Win a copy of Design for the Mind this week in the Design forum!
  • Post Reply
  • Bookmark Topic Watch Topic
  • New Topic

further questions re: voice xml

 
Joel Carklin
Greenhorn
Posts: 28
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
Hi,
I am a jsp, java programmer and the voice xml giveaway caught my eye. I have briefly had a look at the voicexml.org site, but have not yet had the time to spend on it. I just have a couple of questions which will give me a better idea if it is what I think it is, and if voicexml can be used for what I want to use it for:
1. It appears that voiceXml is primarily used for speech recognition. is this correct? Or is it used for producing speech? or both
2. Is it language based, in terms of human languages I mean. i.e.
Can it be used to create speech recognition and text-to-speech in languages other than English. Like Zulu (for example). Or is this entirely dependant on the speech software using the Xml?
Basically, here in Africa the great 'digital divide' is so big it doesn't seem bridgable(?). One of the major problems facing disadvantaged communities here when it comes to technology (besides lack of electricity I mean) is illeteracy. Many millions cannot read or write, many cannot speak english. Is VoiceXml worth investigating as part of a solution to this, I guess is basically my question, with the idea of creating speech recognition and text-to-speech software that can be used here in the third world?
Thanks
Joel
 
Hema Menon
Ranch Hand
Posts: 569
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
Joel,
VoiceXML is neither for producing speech nor used for speech recognition. VoiceXML is an XML based language used to develop telephony applications. You can have speech recognition as a part of the voicexml applications. You need speech recognition system to enable this. VoiceXML 1.0 specifications can have a <dtmf> tag to take telephone inputs from the key pad of a telephone. Or you could have grammar specified for speech recognition or both.
VoiceXML is XML format with a specific DTD for validation. VoiceXML has <prompt> tags which can be used by text to speech engines to provide prompts to the user. Speech and Text to speech are an important part of VoiceXML.
HTH
Hema
 
Xiaofei-Jeff Tang
Greenhorn
Posts: 19
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
Hi Joel,
That's a very interesting idea. VoiceXML basically relies on, rather than uses, automatic speech recognition (ASR) and text-to-speech (TTS) technologies. The leading vendors of ASR and TTS engines all support many languages, e.g. Nuance supports the following languages:
Cantonese
Czech
Dutch
English.America
English.AusNZ
English.Singapore
English.SouthAfrica
English.UK
French
French.Canada
German
Greek
Italian
Japanese
Mandarin
Norwegian
Portuguese.Brazil
Spanish
Spanish.America
Swedish
Turkish
And Speechworks TTS engine supports:
US English
UK English
Castilian Spanish
Mexican Spanish
Canadian French
Standard French
German
Italian
Finnish
Brazilian Portuguese
Mandarin Chinese
Japanese
Korean
VoiceXML just speeds up the development of traditional DTMF-based telephony application and/or the new ASR and TTS based voice applications. So my opinion is yes, if you find some ASR/TTS vendors who can support the language of your interest, and if the overall cost of developing such a service makes sense, then VoiceXML can shorten the time thus save money to build such a service to serve the people there.
Regards, Jeff
Originally posted by Joel Carklin:
Hi,
I am a jsp, java programmer and the voice xml giveaway caught my eye. I have briefly had a look at the voicexml.org site, but have not yet had the time to spend on it. I just have a couple of questions which will give me a better idea if it is what I think it is, and if voicexml can be used for what I want to use it for:
1. It appears that voiceXml is primarily used for speech recognition. is this correct? Or is it used for producing speech? or both
2. Is it language based, in terms of human languages I mean. i.e.
Can it be used to create speech recognition and text-to-speech in languages other than English. Like Zulu (for example). Or is this entirely dependant on the speech software using the Xml?
Basically, here in Africa the great 'digital divide' is so big it doesn't seem bridgable(?). One of the major problems facing disadvantaged communities here when it comes to technology (besides lack of electricity I mean) is illeteracy. Many millions cannot read or write, many cannot speak english. Is VoiceXml worth investigating as part of a solution to this, I guess is basically my question, with the idea of creating speech recognition and text-to-speech software that can be used here in the third world?
Thanks
Joel

 
Xiaofei-Jeff Tang
Greenhorn
Posts: 19
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
Sorry, I meant "VoiceXML basically relies on, rather than creates (rather than uses ;-)), ASR and TTS technologies." Jeff
 
  • Post Reply
  • Bookmark Topic Watch Topic
  • New Topic