Script VoxForge

Following the tutorial offered by VoxForge, this project contains a full script to create acoustic model for the Julius Decoder using the HTK toolkit.

Directories

Type	Name	Function
Directory	clara	Contain grammar, voca and wavs.
Directory	htk	Speech Recognition toolkit.
Directory	julius-4.3.1	Speech recognition engine.
Directory	scripts	Contains the base scripts to generate LM.
File	grammar.sh	Executes programs to generate the LM.
File	manage_files.sh	Manages files to run the scripts.
File	setup.sh	Installs dependencies.
File	run_all.sh	Runs all scripts.
File	README.md	General instructions of the project.
File	.gitignore	Files/folders ignored by git.

Prepare Data

Add your grammar, voca, jconf, wavs, lexicon and prompts on clara folder.

Run

$ chmod +x setup.sh
$ chmod +x run_all.sh
$ ./setup.sh
$ ./run_all.sh

TODO

• Let more general on scripts the folder that contains voca, grammar and wav.

Author

• @Dayof
  • Dayanne Fernandes da Cunha
  • dayannefernandesc@gmail.com