Commercial speech recognition systems like Speechworks and Nuance have drop-in modules for different speech recognition tasks.
For example, Speechworks has:
CreditCardNumber
CreditCardExpiration
SocialSecurity
Time
NaturalNumbers
YesNo
ItemList
VoiceMenu
ContinuousDigits
AlphanumericString
TelephoneNumber
ZipCode
Spelling
Currency
Date
Name
CustomContext
and Nuance has:
Active Confirmation
Audio Recording
Browable List
Browsable Selection
Browsable Action List
Confirm and Correct
Credit Card Information
Currency
Date
Dialog Manager
Digit String
Menu
Quantity
Sectioned Digit String
Silent Confirmation
Social Security Number
Speaker Verification
Speaker Verification Enrollment
Telephone Number
Time
Universal
Voice Enrolled Phrase
Yes/No
Zip Code
Not withstanding that they would probably also sell their demo modules, like stock quotes and the like.
Does the Sphinx architecture support this type of drop-in ability? Are there plans to develop such modules? If not, is there interest in starting such development?
Philip Trauring
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
There is no simple way to do this with Sphinx yet, but we met and talked about it today and i think it was unanimous that we liked this idea very much. Now we have to figure out how to expose the functionality easily -- whether as a collection of finite grammars or nets, with visualization tools, or what. So I guess I'll go for the last option and take "Yes there's interest in starting such development!"
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Commercial speech recognition systems like Speechworks and Nuance have drop-in modules for different speech recognition tasks.
For example, Speechworks has:
CreditCardNumber
CreditCardExpiration
SocialSecurity
Time
NaturalNumbers
YesNo
ItemList
VoiceMenu
ContinuousDigits
AlphanumericString
TelephoneNumber
ZipCode
Spelling
Currency
Date
Name
CustomContext
and Nuance has:
Active Confirmation
Audio Recording
Browable List
Browsable Selection
Browsable Action List
Confirm and Correct
Credit Card Information
Currency
Date
Dialog Manager
Digit String
Menu
Quantity
Sectioned Digit String
Silent Confirmation
Social Security Number
Speaker Verification
Speaker Verification Enrollment
Telephone Number
Time
Universal
Voice Enrolled Phrase
Yes/No
Zip Code
Not withstanding that they would probably also sell their demo modules, like stock quotes and the like.
Does the Sphinx architecture support this type of drop-in ability? Are there plans to develop such modules? If not, is there interest in starting such development?
Philip Trauring
There is no simple way to do this with Sphinx yet, but we met and talked about it today and i think it was unanimous that we liked this idea very much. Now we have to figure out how to expose the functionality easily -- whether as a collection of finite grammars or nets, with visualization tools, or what. So I guess I'll go for the last option and take "Yes there's interest in starting such development!"