The use of speech recognition technology can be used to enhance Interactive Voice Response (IVR), Call-Centre and Telephone information services. Speech recognition technology is particularly useful in the following situations:-
- Where there are too many menu options for DTMF to be a viable option. For example, requesting the share price of companies listed in the FTSE100 index.
- Where the user is likely to be using a DECT or mobile telephone. In these situations, it is difficult for callers to repeatedly move the phone away from the ear to press a button and then move it back again quickly to hear the subsequent prompt.
- Where there is a need to access a very large vocabulary. e.g. selecting an employee name from a large company directory in an automated auto-attendant deployment
eurovoice objects, with its optional speech recogntion add-on, supports the use of Loquendo's award-winning speech recognition engine via the industry standard MRCPv2 interface. MRCPv2 allows high volume, high availability systems to be deployed by the use of distributed speech recognition engines. Furthermore, Loquendo's speech recognition algorithms, in conjunction with MRCPv2 technology, avoids the need for expensive digital signal processing hardware to be deployed in the speech recognition servers.
In a typical call:-
- Incoming speech from the caller is streamed to a speech recognition engine, located on the LAN.
- Echo cancellation, within the IVR Server, stops any echo - as a result of the outgoing message - arriving at the speech recogniser. This allows applications to be deployed that make use of 'barge-in'. 'Barge-in' is the ability to speak a request before waiting for the outgoing message to finish, in a similar manner to IVR systems that allow DTMF type-ahead.
- The speech recogniser interprets the incoming speech and returns the result of the recognition attempt to the application, via eurovoice objects. The recogniser may be configured to interpret all the incoming speech, or to look for specific keywords.
- Candidate word and confidence level information is processed by the application and the appropriate actions taken.
- eurovoice objects allows fine control of the recogniser to maximise its performance, especially in non-ideal acoustical environments.
To read further technical information regarding eurovoice objects 3.0 speech recognition support, click here.
Business benefits:-
- Applications can be deployed which would not be practical without the use of speech input.
- Users may interact with services in a more natural means, resulting in faster access to information, thus reducing call times.
- Businesses deploying speech recognition technology are often seen as being 'high-tech' and 'forward-thinking' by customers and potential investors.