Skip to content

A tool/script for batch speech data enhancement with speed/volume/RIRS/MUSAN

Notifications You must be signed in to change notification settings

zhaoyi2/audio_augment

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

20 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Audio augmentation

(README.md)

Audio augment is a tool/script for batch audio data augmenation through speed, volume, reverb, noise based on kaldi and sox.

Installation

  • kaldi
  • sox
  • cd audio_augment
  • cd tools; make KALDI=/your_kaldi_path

Usage

  • cd audio_augment
  • vim run_aug.sh to change your input_path and out_path and save
  • bash run_aug.sh, so easy!

Workflow(run_aug.sh)

  • Stage 1: Data Preparation contain data and text
  • Stage 2: speed 0.9/1.1
  • Stage 3: volume +-db
  • Stage 4: reverberation(RIRS)
  • Stage 5: MUSAN(noise/music/babble)
  • Stage 6: combine above data and select a subset of the augmend data list about twice the origin data
  • Stage 7: data and label generation

generated Examples

  • cd data/wav/train_aug listen a few enhanced aishell1 audio example through speed/volume/RIRS/MUSAN.

Reference

About

A tool/script for batch speech data enhancement with speed/volume/RIRS/MUSAN

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published