Speaker Identification Using Various Machine Learning Algorithms
Surbhi Bharti, Anamika Kumari, Divya Verma, Ashwni Kumar · 2024
Speaker recognition, a specialized field within the domain of audio-based biometric authentication has gain a significant attention due to its non-intrusive nature and applicability across various domains such as security, access control and identity verification. Owing to its enhanced detecting nature between a natural and difficult-to-fake trait, speaker recognition is considered a reliable method for establishing identity in these diverse contexts. This work reviews all the essential fields of speaker recognition taken into consideration favouring in the present era that boosts up Deep learning environment. Moreover, this paper highlights their advancements, challenges and potential applications of different machine learning (ML) models including Support Vector Machines (SVM) and Long Short-Term Memory Networks (LSTM). Our study aims to compare these models in terms of their F1 score and accuracy in Vox-Celeb dataset which are crucial for overall effectiveness of biometric authentication systems. Further, we will be comparing the foresaid with the existing techniques and their respective areas.