Translating text to AI Voice Speech
Author: 
Axel Fernández Curros
Guide to transform text to AI voice audio. To continue the work of the whisper audio-to-text and text-to-text translate.py script from the previous post we create a script to convert any given text to an audio file with our own voice tone from a given 10 second audio .wav file as input. The script has been modified to accept large amount of phrases so it would be possible to use a large amount of text and would not affect the audio quality.
Record your own voice
The clip should have about 8-10 seconds in a good quality. Audacity can be used to record and export the voice. Remind to export it as .wav format.
Install python dependencies
- Linux
pip3 install gitpython pip3 install gdown - Windows (chocolatey)
pip install gitpython pip install gdown
Clone repository
git clone https://github.com/Axlfc/UE5-python
Navigate into the script folder from the repository directory
cd UE5-python/Content/Python
Run the python script
- Linux
python3 voice_cloning.py "Whatever I write here is going to be read with the same voice" audio.wav - Windows
python voice_cloning.py "Whatever I write here is going to be read with the same voice" audio.wav