REPRESENTATIONAL AXES AND TEMPORAL COOPERATIVE PROCESSES

Authors
  1. Tsotsos, J.K.
Corporate Authors
Toronto Univ, Toronto ONT (CAN) Dept of Computer Science;Defence Research Establishment Atlantic, Dartmouth NS (CAN)
Abstract
A framework for the integration of time into high level (attentive) vision is described. The key elements are an organization of knowledge along several axes, including time, several search modes facilitated by the knowledge organization, a hypothesize and test reasoning framework and a temporal cooperative process, driven by the knowledge organization, for hypothesis ranking. The major aspects of the cooperative process, namely, the definition of consistency, process neighbourhoods, initial certainties, compatibility factors, are all defined in terms of the knowledge organization. Iterations are equated with temporal measurements, and this allows temporal sampling to be discussed quantitatively. A qualitative analysis that includes hypothesis response rise and fall times is presented in addition to guidelines for setting compatibilities such that performance is appropriate. TRUNCATED
Report Number
RCBV-TR-84-2 — Contractor Report, Research in Biological and Computational Vision
Date of publication
15 Apr 1984
Number of Pages
66
DSTKIM No
85-00625
CANDIS No
45456
Format(s):
Hardcopy;Originator's fiche received by DSIS

Permanent link

Document 1 of 1

Date modified: