International Journal of Computer Science and Mobile Computing (IJCSMC)
Speaker recognition is one of the cost effective biometric authentication method adaptive by many online and offline applications. The mobile phone integrated passwords, voice commanders have increased the use of speaker authentication. But these recognitions are having various challenges in terms of background noise, instrumentation noise, etc. To provide the robust system authentication, it is required to provide the impurity preserved recognition. In this paper, an improved layered model is presented. In first stage of this mode, the feature exploration is achieved using LMS approach.