Hello,
I need an information about a possible solution for replacement of our old Speech Recognition System.
The function that it currently performs is described in an example below:
User dials the number ####
The system asking : Please say the name
User says a name
The system provides the extention of required name and dials the extention.
Our phone system is Alcatel based. The requirement for a Speech recognition is around 2000 entries.
The main goal is to replace an old system based on Wnidows 2000 (12 phone ports) but still working fine with a reliable and cost effective solution.
What components (software and hardware) should be in place for accomplishing the above scenario?
I guess dialer part could be done by Asterisk. What about actual Speech recognition?
We got a quotation from one supplier… it’s really crazy price
So looking for a realistic solution.
You can use too the Google speech API. zaf.github.io/asterisk-speech-recog/ I developed something similar of what you described above using this API and it is working fine.
Its not recommendable to use the API engine to sell or use in production servers, If you read carefully you are incurring in a violation of TOS also the google engine is not intended to commercial use since you aren’t paying license fees, and don’t be surprised if in the future google apply charges for violating their code. Besides that the I guess the API is limited to few minutes.
Honestly you got the reason navaismo ,And that is good point to take under consideration. but if someone would like to give a try with the Google API , i absolutely recommended under his own risk.
