Abstract:
This project aims to develop an application that can generate voice clones based on
provided text using a trained model, we named this app Deepcloning.ai. Many Text-toSpeech models exist, they are limited to specific voices. Our focus is on training a model
with an English dataset to replicate a particular person's voice. Once trained, the model can
produce Deepfakes voices for the provided text, offering advantages in education, costeffective advertising, and the media industry. However, it's essential to acknowledge the
potential for misuse, especially in serious crimes and impersonation, necessitating
subsequent efforts for detection and prevention. This project's primary objective is
Deepfakes text to speech creation, with ethical and security considerations in mind.
