Skip to content

Latest commit

 

History

History
29 lines (22 loc) · 1.35 KB

README.md

File metadata and controls

29 lines (22 loc) · 1.35 KB
Raw

Arabic-Phonetiser

Convert Arabic diacritised text to a sequence of phonemes and create a pronunciation dictionary from them for alignment using HTK

Usage

  from phonetise-Arabic import phonetise
  phonemes = phonetise(Arabic_text)

or

  phonetise-Buckwalter.py [inputfile]

[inputfile] should be a utf8 text file contianing in every line:

  "[sound-filename]" "[arabic-text-in-buckwalter]"
  "[sound-filename]" "[arabic-text-in-buckwalter]"
  "[sound-filename]" "[arabic-text-in-buckwalter]"
  "[sound-filename]" "[arabic-text-in-buckwalter]"
  ...

the output will be two files: dict: contianing the sorted pronunciation dicationary with a carrage return at the end for use with tools like HTK utterance-pronunciations.txt: A file contianing in every line:

  "[sound-filename]" "[phoneme-sequence]"
  "[sound-filename]" "[phoneme-sequence]"
  "[sound-filename]" "[phoneme-sequence]"
  "[sound-filename]" "[phoneme-sequence]"

License

Creative Commons Licence
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.