ABSTRACT This project is an audio-based system built using PyTorch and Python. It aims to create an Al that can 'read' a person's voice and mimic it to a passable degree. The Al, after learning thesubjects voice, can potentially say anything we want it to say via a text feeder. The AI needs to be provided with an audio sample of the person's voice which can be recorded on the spot or be a previously recorded audio file. Users can record their voices or find samples of voices and generate completely new audio files from scratch in that voice.

Submitted in the partial fulfillment of the requirements for the degree or Bachelor of Science.

1 Introduction 2 Design 3 Implementation 4 Result 5 Code.