EFFICIENT DEEP LEARNING MODEL FOR RECOGNIZE ARTISTS VOICE

Abstract

Speech is the basic, common and economical sort of communication methodology for everyone to move with each another. These days speech technologies area limited accessible normally however applicable to attention-grabbing vary of task. These technologies alter machines to reply properly and dependably to human voices. Speaker recognition is that the identification of an individual from characteristics of his/her voice is a vital human attribute most regard granted in natural human to human interaction or communication. The most purpose of voice recognition is to spot or acknowledge the voice or audio files by extracting the options from the audio files by exploitation spectrum of audio to acknowledge the voice of a specific creative person. In this, we apply the different deep learning techniques like GMM, HMM and CNN to identify the artist by his songs. We observed that CNN provides good accuracy compare to remaining.