Speech Recognition Engines

Flat
RE: Julius Speech Recognition
User: Gururaj
Date: 1/1/2017 11:26 pm
Views: 9396
Rating: 0

Hello everyone,

i am using julius speech recognition for my application. i have one doubt regarding julius. i have downloaded the latest version and was successful in using its lib and making it work. the problem i am facing is..once the app starts and i call the voice recognition function in my application...it takes the input from mic and displays whatever is said in the mic. but the function still continues to do so again and againn. The control will never come out of that function. here i am facing problem since the control is not returning back i am not able to proceed further. what i want is once the engine gets input from mic it should rcognize and stop there..which i tried to do by deleting the callback function but was unsuccessful.

Can anyone please guide me in this matter, what i need to do to get the desired output. it will be helpful for me.

Thanks in advance.

 

--- (Edited on 1/1/2017 11:26 pm [GMT-0600] by Gururaj) ---

RE: Julius Speech Recognition
User: colbec
Date: 1/2/2017 2:51 am
Views: 1
Rating: 0

You have a couple of choices: first to use the Julius -input control to get the sound data from a list of files (see the .jconf sample file), so that when the list (even if only length one) is exhausted then Julius stops. It is quite easy to record the voice input to a file and then feed the file into Julius. Second you can put a dialog manager in control. If you need more information on what a dialog manager does there are many posts on this forum on that subject accessible by a search.

The basic function of Julius is to start up and then keep on decoding input. When you get more experience you can run Julius as a server, and then tell the server to respond, not respond or shut down as required. It's more efficient than having Julius start and stop all the time.

--- (Edited on 2017-01-02 3:51 am [GMT-0500] by colbec) ---

RE: Julius Speech Recognition
User: Gururaj
Date: 1/2/2017 3:02 am
Views: 3
Rating: 0

Thank you for your reply...

As said by you i am recording voice input to a file and then feed the file into julius. but according to my requirement i want julius to recognize a command and stop there and again start when i want to start it, which i am unable to do.

so is there any way to do like this, as i am new to julius. please guide me further in this matter. Also i will check with the dialog maager.

Thank

--- (Edited on 1/2/2017 3:02 am [GMT-0600] by Gururaj) ---

RE: Julius Speech Recognition
User: colbec
Date: 1/2/2017 3:27 am
Views: 1
Rating: 0

From https://julius.osdn.jp/juliusbook/en/desc_adin.html#id2527780:

"Option -input rawfile tells Julius to read an audio input from file. You can give a file name to be processed to the standard input of Julius. Multiple files can be processed one by one by listing the file names to a text file and specify it by -filelist"

--- (Edited on 2017-01-02 4:27 am [GMT-0500] by colbec) ---

RE: Julius Speech Recognition
User: Gururaj
Date: 1/2/2017 3:31 am
Views: 3
Rating: 0

i checked with the dialog manager, Simon is used as frontend for julius, but I am using julius with Qt, i want to make changes in julius code so that it can recognize only one command at a time. Please suggest me how to do this in julius. Thank

--- (Edited on 1/2/2017 3:31 am [GMT-0600] by Gururaj) ---

RE: Julius Speech Recognition
User: Gururaj
Date: 1/2/2017 3:44 am
Views: 0
Rating: 0

I am using live microphone for giving inputs.(Sorry was confused with files).

--- (Edited on 1/2/2017 3:44 am [GMT-0600] by Gururaj) ---

RE: Julius Speech Recognition
User: colbec
Date: 1/2/2017 5:34 am
Views: 1
Rating: 1

It's not clear to me why anyone would want or need to do this, but in any case probably the best place to explain your situation is at the Julius github site https://github.com/julius-speech/julius/issues where you will meet all sorts of Julius specialists.

--- (Edited on 2017-01-02 6:34 am [GMT-0500] by colbec) ---

RE: Julius Speech Recognition
User: Gururaj
Date: 1/2/2017 5:38 am
Views: 45
Rating: 0

ok thanks for your reply.

--- (Edited on 1/2/2017 5:38 am [GMT-0600] by Gururaj) ---

RE: Julius Speech Recognition
User: Gururaj
Date: 1/4/2017 5:42 am
Views: 119
Rating: 0

Hey i asked in the github regarding the julius problem no one replied on that thread...the problem still persists and i am not able to proceed further. can anyone please help me over this.

--- (Edited on 1/4/2017 5:42 am [GMT-0600] by Gururaj) ---

RE: Julius Speech Recognition
User: colbec
Date: 1/7/2017 8:08 am
Views: 203
Rating: 1

Well, I guess the silence is meaningful too. Julius is like a car, it is designed to start, go and keep going until you tell it to stop. You just have to learn the controls. You can go to a mechanic and ask him to take half the gears out of the gearbox but he is likely to say "Why don't you just press the brake?"

On page http://www.voxforge.org/home/dev/acousticmodels/linux/test/htk--julius/julius Ken gives some information on testing an audio model with Julius. Look at step 4. It will give you some strong hints how to process just one file and stop.

--- (Edited on 2017-01-07 9:08 am [GMT-0500] by colbec) ---

PreviousNext