Sylvain Chartier1, Mounir Boukadoum
1Departement d'Informatique, Université du Québec à Montréal, Montreal, QC H3P 3P8, Canada.
This study introduces a novel Bidirectional Associative Memory (BAM) model for multistep pattern recognition. The enhanced BAM model efficiently learns multiple, variable-length pattern series, including gray-level images, without complex algorithms.
You might also read
Articles linked to this work by shared authors, journal, and citation graph.
Area of Science:
Background:
Purpose of the Study:
Main Methods:
Main Results:
Conclusions: