This project aims to evaluate different methods of extraction and classification in order to develop the most efficient speech recognition system possible.

To use this project, you must have a copy of the database Speaker in the Wild. You can follow the instructions at this link. Unzip the database in the folder database, located at the root of the project.

The project contains certain dependencies to python libraries and requires Python 3.6 or above. You can install these librairies and this version of python in a specific virtual environment. On Linux, this can be achieved by typing the commands on a Linux/Mac terminal on the root of the project

python3 -m venv
source venv/bin/activate
pip3 install -r requirements.txt

Note The Soundfile library cannot currently read flac audio file on Windows, the project can still be used on Windows by downloading Ubuntu from the Windows store.

You can execute the project by simply run the command :

python3 main.py