Progress in pattern recognition, image analysis and applications (11th Iberoamerican congress in pattern recognition, CIARP 2006, Cancun, Mexico, November 14-17, 2006)0CIARP 2006. :539-548
Copyright 2007 INIST-CNRS CC BY 4.0 Sauf mention contraire ci-dessus, le contenu de cette notice bibliographique peut être utilisé dans le cadre d’une licence CC BY 4.0 Inist-CNRS / Unless otherwise stated above, the content of this bibliographic record may be used under a CC BY 4.0 licence by Inist-CNRS / A menos que se haya señalado antes, el contenido de este registro bibliográfico puede ser utilizado al amparo de una licencia CC BY 4.0 Inist-CNRS
Notes:
Computer science; theoretical automation; systems
Accession Number:
edscal.19078988
Database:
PASCAL Archive
Further Information
The strategies for an associative recall can be based on associative memory models. However, the performance of standard associative memories is very sensitive to the number of stored patterns and their mutual correlations. With respect to huge amounts of spatial patterns (mostly correlated) to be processed, we have focused on an arbitrary number of associative memories grouped into several layers (Hierarchical Associative Memories - HAM). In the newly presented HAM2-model, the patterns are hierarchically grouped according to the previous-layer patterns. The HAM2-model uses the information recalled by the previous-layer to find an appropriate subset of next-level associative memories. To evaluate the performance of the HAM2-model, extensive simulations are carried out. The experimental results show the recall ability of the model in the area of associative pattern recall.
