Text to speech service (TTS). Neural Network (NN) based Speech Synthesis System

Description

Based on the Merlin project (https://github.com/CSTR-Edinburgh/merlin/). NN trained on 'cmu_us_bdl_arctic' dataset (male voice) prepared by Carnegie Mellon University.

Usage

In the command line run following commands

docker build -t tts .
docker run -it -v `pwd`/out:/opt/out tts

Default text ('Hello, my name is Joker ...') will be used. Resulting audio file location: 'out/tts.wav'.

Provide your own text

run docker container with your text as argument:

docker run -it -v `pwd`/out:/opt/out tts "Hello. This is text to speech service"

Resulting audio file location: 'out/tts.wav'.

(c) Abylay Ospan aospan@jokersys.com, 2017 https://jokersys.com