Error Minimization For Language-Independent Speaker Identification System
Ajay Kumar Mishra, Amit Kaul · Proceedings of the Conference on Advances in Communication and Control Systems-2013 · 2013
In this paper, error of intra-speaker speech of a language-independent speaker identification system is minimized. Well-known Mel Frequency Cepstral coefficients {MFCCs} have been used for feature extraction. The decision threshold is located at the point where the probabilities of both errors are equal. To make system realistic, we will make system model by using GMM (Gaussian Mixture Model). Using these models, we build speaker identifiers that are capable of identify a speaker regardless of the language of speaker.