I'm a total newbie to Sphinx4. Just trying to get the Transcriber demo to work. I have a wav file which is PCM, 16kHz, 16Bit, Mono and the same thing as Stereo. The speech in it is just "this is a test". When I run the demo, all I get is stuff like
six six oh six four zero six
six two seven
or
five oh oh five
five oh oh
Can someone explain to me why this is happening and how I could get the demo to run? I'd appreciate any help.
Hello,
The transcriber demo recognizes only digits,see the .gram file.Try wav files containing only digits and see it works well.
If you want to make it recognize words you should change the config file...it is using the TIDIGITS models change it to WSJ!!..also modify the .manifest file change it WSJ and in the .gram file add the appropriate words.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hello,
I'm a total newbie to Sphinx4. Just trying to get the Transcriber demo to work. I have a wav file which is PCM, 16kHz, 16Bit, Mono and the same thing as Stereo. The speech in it is just "this is a test". When I run the demo, all I get is stuff like
six six oh six four zero six
six two seven
or
five oh oh five
five oh oh
Can someone explain to me why this is happening and how I could get the demo to run? I'd appreciate any help.
My config has this in it:
<component name="streamDataSource" type="edu.cmu.sphinx.frontend.util.StreamDataSource"> <property name="sampleRate" value="16000"/> <property name="bitsPerSample" value="16"/> <property name="bigEndianData" value="false"/> <property name="signedData" value="true"/> <property name="bytesPerRead" value="320"/> </component>Hello,
The transcriber demo recognizes only digits,see the .gram file.Try wav files containing only digits and see it works well.
If you want to make it recognize words you should change the config file...it is using the TIDIGITS models change it to WSJ!!..also modify the .manifest file change it WSJ and in the .gram file add the appropriate words.