Speech recognition for Arabic numbers using empirical mode decomposition

Abstract

One of the main uses of artificial intelligence is automatic speech recognition. In this study, empirical mode decomposition is used to enhance the detection of speech dysarthria for speech recognition the problem in this paper is how to distinguish between the spoken words in Arabic we chose speech recognition by Empirical Mode Decomposition (EMD) to analyze the sound they were results as a confusion matrix results which indicates a classifier high accuracy (0.8966). Using 29 sounds to recognize by the EMD method the primary objective of this paper is to use EMD to Dysarthria speech recognize Arabic speech in real-time.