PROJECT TITLE :

Speaker Recognition by Machines and Humans: A tutorial review

ABSTRACT:

Identifying someone by his or her voice is a vital human trait most take with a pinch of salt in natural human-to-human interaction/communication. Speaking to somebody over the telephone typically begins by identifying who is speaking and, a minimum of in cases of familiar speakers, a subjective verification by the listener that the identity is correct and the conversation will proceed. Automatic speaker-recognition systems have emerged as an vital means that of verifying identity in many e-commerce applications also normally business interactions, forensics, and law enforcement. Human specialists trained in forensic speaker recognition can perform this task even higher by examining a set of acoustic, prosodic, and linguistic characteristics of speech during a general approach referred to as structured listening. Techniques in forensic speaker recognition have been developed for many years by forensic speech scientists and linguists to assist scale back any potential bias or preconceived understanding on the validity of an unknown audio sample and a reference template from a possible suspect. Experienced researchers in signal processing and machine learning still develop automatic algorithms to effectively perform speaker recognition?with ever-improving performance?to the point where automatic systems begin to perform on par with human listeners. In this text, we review the literature on speaker recognition by machines and humans, with an emphasis on outstanding speaker-modeling techniques that have emerged within the last decade for automatic systems. We have a tendency to discuss totally different aspects of automatic systems, including voice-activity detection (VAD), options, speaker models, customary evaluation information sets, and performance metrics. Human speaker recognition is discussed in two parts?the primary half involves forensic speaker-recognition strategies, and therefore the second illustrates how a na?ve listener performs this task from a neuroscience perspective. We conclude this review with a comparative- study of human versus machine speaker recognition and attempt to point out strengths and weaknesses of each.