Text-dependent Speaker Recognition System Based on Speaking Frequency Characteristics

Voice recognition is one of the various applications of Digital Signal Processing and has many important real-world impacts. The topic has been investigated for quite a long time and is usually divided into two major divisions which are speaker recognition and speech recognition. Speaker recognition identifies the person who is speaking based on characteristics of the vocal utterance. On the other hand, speech recognition focuses on determining the content of the spoken message. In this project, we designed and implemented a speaker recognition system that identifies different users based on their previously stored voice samples. The samples were gathered and its features were extracted using the Mel-frequency Cepstrum Coefficient feature extraction method. These coefficients, which characterize its corresponding voice, would be stored in a database for the purpose of later comparison with future audio inputs to identify an unknown speaker. The module is currently designed to be used as a standalone device. In the future, the module is equipped with the Internet of Things (IoT) for various security systems based on human biometrics.