yes ,i have check it. I find the resource say that pocketsphinx is only used
in embedded system. Is it right? i now setup ASR system with sphinx3 in PC.
and i download the Acoustic Model with version 0.3, it will run with no error.
and the up error is
"FATAL_ERROR: "mdef.c", line 705: Version error: Expecing 0.3, but read BMDF"
which wll be decode by pocketsphinx .
so , when can i download Chinese DEMO Acostics Model and Language Model for
Sphinx3, which the Acoustic Model version is 0.3. or if there are no these two
models ,how can i create these two models, thanks!
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
All the models available for Chinese, including accurate broadcast news PTM
model available in downloads section is supported only in pocketsphinx. There
is no sphinx3 Chinese model.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I setup Sphinx3, run in sphinx3_livecode , error occur:
INFO: Transition Matrices File:
D:\sphinx3\run_demo\zh\model\hmm\zh_cn\tdt_sc_8k/transition_matrices
INFO: mdef.c(694): Reading model definition:
D:\sphinx3\run_demo\zh\model\hmm\zh_cn\tdt_sc_8k/mdef
FATAL_ERROR: "mdef.c", line 705: Version error: Expecing 0.3, but read BMDF
the cfg file is :
-samprate 16000
-hmm D:\sphinx3\run_demo\zh\model\hmm\zh_cn\tdt_sc_8k
-dict D:\sphinx3\run_demo\zh\model\lm\zh_cn\zh_CN\mandarin_notone.dic
-lm D:\sphinx3\run_demo\zh\model\lm\zh_cn\zh_CN\gigatdt.5000.DMP
Idownload the Acoustic Model and Language Model from
http://cmusphinx.svn.sourceforge.net
{
http://cmusphinx.svn.sourceforge.net/viewvc/cmusphinx/trunk/pocketsphinx/mode
l/hmm/
http://cmusphinx.svn.sourceforge.net/viewvc/cmusphinx/trunk/pocketsphinx/mode
l/lm/
}
Does the HMM can be used which download from the website,
and how to deal with this issue, tks!
I use the english model , the same error occur.
INFO: Mixture Weight File: (null)
INFO: Transition Matrices File:
D:\sphinx3\run_demo\en\hmm\en_us\hub4wsj_sc_8k/transition_matrices
INFO: mdef.c(694): Reading model definition:
D:\sphinx3\run_demo\en\hmm\en_us\hub4wsj_sc_8k/mdef
FATAL_ERROR: "mdef.c", line 705: Version error: Expecing 0.3, but read BMDF
.....
You can use pocketsphinx to decode with those models
yes ,i have check it. I find the resource say that pocketsphinx is only used
in embedded system. Is it right? i now setup ASR system with sphinx3 in PC.
and i download the Acoustic Model with version 0.3, it will run with no error.
and the up error is
"FATAL_ERROR: "mdef.c", line 705: Version error: Expecing 0.3, but read BMDF"
which wll be decode by pocketsphinx .
so , when can i download Chinese DEMO Acostics Model and Language Model for
Sphinx3, which the Acoustic Model version is 0.3. or if there are no these two
models ,how can i create these two models, thanks!
Your resource is wrong. Pocketsphinx is a recommended engine for decoding. You
can learn more about CMUSphinx versions on our website
http://cmusphinx.sourceforge.net/wiki/versions
All the models available for Chinese, including accurate broadcast news PTM
model available in downloads section is supported only in pocketsphinx. There
is no sphinx3 Chinese model.
Thanks all, especial nshmyrev.
does the resource I download right, and does there are the lasted ?