JCOM_V4_N4_RP4 Indraneel Sreeram Venkata Praveen Kumar Vuppala Journal on Computer Science 2347–6141 4 4 25 31 Automated Teller Machines, Feature Extraction, Discrete Wavelet Transforms, Wavelet Packet Decomposition, Pattern Recognition, Artificial Neural Networks Automated Teller Machines (ATMs) enable customers to perform financial transactions like cash withdrawal, check balances or credit mobile phones with the help of the machine and without any human intervention. In this paper, a speech recognition system is developed for ATMs for performing financial transactions. Since speech is the most natural and easiest mode of communication, communication between the customer and the ATM can be performed easily through speech. This allows ATM machines to communicate with the customers using the stored speech samples and the user communicates with the machine through spoken digits. Here, two speech recognition systems are developed using the feature extraction techniques Discrete Wavelet Transforms (DWT) and Wavelet Packet Decomposition (WPD). The classifier used along with both the methods is Artificial Neural Network (ANN). Both methods produced good recognition accuracies of 87.575% and 85.5% each, but DWT and ANN combination produced better results than WPD and ANN. December 2016 - February 2017 Copyright © 2017 i-manager publications. All rights reserved. i-manager Publications http://www.imanagerpublications.com/Article.aspx?ArticleId=13417