ABSTRACT: Background: Human biometrics encompasses various types of data used for identification and authentication, including fingerprints, iris patterns, voiceprints etc. Each type has unique characteristics suitable for different security levels. Voiceprints or speaker recognition analyses unique voice features and determines a speaker's identity using the acoustic features of their voice. Voice prints can enhance user experience and security in mobile applications. Malayalam is s a low resource south Indian language in terms of availability of speech corpora. Transformers in machine learning have shown promising results in natural language processing tasks, including speech........
KeyWord: Speaker Identification, Malayalam Language, transformers, Encoder-only model
[1].
Marcos Faundez-Zanuy,Enric Monte-Moreno,Escolauniversitariapolitècnica De Mataró, State-Of-The-Art In Speaker Recognition,Talp Research Center (Bercelona)Avda. Puig I Cadafalch 101-111,08303 Mataro (Barcelona) Spain
[2].
A Review On Speech Recognition Technique, International Journal Of Computer Applications, Vol 10, No.3, Pp. 16-24. 2010.
[3].
Adwa Abakarim, Abdenbiabenaou,Comparative Study To Realize An Automatic Speaker Recognition System,International Journal Of Electrical And Computer Engineering (Ijece),Vol. 12, No. 1, Pp. 376~382,2022
[4].
Xu Xiang, Shuai Wang, Houjun Huang, Yanminqian, Kai Yu,Margin Matters: Towards More Discriminative Deep Neural Network Embeddings For Speaker Recognition,2019
[5].
R. Sharma, D. Govind,J. Mishra, A. K. Dubey, K. T. Deepak, S. R. M. Prasanna,Milestones In Speaker Recognition,Artifcial Intelligence Review 57:58,2024