Speech Recognition in ATMs: Application of Linear Predictive Coding and Support Vector Machines
Sonia Sunny · International Journal for Research in Applied Science and Engineering Technology · 2017
Today, Automated Teller Machines (ATMs) are extensively used by people for financial transactions.It provides a convenient, fast and easy way for customers to access cash.In this paper, a speech recognition system is developed for financial transactions in ATMs using Linear Predictive Coding (LPC) and Support Vector Machines (SVM).Voice signals are sampled directly from the microphone and then they are processed using LPC for extracting the features.Training, testing and pattern recognition are performed using Support Vector Machines.The proposed method is implemented for 200 speakers uttering 10 spoken digits in English.This hybrid architecture of LPC and SVM produced rather good recognition accuracy of 80.25%.