A combined self-organizing feature map and multilayer perceptron for isolated word recognition

• DOI: 10.1109/78.165652
• Publication Date: 2002-08-24T19:31:03Z
• Authors: Z. Kuang; A. Kuh
• Keywords: Self-organizing map; Feature (linguistics)

A neural network system which combines a self-organizing feature map and multilayer perception for the problem of isolated word speech recognition is presented. new method combining self-organization learning K-means clustering used training map, an efficient adaptive nearby-search coding based on 'locality' designed. The shown to save about 50% computation without degradation in rate compared full-search coding. Various experiments different choices parameters were conducted TI 20 database with best rates as high 99.5% both speaker-dependent multispeaker-dependent tests.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>