Phoneme recognition system

Date of Publication

1994

Document Type

Bachelor's Thesis

Degree Name

Bachelor of Science in Computer Science

College

College of Computer Studies

Department/Unit

Computer Science

Abstract/Summary

Executive Summary. Back propagation algorithm is currently the most widely used algorithm in neural network learning and has been successfully used in various tasks including knowledge representation in semantic networks, hand-written character recognition, speech synthesis and speech recognition. Principal reasons for the widespread use of Backpropagation algorithm lies in its ability to generalized on the unseen patterns and its inherent capability to build arbitrary non-linear boundaries between input and output representation.

Based on the given strength of Back Propagation Network, the speaker dependent Phoneme Recognition System was designed and implemented using this algorithm. It was applied in the identification of the smallest unit of speech known as Phonemes. However, before the system can perform recognition, the input signal acquired from the user must be first segmented based from the acoustic changes being observed. Analysis of the segments follows by extracting the pertinent frequency patterns of the signal to be learned by the network. Once sufficiently trained, recognition can now be performed.

The Phoneme Recognition System can be used as a stepping stone for enhanced word recognition systems. Adding the grammar or the structure of the permitted phoneme sequences can increase the recognition rate by eliminating candidate phone sequences at variance with the grammar.

Abstract Format

html

Language

English

Format

Print

Accession Number

TU09114

Shelf Location

Archives, The Learning Commons, 12F, Henry Sy Sr. Hall

Physical Description

73 numb. leaves ; Computer print-out.

This document is currently not available here.

Share

COinS