We are pleased to announce the release of EUSTACE, the Edinburgh
University Speech Timing Archive and Corpus of English, available at
http://www.cstr.ed.ac.uk/projects/eustace
We hope that this resource will be useful for phonetics researchers
and speech technologists working on synthesis and recognition.
The EUSTACE speech corpus comprises 4608 spoken sentences recorded at
the department of Theoretical and Applied Linguistics of Edinburgh
University. These sentences, spoken by six speakers of British
English, were designed to examine a number of durational effects in
speech and are controlled for length and phonetic content.
Subconstituents of key words in each sentence have been identified by
labels in "xlabel" (ESPS) format, which also include notes about the
prosodic realisation of some of the sentences.
Example sentences are available for playback on the website. The
complete archive, available for downloading, includes a structured
list of the sentences, the speech recordings and the label files,
together with full documentation including details of the experimental
design, recording procedure and labelling methodology. Speech waveform
files are available in ".wav" (RIFF) format and ".sd" (ESPS) format.
The downloadable corpus is free, and licensed for non-commercial use
only.
Details of the results of the original research, as described in
Laurence White's PhD dissertation, "English speech timing: a domain
and locus approach", are also included on the website, and the full
text of the dissertation is available for downloading.
Comments and questions are welcome. Please contact Laurence White
([log in to unmask]) or Simon King ([log in to unmask]).
|