We have created our dictionary containing 3,300+ words of our native language which is Tagalog. We also have created our own language model containing those words written in our dictionary. We are using the acoustic model provided by this link. The problem is, when we try to convert speech, it gives us wrong results. The results aren't even close to the words we spoke. When we say aba, it gives us ambag at buod amain alis. We are aware of it could be the background noise so we record our speech on a quiet place and speak our speech very well to receive an accurate result. Please enlighten us why is that happening? What could be the cause of such inaccuracy where the spoken words aren't even showing on the results?
Last edit: Randall Castillo 2016-10-27
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Command line, logs, data files of what specifically? Or where can I get those specifically? Sorry for the stupid questions but I really don't know where to get those.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Yes, that should give me more understanding of your problem, however, the information you provided is still not complete. You still need to provide logs and data files to reproduce your problem. The more information you provide the faster you get an answer.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
We have created our dictionary containing 3,300+ words of our native language which is Tagalog. We also have created our own language model containing those words written in our dictionary. We are using the acoustic model provided by this link. The problem is, when we try to convert speech, it gives us wrong results. The results aren't even close to the words we spoke. When we say aba, it gives us ambag at buod amain alis. We are aware of it could be the background noise so we record our speech on a quiet place and speak our speech very well to receive an accurate result. Please enlighten us why is that happening? What could be the cause of such inaccuracy where the spoken words aren't even showing on the results?
Last edit: Randall Castillo 2016-10-27
You need to provide command line, logs, data files required reproduce your problem in order to get help on this issue.
Command line, logs, data files of what specifically? Or where can I get those specifically? Sorry for the stupid questions but I really don't know where to get those.
Of the command you run to recognize "ambag at buod amain alis".
Is this the command line you're talking about, Mr. Nikolay?
Last edit: Randall Castillo 2016-10-31
Yes, that should give me more understanding of your problem, however, the information you provided is still not complete. You still need to provide logs and data files to reproduce your problem. The more information you provide the faster you get an answer.