Speech Recognition Using Artificial Neural Networks

Abstract

The objective of this study is to evaluate the potentiality of using Artificial Neural Networks (ANNs) for Speech Recognition. The Linear Predictive Code (LPC) was used for the feature extractions of the word used. The speech data (spoken words) has been converted in to voice signals in digital format. MS-Excel package has been used to generate 600 learning pattern. 540 were used to train General Regression Neural Network (GRNN) and Back Propagation Network (BPN) architecture. The reminder 60 patterns were used to test the performance of the trained shell. The General Regression Neural Network (GRNN) was found to be able to recognize speech patterns and process test patterns with an average error of ±0.016667 while the standard deviation (STVD) was ±0.129099. Otherwise, the average of the BPN was ±1.168 ×10-4.