SpeechDatasets.jl

Convenient and unified way to load a speech dataset. It can then be harnessed with other PTAL tools.

A SpeechDataset instance consists of a set of recordings (info about audio data) and annotations.

Contents

• API
• • Load a Dataset
  • Types
  • Lexicons
  • Index
• Add artifact
• • 1. Create an artifact
  • 2. Upload it to GitLab
  • 3. Add it to Artifacts.toml
  • 4. Use it in the package
• Supported Datasets
• Examples
• SpeechDatasets.jl
• • Contents
  • License
  • Authors
• Installation
• Add a new dataset
• • Functions details

License

This software is provided under the CeCILL-C license

Authors

• Lucas Ondel Yang
• Nicolas Denier
• Simon Devauchelle