telisSpeech Recognizer

Latest News

Belgique : l’horaire des trains en français et Néerlandais avec teliSpeech

Based on market standards, the core of teliSpeech Recognizer satisfies four fundamental objectives: giving a high level of recognition performance, directly contributing to dialogue efficiency, and increasing the level of interactivity whilst allowing for spontaneous Natural Language expression.
This technological core is slotted into an infrastructure composed of four essential cornerstones: a distributed architecture, an administration framework, an interface with the telephony platforms and a Quality of Service Management environment.
Together, the core and the infrastructure, allow you to master the implementation of your voice services.

In order to accelerate the expansion of voice services, telisma has invested heavily in core technology and the quality of speech recognition. Whoever the user is, wherever they call from, the service will understand them better with teliSpeech:

its intelligent speech detector automatically adapts to the surrounding noise levels and manages the possible ambiguities between a pause and the end of speech.
an evolved noise filter together with the speech detector guarantees a high level of recognition, even for interaction with a particularly noisy environment, such as speech in a car or on a mobile.
the Noise Catcher marks and rejects the ambient or background noise (such as door slamming).

Dialogue efficiency mainly depends on dealing with the different possible forms of expression. Your service can offer richer and more fluid dialogues thanks to teliSpeech:

The eXtra Large Vocabulary technology can handle up to 2 million words in real time. This allows you, for example, to compile a directory of several hundreds of thousands of names in a few seconds
teliSpeech can recognise voice and DTMF. The DTMF sequences keyed in by users are interpreted in the same way as a voice, through ad hoc grammars.
The multi-language recognition engine can handle as many languages as you like, and even within a single call. In particular it can recognise pronunciation errors made by foreign speakers.

Response time for a service is an essential key to what users feel; it becomes critical when considering a voice service. Over and above user satisfaction, a reactive service also reduces the duration of a call, and therefore the associated costs. Your voice service will be more interactive and dynamic with teliSpeech.

TeliSpeech Recognizer is available in the following configurations:

operating system	compiler
Windows 2003 with Service Pack 2	Microsoft Visual C++ 6.x or 8.x
Linux Enterprise 4.0	GCC 3.4.3 + binutils 2.15

products

Latest News