252 lines
9.1 KiB
Plaintext
252 lines
9.1 KiB
Plaintext
EDINBURGH SPEECH TOOLS LIBRARY
|
|
VERSION 2.4 December 2014
|
|
|
|
Principal authors: Alan W Black, Paul Taylor, Richard Caley,
|
|
Simon King, Rob Clark and Korin Richmond
|
|
|
|
Centre for Speech Technology
|
|
University of Edinburgh
|
|
Informatics Forum
|
|
10 Crichton Street
|
|
Edinburgh EH8 9AB, UK
|
|
http://www.cstr.ed.ac.uk/projects/speech_tools
|
|
|
|
Although this is a considered stable release, there are still a number
|
|
of known parts are missing or incomplete.
|
|
|
|
------------------------------------------------------------------------
|
|
|
|
The Edinburgh speech tools system is a library of C++ classes, functions
|
|
and utility programs that are frequently used in speech software. The
|
|
system compiles to a single unix library .a file which can be linked
|
|
with software. At present, C++ classes for several useful speech and
|
|
language classes have been written, along with audio software and some
|
|
basic signal processing software.
|
|
|
|
========================================================================
|
|
|
|
|
|
Edinburgh Speech Tools Library version 2.4
|
|
Centre for Speech Technology Research
|
|
University of Edinburgh, UK
|
|
Copyright (c) 1994-2014
|
|
All Rights Reserved.
|
|
|
|
Permission is hereby granted, free of charge, to use and distribute
|
|
this software and its documentation without restriction, including
|
|
without limitation the rights to use, copy, modify, merge, publish,
|
|
distribute, sublicense, and/or sell copies of this work, and to
|
|
permit persons to whom this work is furnished to do so, subject to
|
|
the following conditions:
|
|
1. The code must retain the above copyright notice, this list of
|
|
conditions and the following disclaimer.
|
|
2. Any modifications must be clearly marked as such.
|
|
3. Original authors' names are not deleted.
|
|
4. The authors' names are not used to endorse or promote products
|
|
derived from this software without specific prior written
|
|
permission.
|
|
|
|
THE UNIVERSITY OF EDINBURGH AND THE CONTRIBUTORS TO THIS WORK
|
|
DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, INCLUDING
|
|
ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO EVENT
|
|
SHALL THE UNIVERSITY OF EDINBURGH NOR THE CONTRIBUTORS BE LIABLE
|
|
FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
|
|
WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN
|
|
AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION,
|
|
ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF
|
|
THIS SOFTWARE.
|
|
|
|
========================================================================
|
|
|
|
The following c++ programs are available:
|
|
|
|
na_play: generic playback program for use with net_audio and CSTR ao.
|
|
|
|
ch_wave: Waveform file conversion program.
|
|
|
|
ch_lab: label file conversion program.
|
|
|
|
ch_track: Track file conversion program.
|
|
|
|
wagon: a CART tree build and test program
|
|
|
|
And others
|
|
|
|
The following C++ sub-libraries are available
|
|
|
|
audio: C++ audio functions for Network Audio system, Suns
|
|
Linux and FreeBSD
|
|
|
|
speech_class: C++ speech classes, including waveform and track.
|
|
|
|
ling_class: C++ linguistic classes.
|
|
|
|
sigpr: Signal processing
|
|
|
|
utils: Various utilities.
|
|
|
|
|
|
COPYING
|
|
------------------------------------------------------------------------
|
|
|
|
Some files in the distribution do not fall under the above copyright.
|
|
However none have any commercial restrictions, and no files fall under
|
|
the GPL. Specifically files that have different licences are as
|
|
follows:
|
|
|
|
grammar/wfst/wfst_train.cc
|
|
Copyright Carnegie Mellon University (under the same
|
|
open licences as much of the rest of the code).
|
|
siod/
|
|
include/siod*.h
|
|
lib/siod/siod.scm
|
|
Apart from some new functionality under our licence the SIOD
|
|
files fall under the licence given in siod/slib.cc copyright
|
|
Paradigm Associates. This code has been substantially modified
|
|
by us though our changes are under that licence.
|
|
|
|
editline.c editline.h el_complete.c el_sysunix.c el_unix.h
|
|
are copyright 1992 Simmule Turner and Rich Salz under the
|
|
"BSD like" licence in siod/editline.c Again substantial
|
|
changes have been made by us which continue to be under that
|
|
licence
|
|
|
|
rxp/
|
|
include/rxp/
|
|
Is Copyright Richard Tobin. We have a specific version of RXP
|
|
which allows free (including non-commercial) use.
|
|
|
|
base_class/raveconv.cc
|
|
Is copyright 1992, 1995 by Markus Mummert
|
|
|
|
base_class/string/reg*.{c,cc}
|
|
Copyright (c) 1986 by University of Toronto.
