CountPaTaKa

This task is based on ALICE to produce PaTaKa detection
See the Automatic LInguistic Unit Count Estimator (ALICE) for the reference

Requirements

ALICE has been developed and so-far tested for a range of Linux and macOS environments.

Packages:

• Conda (https://docs.conda.io/en/latest/)
• CMake (pip install cmake or conda install cmake)
• Sox (http://sox.sourceforge.net/ , brew install sox (macOS with Homebrew installed), sudo apt-get install sox(Ubuntu))

(other packages are automatically installed by conda environment)

Installation

• Clone the repository with submodules :

git clone --recurse-submodules https://github.com/orasanen/ALICE/

• Make sure you have Conda, Cmake and Sox installed.

• Create the conda environment installing all the dependencies. Note that this is OS dependent:

cd ALICE

On Linux:

conda env create -f ALICE_Linux.yml 

On macOS:

conda env create -f ALICE_macOS.yml 

Usage

Always activate the ALICE conda environment before usage. To do this, run:

  $ conda activate ALICE

Since ALICE need 16khz sampling rate and ptk_2.wav has a sampling rate of 44100.

I convert ptk_2.wav audio files to 16kHz before using the ALICE.

To process .wav files containing the audio of interest, run:

  $ ./run_ALICE.sh ptk_2_2.wav

where <data_location> = folder of .wavs, path to a .wav, or path to a .txt file with a list of .wav paths, one per row.

For GPU use during diarization , use

  $ ./run_ALICE.sh ptk_2_2.wav gpu

Result

Open ALICE_output.txt

FileID 	 phonemes 	 syllables 	 words
ptk_2_2	182	94	60

You can use ALICE read 94 syllables.

Pa , Ta and Ka each word represent one syllables.

if we need to the entire phrase “Pa Ta Ka” as one unit.

the result should be 94/3=31 times (Pa Ta Ka) for the .wav files