LumenVox announces v6.0.

Speech recognition technologies provider LumenVox has announced version 6.0 of the company's Speech Recognition Engine and Speech Platform. The Speech Recognition Engine, which performs speech recognition on audio data from any audio source, is speaker- and hardware-independent. The Engine is

The Speech Recognition Engine introduces new supported features, including the following: server-side grammars, which offers clients "a more efficient way to support large grammars, by pre-loading grammars onto the server"; n-best results from the Speech Engine, for allowance of flexible error handling; and MRCP (Media Resource Control Protocol) support, which allows developers to work with a common language for seamless management of diverse media resources (e.g., text-to-speech, voice verification, etc.).

Both the n-best results, wherein the engine can confirm several likely results instead of only the top-scoring result, and the MRCP support, according to the company, "further establish LumenVox as one of the most viable vendors to provide a high-quality, robust speech recognition engine."

In addition to the premiered new features, LumenVox also offers further improvements to its Speech Recognition Engine and Speech Platform. Included improvements offered: an improved Out-of-Vocabulary filter; barge-in and end-of-speech detection; confidence scoring; and increased speech recognizer search speedup.

www.lumenvox.com