|
|
|
|
base_class/string/EST_strcasecmp.c
|
|
Copyright (c) 1987, 1993
|
|
The Regents of the University of California. All rights reserved.
|
|
|
|
The standard BSD licence
|
|
|
|
|
|
ACKNOWLEDGEMENTS
|
|
------------------------------------------------------------------------
|
|
|
|
Most of the software was written by Paul Taylor, Alan W Black, Simon
|
|
King, Richard Caley, Rob Clark and Korin Richmond. Additional
|
|
contributions have been made by the following people. Slightly
|
|
different licence agreements accompany these pieces of code, so it is
|
|
important that the various agreements be properly understood.
|
|
|
|
Tony Robinson: some signal processing code.
|
|
Markus Mummert: sample rate converter (rateconv.cc).
|
|
Craig Reese and Joe Campbell: ulaw conversion code.
|
|
Paul Bagshaw: pitch tracker.
|
|
Theo Veenker: IRIX audio support.
|
|
Stan Chen: AIX support
|
|
Paradigm Assoc. and George Carrett: For Scheme In One Defun
|
|
Simmule Turner and Rich Salz: for editline
|
|
Henry Spencer: for regex code
|
|
The Regents of the University of California: some string comparison code
|
|
Richard Tobin: RXP, XML parser.
|
|
Arthur Toth: Windows server code
|
|
Pierre Lorenzon <devel@pollock-nageoire.net> gcc-4.7 patches
|
|
Sergio Oller: for making the documentation work again
|
|
|
|
INSTALLATION
|
|
************
|
|
|
|
see INSTALL file for details
|
|
|
|
FILE/DIRECTORY STRUCTURE
|
|
------------------------------------------------------------------------
|
|
|
|
The main directory contains minimal wrap-around main() functions
|
|
which compile to form executables which use the algorithms in the
|
|
speech tools. All these main() functions really do is to read
|
|
command line options and read and write files, the algorithms
|
|
themselves are completely contained within the libraries.
|
|
|
|
The directory structure is as follows:
|
|
|
|
config:
|
|
shared make rules and configuration make rules
|
|
|
|
speech_class:
|
|
general speech class files, including waveforms, contours etc.
|
|
|
|
ling_class:
|
|
general linguistic class files, including a stream-based architecture.
|
|
|
|
sigpr:
|
|
signal processing code.
|
|
|
|
doc:
|
|
basic documentation, generates html, info and postscript
|
|
|
|
include:
|
|
class and general header files.
|
|
|
|
lib:
|
|
where the .a files are kept
|
|
|
|
main:
|
|
Directory containing main files which compile to executables.
|
|
|
|
|
|
DOCUMENTATION
|
|
------------------------------------------------------------------------
|
|
|
|
There are two sources of documentation.
|
|
|
|
Man pages exist for the more important stand-alone programs. These are
|
|
found in the main directory.
|
|
|
|
The library documentation has substantiall improved in this version
|
|
using an integrated DOC++, JADE and docbook combination. Classes are
|
|
described at the overview, theorectical, practical and code level.
|
|
The build process for the documentation is still dependent on some
|
|
more obscure packages so it is recommended that you take the
|
|
pre-built documentation. Or access it on-line through
|
|
http://www.cstr.ed.ac.uk/projects/speech_tools
|
|
|
|
The library documentation is in texinfo format which can be converted
|
|
into GNU info HTML and postscript, see Makefile in doc/
|
|
|
|
Note the documentation is not up to date. We are in the process of moving the documentation to Doxygen.
|
|
|
|
BUGS/COMMENTS ETC
|
|
------------------------------------------------------------------------
|
|
|
|
Substantial parts of this release are considered mature and should be
|
|
relatively free of bugs, however there are newer parts which have not
|
|
been rigourously tested yet. If you find any bugs, please send a bug
|
|
report to speechtools@cstr.ed.ac.uk. Any useful comments will also be
|
|
appreciated. See also the bugs file in the doc directory.
|
|
|
|
If you have some software that you think could be integrated into the
|
|
speechtools system, please contact us at
|
|
speechtools@cstr.ed.ac.uk. The whole point of writing and releasing
|
|
code such as this is to prevent duplication of effort, and therefore
|
|
we would appreciate greatly any code which will add to the usefulness
|
|
of the system. Signal processing routines are particularly welcome.
|
|
|
|
|
|
======================================================================
|
|
ADDITIONAL SOFTWARE
|
|
======================================================================
|
|
|
|
CSTR supplies a number of different speech technology systems
|
|
including speech synthesis systems and (in the future) a recognition
|
|
system. To find out more information checkout http://www.cstr.ed.ac.uk/
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|