julius-speech / grammar-kit Goto Github PK

I'd like to generate some sample sentences from my grammar, but every time I use the 'generate' binary located at bin/linux/generate I just get the message:

grammar-kit/SampleGrammars$ ../bin/linux/generate fruit
-bash: ../bin/linux/generate: No such file or directory

The actual file ../bin/linux/generate does exist and I made it executable, so I assume the error message is the program itself reporting that it is unable to find some file, but I have no idea what is missing. I don't have a "generate" command anywhere on my path, and I did install both Julius and grammar-kit from source.

Thanks!

Hello,

I am trying to understand how to use julius. Unfortunately, I cannot find any practical examples in the documentation or on the internet demonstrating usage. I am attempting to run a server that will receive audio streams and return the recognition results in realtime.

I finally figured out how to run a server using adinnet (and the example scripts in dictation-kit) like so:

./run-linux-dnn.sh -input adinnet

Note of course that ,/run-linux-dnn.sh can be any julius command. I'm just using the example script in dictation-kit here, since I don't really understand how to fully configure julius yet.

This will parse Japanese text and accept streams given to it by adintool (which was difficult to figure out, as I could not find any examples of how to use it either. For anyone who stumbles on this after hours upon hours of searching like I did, use it something like this to send files to a server: echo "file.wav" | adintool -in file -out adinnet -server localhost)
However, the above command will not send results back to the client. From what I understand, I need to use the -module option, but it does not appear to work when used in combination with adinnet. IE, if I type the command:

./run-linux-dnn.sh -input adinnet -module

When I start the server this way, it nolonger acts as a server and accepts connections. It just blocks the server, it seems.

How am I supposed to use this option, and can you give a super simple, concrete example? Do I need to do something on the client side as well?

Note: I have read this thread and even looked through the code that is recommended as a reference, but I do not understand it. It's way above my head: julius-speech/julius#2