Neural Architecture for Mental Imaging of Sequences Based on Optical Flow Predictions

  • Authors:
  • Volker Stephan;Horst-Michael Gross

  • Affiliations:
  • -;-

  • Venue:
  • ICANN '01 Proceedings of the International Conference on Artificial Neural Networks
  • Year:
  • 2001

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this paper we present a neural architecture for a mental imaging like generation of image sequences. Mental imaging plays a central role for various perception processes. Thereto, we investigated mechanisms to model this ability of biological systems at a functional level for sequences of images. Because it is impossible to memorize many experienced sequences, we developed an universal, general and very powerful approach based on the ability to predict optic flow fields as consequences of the systems own actions and tested the resulting architecture on a real mobile system.