San Diego, CA (MAY 6, 2002) — LumenVox, a provider of speech recognition technologies, unveiled version 2.5 of their Speech Platform , an updated version of their product that is setting the new standard for price, quality and flexibility within the speech recognition market. V2.5 introduces Custom Actions Programming Interface (CAPI).
CAPI allows C++ developers to write their own dynamic link library (dll), to receive control over the call flow. The application will notify the specified function in the Custom Action dll with specific events. The dll can gather information or alter the behavior of the call, i.e. add audio, make additions to grammars, evaluate previous recognition, direct the caller to a new module, interface with databases or even transfer the caller. With this new flexibility, the Speech Platform can fill most telephony needs.
This feature is being added to the already feature-rich Platform. Other features include: natural language, which allows callers to speak more naturally by combining a pool of "non-critical phrases" with secondary items for further clarification; multi-level, that allows for a series of questions to be asked that is dependent on the question before; and overview, that allows for a long introductory or explanatory sound file, that can be interrupted with extended explanations, then returns to the original sound file, where the caller had left off.
Completely free single-user systems are available for download from the LumenVox website, www.LumenVox.com.
The new Speech Platform V2.5 includes LumenVox's Speech Recognition Engine as the foundation. The Speech Engine has a low level API set that is order independent and an automatic out-of-vocabulary filter and is also available for download from the LumenVox website. The Engine will automatically produce pronunciations for any word not currently in the 120,000 plus word dictionary.
© 2017 LumenVox, LLC. All rights reserved.