Novel Approach of Implementing Speech Recognition using Neural Networks for Information Retrieval

Abstract

Objective: Retrieval of information using speech recognition and Neural Network. Methods: A novel method is proposed for information retrieval using speech recognition and Neural Network. Automatic Speech Recognition (ASR) is the technological process that allows translation of information spoken by a human being into corresponding text. Marcovian technique is used to update the sampling weight generated from the input speech. The Kalman filtering technique is used to extract the feature vector (s). Neural Network is used to classify the feature vector (s) based on the energy updation. Results: The performance of the proposed method is evaluated in terms of SNR. The system components are speech processing, feature extraction, training and testing by using neural networks and information retrieval. Conclusion: In the proposed method the retrieve process proved >90% success.