proposes how primates use motion information to segment objects and determine heading. These competences reflect complementary properties of MT- / MSTv and MT+ / MSTd pathways. The model clarifies the functions of magnocellular cells in primate retina, V1, MT and MST that relate to heading and describes how feedforward, feedback and lateral brain circuits are used to perform motion estimation calculations. The model retina responds to transients in the input stream, producing moving boundary representations. Model V1 generates a local speed and direction estimate that is noisy due to the neural aperture problem. Model MT+ computes a global motion estimate supporting accurate heading estimation in MSTd. Modulatory attentional feedback from MSTd reduces motion ambiguities in MT+. The model quantitatively simulates properties of human data during heading estimation tasks. Simulated rotations less than 1 degree per second do not affect accuracy, whereas faster simulated rotations do (Royden et al. 1994, Vis Res 34). MT- segments objects using differential motion. Feedback from MSTv to MT- helps resolve the aperture problem and drives global motion capture: When an object moves differently from the background, MT- segments the object and computes accurate estimates of object motion. The resolution of the aperture problem in the model MT- displays the same time-course as primate MT (Pack and Born 2001, Nature 409). Model representations activate processes of visually guided steering and navigation. The model is tested both through simulations of the cellular dynamics of cortical cells and on complex natural image sequences. This distinguishes our model from other models which either include cellular dynamics or focus on real world applications.

Mingolla, E. Browning, N. A. Grossberg, S. (2008). Neural dynamics of visually-based object segmentation and navigation in complex environments [Abstract]. Journal of Vision, 8(6):1154, 1154a, http://journalofvision.org/8/6/1154/, doi:10.1167/8.6.1154. [CrossRef]
Footnotes
 Supported in part by the National Science Foundation (NSF BCS-0235398 and NSF SBE-0354378), the National-Geospatial Intelligence Agency (NMS201-01-1-2016), and the Office of Naval Research (ONR N00014-01-1-0624).
×
×

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.

×