Inference Group
.
.




Search :
.

Downloading Dasher

Dasher (versions 3 and above) is available for general use under the conditions of the GNU General Public Licence. You should be aware of the terms and conditions of this licence, which is also contained in the distribution itself.

Please fill in the form below to download Dasher. It's not required, but we'd be grateful if you filled it in. This information is purely for our own Dasher-related use and will not be disclosed to any other party. Neither will we, in general, send any mail to any address given here. If you want to keep in touch with Dasher developments after downloading, please see this page for mailing list information.


Dasher download - step 1 of 2
Your Name:  
email address:  
country:  
first language:  
Do you have a disability relevant to Dasher?  
Age: 
Where did you hear about Dasher? 
select the packages you require

Dasher

Stable Release (4.11)

 

Microsoft Windows

  version 4.11 (Windows XP, Vista, Windows 7)
  version 4.6.2 (Windows 2000)
  32 bit Windows XP version is confirmed to work under 64 bit Windows 7.
  Please manually uninstall any versions prior to 4.4.2 before installing version 4.6.0.
  For speech production, the Microsoft Speech API is required. This is included in Windows XP or may be obtained from http://www.microsoft.com/speech/download/sdk51/
 

Mac OS X

  version 4.11
  expat 2.0.1 necessary for MacOS X less than 10.5
 

Source Code

  version 4.11 (.tar.bz2)
  version 4.11 (.tar.gz)
  Android beta (.tgz)

Speech Dasher

  Speech Dasher is a post-processor for your computer's speech-recognition system (eg Microsoft or Dragon) (More information)

Development Release

 

Microsoft Windows

  Speech Dasher prototype (XP)

Dasher For Mobile Devices

Early/beta release based on 4.11+

  version 1.0 (android)

Old Release (2.0.3)

 

Pocket PC (list of pocket PCs)

  version 2.0.3

Additional Files

Training texts (for use with version 3 or 4)

 

European languages training texts

  Albanian training text (by Alban Rrustemi) (317K)
  Armenian training text (by kostas savvidis) (K)
  Czech COMBINING training text (456K)
  Czech training text (456K)
  Danish training text (provided by Ole Winther) (600K)
  Dutch training text (provided by Peter Baert) (1.1M)
  English training text (already included in main Dasher distribution) (300K)
  English training text Lower Case (300K)
  Esperanto training text (provided by Edmund Grimley Evans) (500K)
  Finnish training text (by Leena Krohn) (460K)
  French training text (by Gilles Brodeur and Peter Baert) (640K)
  French COMBINING training text (by Gilles Brodeur and Peter Baert) (640K)
  German training text (by Michael Muhlhaus) (919K)
  Greek training text (by Pantelis Nasikas) (496K)
  Hungarian training text ((c) Hungarian Electronic Library, for use in non-profit applications only) (500K)
  Italian training text (provided by Matteo Fasanella) (322K)
  Lithuanian training text (provided by aleksandras steponavicius) (301K)
  Polish training text (by Blazej L. Popielski) (1M)
  Portuguese (Brazilian) COMBINING training text (by Martin Whittle and CETENFolha) (656K)
  Portuguese (Brazilian) training text (by Martin Whittle and CETENFolha) (656K)
  Russian training text (by Alex Ziranov) (589K)
  Russian Latin training text (by Nikon Rasumov) (300K)
  Spanish training text (by Andrea Prado) (1.6M)
  Spanish COMBINING training text (by Andrea Prado) (1.6M)
  Swedish training text (thanks, Johan Bouvin) (413K)
  Turkish training text (thanks, Bulent Celasun) (200K)
 

Asian languages training texts

  Japanese Hiragana(60) training text (130K)
  Japanese Hiragana(83) training text (130K)
  Thai training text (thanks, Ratthachat Chatpatanasiri and Jung Dalglish) (1M)
  Mongolian training text (271K)
 

Indian languages training texts - generously provided by the Emille project

  Assamese (688K)
  Bengali (816K)
  Gujarati (631K)
  Hindi (800K)
  Kannada (894K)
  Kashmiri (491K)
  Malayalam (956K)
  Marathi (1M)
  Oriya (581K)
  Punjabi (gurmukhi) (746K)
  Punjabi (indopersoarabic) (822K)
  Sinhala (649K)
  Tamil (581K)
  Telugu (800K)
  Urdu (653K)
 

African languages training texts (South African languages kindly provided by De Schryver, Gilles-Maurice. 2002)

  Swahili training text (provided by Arvi Hurskainen) (300K)
  Afrikaans training text (1M)
  Malagasy training text (82K)
  Ndebele training text (1M)
  Sepedi training text (1M)
  Sesotho training text (1M)
  Swati training text (918K)
  Tsonga training text (1.2M)
  Tswana training text (1M)
  Venda training text (548K)
  Xhosa training text (933K)
  Zulu training text (855K)
 

Middle-eastern languages

  Arabic training text (3M) provided by Yasser H. Abdel-Haleem
  Hebrew training text (183K)
  Persian training text (1M) provided by Behdad

Scripts for converting files

  Perl program converts Japanese Shift-JIS to UTF8
  Perl program ensures Hiragana(UTF8) files are correct for Hiragana(60) alphabet
  Perl program ensures Hiragana(UTF8) files are correct for Hiragana(83) alphabet

Other files

  Abc (Music) training text (2M)
 

If this form does not work, please let us know.

The Inference Group is supported by the Gatsby Foundation
and by a partnership award from IBM Zurich Research Laboratory
David MacKay
Site last modified Fri Oct 1 10:33:19 BST 2010