A run-through: Text independent speaker identification using deep learning
Pooja Gambhir, Amita Dev · 2021
Speaker identification determines an unknown speaker’s identity which is a 1: N match where the speech is compared against multiple templates. Speaker identification is categorized under speaker recognition which can be implemented covertly without the user’s knowledge to identify the speakers in the discussion, designing automated alert systems of speaker’s voice change, and many more areas of applications. Empirical solutions for speaker identification are often subject to noise and speaker’s disguised voice and remain one of the challenging fields of research in speech-processing applications. A study on deep learning-based feature extraction and classification techniques for Speaker Identification is presented in this study. This study summarizes the current trends of DNN practices, their requirement in the field of speaker recognition for Large Vocabulary Speaker dataset.