1. GENERAL INFORMATION:
----------------------------------------

These are recordings and transcriptions to be used for
building an ASR system. There are typically 200 speakers 
per languange and 30 utterances per speaker,  14 of which 
are eliceted speech and the others are read speach. The 
recordings are of telephone data, and the text was selected for phoneme coverage.

Language: One language per directory.
There is a subdirectory per speaker for each language.
The audio and transcription file names indicate speaker and utterance number, e.g. 
<language name>_<speaker number>_<utterance number>.wav for the audio files
<language name>_<speaker number>_<utterance number>.txt for the transcription text file.


2. LICENCE:
----------------------------------------

Refer to the "LICENCE" file in this directory.


3. TECHNICAL:
----------------------------------------

Description of files:

README                 -    This file.
LICENCE                -    Text file containing licensing information.
METADATA.csv	       -    Text file containing metadata regarding landline/modile, age and gender as available
/wavs/*.wav            -    Recordings in wav format.
/transcriptions/*.txt  -    Phonetic transcriptions.