Hi
I am using pocketsphinx, for my college project Audio Video Tagging. In this I
need to search an audio or video file , based on the content being said in the
video. I am using pocketsphinx toolkit to get the text of audio video file,
and then search through the text. Actually Now I have to do exhaustive testing
and automate the process to list all the words which might not be found in the
dictionary. So I have taken 3 Lakhs words from Oxford Dictionary, splitted it
for 5000 words each, and generated audio file of those words. But this
generated wav file, when given to pocketsphinx is giving 0% accuracy, hardly 1
word gets detected. I am using english US dictionary only.
Please help.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Pocketsphinx accuracy is not quite related to your issues. You should rethink
algorithm you want to use to detect keywords in audio. Plain pocketsphinx is
not really usable for that task.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Kindly specify which accuracy you mention.
Because : conversion of original wav to text is working perfectly for us.
But : conversion -> text -> wav -> pocketsphinx -> text , is not working for
us.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi
I am using pocketsphinx, for my college project Audio Video Tagging. In this I
need to search an audio or video file , based on the content being said in the
video. I am using pocketsphinx toolkit to get the text of audio video file,
and then search through the text. Actually Now I have to do exhaustive testing
and automate the process to list all the words which might not be found in the
dictionary. So I have taken 3 Lakhs words from Oxford Dictionary, splitted it
for 5000 words each, and generated audio file of those words. But this
generated wav file, when given to pocketsphinx is giving 0% accuracy, hardly 1
word gets detected. I am using english US dictionary only.
Please help.
I use pocketsphinx_batch to convert the audio wav to text
Pocketsphinx accuracy is not quite related to your issues. You should rethink
algorithm you want to use to detect keywords in audio. Plain pocketsphinx is
not really usable for that task.
Kindly specify which accuracy you mention.
Because : conversion of original wav to text is working perfectly for us.
But : conversion -> text -> wav -> pocketsphinx -> text , is not working for
us.