Skip to content

This repository is an implementation of Transfer Learning from Speaker Verification to Multispeaker Text-To-Speech Synthesis (SV2TTS) with a vocoder that works in real-time. SV2TTS is a three-stage deep learning framework that allows to create a numerical representation of a voice from a few seconds of audio, and to use it to condition a text-to…

smoke-trees/Voice-synthesis

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

43 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Voice Cloning and Text to Speech Synthesis

A Standalone service for cloning your own voice and synthesize any text in English in your own voice.


Read more about the procedure we followed and the findings here


Functionalities

  • Clone voices after feeding samples to it
  • Synthesized voice on custom texts
  • Speech-to-text facility for input using microphones
  • RestAPI with a testing UI for testing the model

Instructions to run the trained models

If you want to try and test out the samples trained and how the model is performing on custom text you can follow these instructions.

  • Pre-requisites:

    • For Windows

      • python (3.6 or 3.7 works best)
      • virtualenv

        If you dont have virtualenv check it out here to install

      • Trained embeddings from here
    • For linux

      • Bash shell for executing the scripts
  • Directions to install:

    • For windows
      • Clone the repo
      • Setup Virtualenv
         virtualenv env
         cd env/scripts
         activate
      • Install all requirements packages
         pip install -r requirements.txt
    • For Linux
      • Run the run.sh file to install the project
         ./run.sh

    After installing all the dependencies and environment prerequisites run the below file to check you are ready and good to go!
     ./test.sh
  • Directions to execute

    • Test through interface

      • Start the python flask server
         python app.py
      • Log on to localhost:5000 to test the model
    • Test through the Synthesize function

      • Follow the instructions given here
    • Test using Docker

        docker build -t smoketrees/voice:latest .
        docker run smoketrees/voice:latest -p 5000:5000

Instructions to train your own models

If you want to work with the source code and want to train your own models on different dataset and different language medium you can check out the instructions mentioned here

For more information about the samples tested and there results you can get all the information from here

Contributors

About

This repository is an implementation of Transfer Learning from Speaker Verification to Multispeaker Text-To-Speech Synthesis (SV2TTS) with a vocoder that works in real-time. SV2TTS is a three-stage deep learning framework that allows to create a numerical representation of a voice from a few seconds of audio, and to use it to condition a text-to…

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published