Gulf Dialect Speech Recognition Using Neural Network

Communication has been an important aspect of human life, civilization, and globalization for thousands of years. Automatic Speech Recognition has found its application on various aspects of our daily lives such as biometric analysis, education, security, healthcare in smart cities. This paper implements Automatic Speech Recognition Mechanism along with Neural Network where the system converts the Arabic and Gulf dialect speech into Arabic text that is written in flask framework. A comprehensive research of utilizing a Convolutional Neural Network, Long Short Term Memory networks in speech recognition besides introducing methods for training of the neural network so that a convenient neural output can be acquired to the desired output. The proposed model was successfully tested using 35 speech recorded files containing 60,000 words, collected from different websites. The model was able to correctly recognize and convert gulf dialect speech record with accuracy 72.7%