Click here to register.

Running Julius Live

1. First you need to create your Julius configuration file.  Copy this sample configuration file (sample.jconf) to you 'voxforge/manual' folder.  For details on the parameters contained in the Sample.jconf file, see the Juliusbook for more information.  The main parameters are shown below:

# VoxForge configurations:
-dfa sample.dfa     # finite state automaton grammar file
-v sample.dict      # pronunciation dictionary
-h hmm15/hmmdefs    # acoustic HMM (ascii or Julius binary)
-hlist tiedlist     # HMMList to map logical phone to physical
-smpFreq 16000	    # sampling rate (Hz)
-spmodel "sp"		    # name of a short-pause silence model
-multipath          # force enable MULTI-PATH model handling
-gprune safe        # Gaussian pruning method
-iwcd1 max          # Inter-word triphone approximation method
-iwsppenalty -70.0	# transition penalty for the appended sp models
-iwsp			          # append a skippable sp model at all word ends
-penalty1 5.0		    # word insertion penalty for grammar (pass1)
-penalty2 20.0	    # word insertion penalty for grammar (pass2)
-b2 200             # beam width on 2nd pass (#words)
-sb 200.0		        # score beam envelope threshold
-n 1                # num of sentences to find

# you may need to adjust your "-lv" value to prevent the recognizer inadvertently 
# recognizing non-speech sounds:
-lv 4000			# level threshold (0-32767)

# comment these out for debugging:
-logfile julius.log
-quiet

 

Make sure your Microphone volume is similar to when you created your audio files. Then run Julius with:

$julius-4.3.1 -input mic -C sample.jconf

 

 

Example of your Julius startup output.

The first 2-3 seconds of your speech will not be recognized - Julius adjusts its recognition levels (that is what the reference to their being "no CMN parameter is available on startup" is all about).  In addition, Julius will only recognize phrases from the grammar you created in Step 1.

Example of what the Julius recognition output looks like when I say "Phone Steve" into my microphone.

You should get fair recognition results.  To improve recognition, your Acoustic Model needs more audio training data.  You need to create new prompts, and record more speech audio files based on these prompts in order to create better acoustic models.  You can speed up the training process by using the Acoustic Model creation script in the How-to (i.e. How-to Create an Acoustic Model - using a script).

Comments

By kmaclean - 4/18/2008 - 1 Replies If you have a small grammar, the following things can help improve recognition performance: using a noise model, using word-based hmm with more states (rather than phone-based hmm), not using context-independent models

By ashkanmradi - 1/25/2020 - 1 Replies when i try the command:

By wildanzd - 8/6/2018 - 2 Replies hello, i'm just finished follow the tutorial and there is no problem until i try to run julius using terminal on windows 10. when i finished run the command,

By prithviraj - 5/3/2018 Hi I am finding the following problem while running julius-4.3.1.

By prithviraj - 5/3/2018 Hi I am finding the following problem while running julius-4.3.1.

By Sudipto - 10/1/2016 - 1 Replies when it is <<please speak>> I try to record "phone steve".I don't know what is happening.it just shows the warning "Too long input(>320000 samples)" in the output.Then again <<please speak>> message are shown.

By Sharma - 7/7/2016 - 2 Replies C:\Users\Sharma\voxforge\tutorial>julius-4.3.1 -input mic -C Sample.jconf

By Shama - 6/3/2016 - 3 Replies julia -input mic -C test.jconf

By vimal01 - 11/29/2015 Hi ken, I made this demo for few Hindi words. I used Ubuntu 14.0.04. I While run this demo it gives me following output. Can you please tell me why it is giving "Search Failed" error? I am unable to apprehend these errors. Are they grammar related issues or anything else! Please Suggest. Thanks Attached Output Below: STAT: include config: sample.jconf <<< please speak >>>pass1_best: <s> ANURODHA KE <search failed> <<< please speak >>>pass1_best: <s> ANURODHA KE <search failed> <<< please speak >>><<< please speak >>><<< please speak >>>pass1_best: <s> ANURODHA UNASE sentence1: <s> ANURODHA UNASE </s> <<< please speak >>>pass1_best: <s> <search failed> pass1_best: <s> ANURODHA KE <search failed> pass1_best: <s> sentence1: <s> KARENGE MEET </s> pass1_best: <s> ANURODHA KE <search failed> pass1_best: <s> ANURODHA KE <search failed> pass1_best: <s> ANURODHA KE <search failed> pass1_best: <s> ANURODHA KE <search failed>

By Ivona - 8/22/2014 Hello,

By tedja - 5/19/2014 root@raspberrypi:/# julian -input mic -C /home/voxforge/manual/julian.jconf include config: /home/voxforge/manual/julian.jconf /home/voxforge/manual/julian.jconf: wrong argument: -iwsp Try `-help' for more information. Terminated

By angga - 5/19/2014 i've got problem when running julius

By Oviya - 2/22/2014 - 1 Replies ### read waveform input <<< please speak >>>adin_mic_read: no data fragment after 300 msec? adin_mic_read: no data fragment after 300 msec? adin_mic_read: no data fragment after 300 msec? ...

By Sam - 3/31/2013 - 6 Replies Tried Julius 4.2 version. Ran it with the same julian.jconf file as in the tutorial.

By Sam - 3/29/2013 - 12 Replies Hi,

By vjdtao - 8/30/2012 - 1 Replies

By Nimrod - 8/23/2012 - 1 Replies Hello,

By seyaw - 12/15/2010 - 2 Replies I have followed the thread at

By seyaw - 12/15/2010 - 3 Replies Sorry the previous error was due to type mismatch. Now I have the following problem

By seyaw - 12/15/2010 - 1 Replies I have reached the last step. But when I run "julian -input mic -C julian.jconf" it doesn't run. It displays a message

By georgi - 11/19/2010 - 2 Replies Dear Ken, I have trained hmm models using HTK.

By aspirant - 11/15/2010 - 1 Replies hi...

By Yo - 4/9/2010 - 4 Replies I've successfully followed the instructions and installed Julian.

By dina - 2/23/2010 - 4 Replies please help me, I need to know how to tied acoustic modle with java application

By fci - 2/19/2010 Thanks for this tutorial ,i tried to execute acoustic model but i finished it but i have remark

By Cecifia - 12/23/2009 - 5 Replies Hi,

By hi - 11/21/2009 - 1 Replies whe i tried to excute the command julian -input mic -C julian.jconf,nothing is displayed.That means no error,warning as well as correct messages at all. So, what is the problem.

By zvaracik - 5/15/2009 - 3 Replies Hi, I have a problem with run a Julian. When I run it, I get this message:

By Visitor - 3/20/2009 - 4 Replies Hi,

By Nick - 3/10/2009 - 4 Replies Hi,