audioretrieval
所属分类:语音合成
开发工具:Visual C++
文件大小:358KB
下载次数:239
上传日期:2008-01-20 21:51:15
上 传 者:
miaowei
说明: 一个音频检索的例子。里面含有VC源码,并且附了一个PDF文档说明。根据自己需要修改一下可以用。
(Examples of an audio retrieval. Which contains a VC-source, and attached a PDF document explains. According to their own needs can be changed a bit.)
文件列表:
audioretrieval\LICENSE.TXT (2113, 2004-08-03)
audioretrieval\overview.txt (2131, 2004-08-03)
audioretrieval\proc\catfff.c (5355, 2003-06-12)
audioretrieval\proc\cep2spec.c (7417, 1999-06-17)
audioretrieval\proc\chopfff.c (5623, 2003-06-12)
audioretrieval\proc\fff2text.c (4452, 2003-07-11)
audioretrieval\proc\fff2trec.c (4370, 2003-07-31)
audioretrieval\proc\fff2xml.c (8966, 2003-07-11)
audioretrieval\proc\INSTALL.csh (1250, 1999-06-17)
audioretrieval\proc\Makefile (1416, 2003-07-24)
audioretrieval\proc\procdata.c (13748, 1999-06-17)
audioretrieval\proc\procfff.c (8901, 2003-06-26)
audioretrieval\proc\raw2fff.c (6149, 2003-07-17)
audioretrieval\proc\text2fff.c (6550, 2003-07-09)
audioretrieval\proc\traingauss.c (6834, 2003-06-14)
audioretrieval\src\choptree.c (9164, 1999-06-17)
audioretrieval\src\growtree.c (24292, 2003-06-10)
audioretrieval\src\histdist.c (14457, 2003-06-12)
audioretrieval\src\Makefile (1008, 2003-06-12)
audioretrieval\src\probtree.c (8255, 2003-06-12)
audioretrieval\src\quantize.c (5755, 2003-06-12)
audioretrieval\src\showhist.c (4216, 2003-06-12)
audioretrieval\src\showtree.c (14527, 2003-06-10)
audioretrieval\src\Ver-history (542, 2004-08-03)
audioretrieval\test\mfiles\box.m (457, 1999-06-17)
audioretrieval\test\mfiles\makendata.m (873, 1999-06-17)
audioretrieval\test\mfiles\plottest.m (333, 1999-06-17)
audioretrieval\test\mfiles\plottrain.m (339, 1999-06-17)
audioretrieval\test\test0.txt (1400, 1999-06-17)
audioretrieval\test\test1.txt (1345, 1999-06-17)
audioretrieval\test\train0.txt (69976, 1999-06-17)
audioretrieval\test\train1.txt (67495, 1999-06-17)
audioretrieval\test\tutorial.bash (2145, 2003-09-18)
audioretrieval\test\tutorial.csh (2136, 2003-09-18)
audioretrieval\tqlib\dbase.c (7309, 2003-06-10)
audioretrieval\tqlib\dbase.h (3859, 1999-06-17)
audioretrieval\tqlib\fffio.c (7198, 2003-07-25)
audioretrieval\tqlib\fffio.h (5781, 2003-07-25)
... ...
Welcome to the TreeQ Package.
V1.4.1, Summer 2004
Jonathan Foote
jtfoote@ieee.org
0) RTFM
Before you do anything, read the documentation in tqman.pdf
1) INSTALLATION
Installation should be simple: first make the tqlib library file in
the tqlib/ directory, then make the tools in the src/ and proc/ dirs. The
makefiles default to putting the binaries in bin/, which you can then add to your
path. If you want 'em someplace else, it's easy to change the makefile.
Once everything is built, test the code with the tutorial script in test/
(you might want to follow along in the tutorial chapter of the
documentation).
2) GETTING STARTED WITH SPEECH AND SIGNAL PROCESSING
The proc/ dir contains signal processing tools that you should use
if you don't have HTK. I've used free code from OGI to do front-end
processing like cepstral analysis. Look at proc/00readme for how to get
the code and how to link it with the proc/ tools.
3/2002: the OGI tools seem to be defunct (there's a new version which
is not compatible with my code). I don't feel comfortable giving away my
copy of their old code.
However HTK is now free and available from http://htk.eng.cam.ac.uk/
The only HTK tool you need is HCopy to paramterize your audio
files. Make sure you use the following config options:
TARGETLABEL=HTK
SAVECOMPRESSED=F
SAVEWITHCRC=F
On cygwin/Windows machines, use the following HConfig options to
get the byte swapping right:
NATURALREADORDER=T
NATURALWRITEORDER=T
If your are new to speech work I would recommend the following book as
a good intro (it will explain things like cepstral coefficients):
Rabiner, L. R. and B. H. Juang (1993). _Fundamentals of Speech Recognition_.
There's a lot of good training data for speech applications. Look at
the at the comp.speech FAQs at
ftp://rtfm.mit.edu/pub/usenet-by-hierarchy/comp/speech/
3) ADMINISTRIVIA
This code is released under a BSD-like license. You are free to use
and modify the code, however I retain the copyright and request credit
for all applications of the TreeQ code base. Please send bug reports
and comments, wise and otherwise, to jtfoote@ieee.org. I will try,
but can't guarantee, to look at them. Bug reports with fixes, of course,
are most welcome and will be rapidly included.
Good luck, have fun, and share + enjoy.
Jonathan Foote, Summer 2004.
近期下载者:
相关文件:
收藏者: