Languages

Read Prompts and Submit Recordings

 

Download Java Webstart jnlp file

 

Then double click the file to start the Speech Submission app (you need Java installed).

Before you Begin Recording Prompts

Turn off any other audio programs on your PC (i.e. music players, audio editors, etc.)

Position your mike so it does not pick up your breathing.  While recording, try to minimize any non-speech noises (i.e. exhaling, taking a breath, lip smacking, ...) or background noise. 

The black box near the bottom of your screen will display the audio waveform of your recording. Please record a test recording to ensure your microphone volume is not too loud or too soft.

 

Recording Prompts 

For each prompt line, please record your speech as follows:

  1. click the Record button,
  2. pause for half a second,
  3. Read the corresponding prompt sentence,
  4. pause for half a second, and then
  5. click the the Stop button.

If you make a mistake, click Record again to re-record your prompt.

Please do not read the punctuation marks out loud.

Once you have completed recording all ten prompts the Upload button will activate.  Click the Upload button to upload your entire submission to the VoxForge repository as a single zip file.

Repeat the process (multiple submissions are encouraged!) 

Comments

By chip_maker - 12/25/2014 Hi, It appears the Java applet has some issues (as of today Dec 25th) even after I installed latest java plug in on my laptop (Windows 8.1 all updates done). I had to minimize Internet explorer and maximize it for the Java applet to show properly.The other method appeared cumbersome and I could not figure out easily. Hence I still somehow uploaded few speech recordings using Java applet. May be one more method using some other software would be helpful. I am fine downloading and installing a recording software like this applet on laptop. Also it appears this comment has to be one paragraph. Multiple paragraphs are not appearing. Regards, Kal Gandikota

By chip_maker - 12/25/2014 Hi,

By Rex7957 - 3/4/2014 - 1 Replies I use stereo audio files to extract MFCC features with HTK. I don't know how HTK processes stereo audio files. I want to use my features which are produced with Matlab,and then compare the speech recognition rate with MFCC. So I need to know the details how HTK works. Does HTK works like this in Matlab code? a=wavread(test.wav);

By Bogdan Bivolaru - 1/26/2014 All modern browsers can nowadays record audio from user's microphone:

By Visitor - 12/27/2013 - 1 Replies I Got the follwoing error while recording sound

By Joey - 8/11/2013 +1 to not using Java for submitting speech. Due to my company's security policy, I can't even use Java on my laptop. Even Flash, terrible as it is, would be better.

By konaya - 6/19/2013 - 1 Replies Using films with subtitles, like someone suggested, is obviously destined to fail. There's too much noise, and the subtitle does not match the actual speech word-for-word.

By Behrooz - 5/15/2013 - 1 Replies I wonder why cannot you use movies + their subtitles for training. Subtitles give you the exact time slot for each sentense.

By Anand - 1/24/2013 Hello friends,

By veritanuda - 1/21/2013 - 6 Replies Can we please update the recording process to not use Java. It is 2013 and Java has had so many exploits of late it is no longer recommended to disable it in browsers but actually disabled automatically by Mozilla in Firefox.

By bhupaesh Ghai - 6/30/2012 - 2 Replies I have gone through all the instructions for installing Cygwin and then HTK on it.(As per 2006 Voxforge documentation). When i did first time, i didn't face any problem. But due to some problems, this time, i am facing the following problem:

By brindafella - 1/3/2012 - 1 Replies Okay, so I made sure that my Java was more up-to-date than the version number required by the page; I updated it further, anyway.

By ssb - 12/24/2011 - 1 Replies I record my voice OK, but when I press Play, it plays only a small segment of what I said, then it skips to near the end. So I didn't upload it.

By alejandro - 12/3/2011 - 2 Replies en el programa no aparece el idioma castellano y las frases se encuentran en ingles es posible podrian agregarlo por favor o alguna idea de como contribuir para dar desarrollo al idioma castellano ya sea traducir las frases o desarrollar una verision para castellano

By MikeJ - 10/30/2011 - 5 Replies The work queue listed on this page has apparently not been processed in well over a year. The Metrics page has not been updated in 5 months. Am I wasting my time submitting recordings?

By ssanfrat - 10/8/2011 It would be more forgiving if you added a "restart" button for when the volunteer is uploading the recordings. I had made a set of recordings, then press upload, but needed to refresh my internet connection. The upload was stalled, so I tried to reload the web page, and the result was a previous page that I had used. That's a problem. So to correct it I had to navigate back to your home page, then to the page to submit recordings, and the java applet ddin't see the old recordings anymore, and so I had to re-create them all, a real nusance.

By Catbells - 8/2/2011 - 1 Replies It seems a bit strange that every time I log in I have to say that I'm a female, an adult and my accent is British. After all, these details don't change very often . Could they be automatically inserted?

By Derrick - 4/29/2011 - 1 Replies Hi, I know some people in New Zealand who are interested in contributing, but bandwidth there is very expensive and they're uncertain how much is being used when they upload a submission. Is the audio compressed, losslessly or otherwise, before upload? What's a typical average size? It'd also be a big help if the "results" page would describe how much bandwidth was ultimately used and how much their user has used so far, so they can track their overall usage and the amount their contribution is costing them. This would also be helpful for other bandwidth-starved nations. Thanks!

By kmaclean - 3/20/2011 - 3 Replies John asks:

By Catbells - 12/24/2010 - 1 Replies Twice the command below appeared with the bottom row of pixels omitted so I can't be sure that the spaces are not dots and vice versa: -

By Airwings - 12/16/2010 For me Firefox was working perfectly. The upload/ record / play commands worked perfectly; i don't know what you were complaining about guys...

By utilator - 11/20/2010 - 1 Replies simply could you make a link so that we upload a wav file, as simply as possible , because firefox is never working...

By dancor - 10/4/2010 - 4 Replies I second the idea to save user/gender/age/dialect/mic settings for multiple recording sessions, via a cookie or otherwise. I am even logged in to my account, perhaps the information could be associated with my account and then passed into the applet.

By BGlassman - 3/8/2010 - 1 Replies After I record a prompt, it plays back at a very high rate of speed. I'm on a MacBook Pro with a Plantronics Bluetooth headset. I can't imagine these recordings will be of any use to you, but if you have any insight into why it's happening, maybe I can produce something you can use. Sorry not to have been able to help out. It's a most worthy project.

By kmaclean - 1/25/2010 see this post: Voxforge java speech applet and Firefox 3.6

By Mariane - 9/6/2009 - 4 Replies Could you please give a link to the file listing the prompts? And to the upload function? I can't see anything of the kind.

By Daniel Teague - 3/22/2009 - 2 Replies I have set pulse audio to record from my USB mic (Sennheiser), and Sound Recorder in Ubuntu can record and play back my speech. However, the applet above does not seem to record the prompts I read in. Also, the Java applet during playback doesn't have any spikes.

By Tom - 1/7/2009 - 1 Replies I get an error in the java console on uploading:

By Visitor - 12/30/2008 - 1 Replies Could you possibly make the user profile (i.e. gender,dialect, language) a forced option?

By awalsh - 11/29/2008 - 4 Replies Hi,