EUSTACE
HOME |
DESIGN |
EXAMPLES |
SENTENCES |
PROCEDURE |
RESULTS |
DISSERTATION |
LICENCE |
DOWNLOAD |
GLOSSARY |
|
---|
There are two version of the EUSTACE downloadable speech corpus, one containing speech files in .wav (RIFF) format and one containing speech files in .sd (ESPS) format. For each version, the top directory contains a README file, with outline information abut the corpus and a directory, speech.wav or speech.esps, which itself contains a directory for each subject, subjectf1 to subjectm3. Within each subject directory are two directories right-headed and left-headed, containing the speech files, named according to the conventions outlined in the description of the organisation of recorded sentences into sound files (see PROCEDURE page, where a full list of the available files is also available).
The downloadable speech files are available as:
The speech files should be downloaded with the documentation and
the label files, as outlined below.
There are two sets of label files, one for the test syllables and one for the additional syllables (see PROCEDURE page). Each label file contains a ESPS xlabel header, followed by a sequence of lines with the label timings on the left and the label itself on the right.
The label files are downloaded with the EUSTACE documentation, as a single gzipped tar file or as a single zip file. The top directory contains a README file and two directories: docs and labels.
The corpus is free, and licensed for non-commercial use only. Please read the licence agreement before downloading the corpus. For our records, please send an email message to Laurence White (laurence@ling.ed.ac.uk) indicating that you have downloaded the corpus.
Up to EUSTACE HOME page |