W3C home > Mailing lists > Public > www-voice@w3.org > July to September 2003

RE: TTS for Semitic languages

From: Dave Raggett <dsr@w3.org>
Date: Fri, 8 Aug 2003 14:02:28 +0100 (BST)
To: David.Pawson@rnib.org.uk
Cc: www-voice@w3.org
Message-ID: <Pine.LNX.4.53.0308081357570.1056@localhost.localdomain>

One data point is the claim of 98% accuracy for unvoweled
Hewbrew text to speech (Kolan), as given in:

http://flrc.mitre.org/reports/All_Products_LID_TYPE.pl?TYPE=SPEECH

p.s. the melingo website itself claims 97% accuracy.

 Dave Raggett <dsr@w3.org>  W3C lead for voice and multimodal.
 http://www.w3.org/People/Raggett +44 1225 866240 (or 867351)


On Fri, 8 Aug 2003 David.Pawson@rnib.org.uk wrote:

>
> > TTS engines are designed to automatically insert vowels
> All semitic tts engines Dave? Not from what we are told by
> a lady from Israel.
>
> The question was intended to find out how good such insertion is,
> if anyone has experience of same.
>
> We've unvalidated information that the success rate is as low as 60%.
>
> regards DaveP
>
>
>
>
>
> > > I was discussing TTS for Semitic languages with the WAI
> > group the other
> > > day, and we had some general questions that I was hoping people with
> > > experience in the area might have be able to offer some input.
> > >
> > > The questions center on experiences when rendering text with versus
> > > without vowels.  A human would most likely not have problems reading
> > > something written whether or not it contained any vowels
> > (assuming the
> > > person spoke the language).  How does a TTS engine do
> > though?  Is the
> > > output reasonable when no vowels are present in the input?
> > I'm not sure
> > > how to qualify reasonable - either "the output is
> > understandable", or
> > > possibly something more quantitative, possibly "60% of the words are
> > > pronounced correctly".
> > >
> > > Thanks,
> > > Jeff
> > >
> >
>
> -
>
> NOTICE: The information contained in this email and any attachments is
> confidential and may be legally privileged. If you are not the
> intended recipient you are hereby notified that you must not use,
> disclose, distribute, copy, print or rely on this email's content. If
> you are not the intended recipient, please notify the sender
> immediately and then delete the email and any attachments from your
> system.
>
> RNIB has made strenuous efforts to ensure that emails and any
> attachments generated by its staff are free from viruses. However, it
> cannot accept any responsibility for any viruses which are
> transmitted. We therefore recommend you scan all attachments.
>
> Please note that the statements and views expressed in this email
> and any attachments are those of the author and do not necessarily
> represent those of RNIB.
>
> RNIB Registered Charity Number: 226227
>
> Website: http://www.rnib.org.uk
>
Received on Friday, 8 August 2003 09:02:23 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 30 October 2006 12:48:58 GMT