Share Email Print

Proceedings Paper

Temporal pattern recognition using one-memory-element-per-state sequential neural network
Author(s): Tet H. Yeap; S. G. Zaky; John K. Tsotsos; H. C. Kwan
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

This paper presents a special sequential neural network for temporal pattern recognition using one memory element per state. The network generates a single response representing a sequence of events by utilizing the process of temporal integration. That is, the response is generated in small increments at each time step by summing in time the recognition result of each event. Models for motion detection and speech recognition based on the proposed network were implemented. Simulation results show that the network is tolerant to noise, and can recognize partial sequences.

Paper Details

Date Published: 19 August 1993
PDF: 9 pages
Proc. SPIE 1966, Science of Artificial Neural Networks II, (19 August 1993); doi: 10.1117/12.152649
Show Author Affiliations
Tet H. Yeap, Univ. of Ottawa (Canada)
S. G. Zaky, Univ. of Toronto (Canada)
John K. Tsotsos, Univ. of Toronto (Canada)
H. C. Kwan, Univ. of Toronto (Canada)

Published in SPIE Proceedings Vol. 1966:
Science of Artificial Neural Networks II
Dennis W. Ruck, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?