Automatic Speaker Recognition in Diverse Data Condition. A System Based on Signal Processing and Deep Learning Techniques

      Аннотация: Speaker recognition is the process of identifying and verifying a speaker’s voiceprints from several voices. The main objective of speaker recognition is to create a system that is capable of verifying a person’s voice and can also be proficient in identifying a person from his/her voice. This book discusses the challenges of the Speaker recognition system in practical environments. It is observed that the input signal conditions and non-speaker information have a large impact on the speaker recognition system. This book provides some solutions to the problems associated with speaker recognition in degraded signal conditions. Here some novel pre-processing, feature extraction, and modeling techniques are discussed for speaker recognition.

Ключевые слова: Speaker Recognition, speaker verificcation, speaker identification, Machine Learning, Pattern recognition