https://github.com/oscarknagg/voicemap
Identifying people from small audio fragments
https://github.com/oscarknagg/voicemap
convolutional-neural-networks machine-learning speaker-identification speaker-recognition
Last synced: about 1 month ago
JSON representation
Identifying people from small audio fragments
- Host: GitHub
- URL: https://github.com/oscarknagg/voicemap
- Owner: oscarknagg
- Created: 2018-09-01T21:04:00.000Z (almost 7 years ago)
- Default Branch: master
- Last Pushed: 2020-04-10T08:03:57.000Z (about 5 years ago)
- Last Synced: 2025-04-07T07:42:52.145Z (2 months ago)
- Topics: convolutional-neural-networks, machine-learning, speaker-identification, speaker-recognition
- Language: Python
- Homepage:
- Size: 3.18 MB
- Stars: 170
- Watchers: 6
- Forks: 73
- Open Issues: 9
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# voicemap
This repository contains code to build deep learning models to identify
different speakers based on audio samples containg their voice.
The eventual aim is for this repository to become a pip-installable
python package for quickly and easily performing speaker identification
related tasks.
**This tensorflow/Keras/python2.7 branch is discontinued. Work is
continuing on the pytorch-python-3.6 branch which will become the
master branch.**
## Instructions
#### Requirements
Make a new virtualenv and install requirements from `requirements.txt`
with the following command.
```
pip install -r requirements.txt
```
This project was written in Python 2.7.12 so I cannot guarantee it works
on any other version.
#### Data
Get training data here: http://www.openslr.org/12
- train-clean-100.tar.gz
- train-clean-360.tar.gz
- dev-clean.tar.gz
Place the unzipped training data into the `data/` folder so the file
structure is as follows:
```
data/
LibriSpeech/
dev-clean/
train-clean-100/
train-clean-360/
SPEAKERS.TXT
```
Please use the `SPEAKERS.TXT` supplied in the repo as I've made a few
corrections to the one found at openslr.org.
#### Run tests
This requires the LibriSpeech data.
```
python -m unittest tests.tests
```
## Contents
### voicemap
This package contains re-usable code for defining network architectures,
interacting with datasets and many utility functions.
### experiments
This package contains experiments in the form of python scripts.
### notebooks
This folder contains Jupyter notebooks used for interactive
visualisation and analysis.