Automatic Speech Recognition
Automatic Speech Recognition (ASR) technology is used for building voice driven user interfaces for Service Provider and Enterprise applications. ASR provides an efficient and intuitive spoken alternative to touch-tone (DTMF) applications.
OmniVox3D Application Server Software's ASR Module integrates to several speech engines (Nuance/Scansoft's OSR and Lumenvox Speech Engine) enabling application developers to apply a speech driven user interface to existing applications in the language of their choice. Through the use of ASR-driven applications, Service Providers and Enterprises can improve workforce productivity and operational process efficiency. ASR creates value-add and differentiation for the services provided.
The ASR Module minimizes the expense of implementing voice driven applications by providing an intuitive GUI based Service Creation Environment as well as standards-based VXML creation. The flexibility of the ASR speech engine is fully maintained within OmniVox3D allowing application developers to take full advantage of features such as Confidence Ratings and N-Best results, which are returned from the speech engine up to the application management layer.
ASR functionality is achieved by high level programming of the ASR commands available in OmniView®, OmniVox3D's Service Creation Environment, combined with the building of ASR grammars that contain the universe of words to be recognized at each specific speech utterance.
The integration of the speech engines, combined with the stability, flexibility and powerful Service Creation Environment included in OmniVox3D provides application developers a rich feature set of functions allowing advanced application development. The provided technological features include:
- Utterance recording detected by the speech recognition engines into a new file, or appending to an existing voice file.
- Barge-in capability, allowing expert callers to interrupt prompt messages by speaking responses at any time during the call.
- "N-Best results" which is a feature provided by the speech engine that returns the number of times a word or sentence detected in the utterance spoken, from Grammar file to the OmniVox3D application variables.
- Confidence rating (%), which is a rating returned by the speech engines informing the application of how accurate it thinks the utterance matches what is configured in the Grammar files.
- All results returned from the speech engines can be assigned to OmniVox3D application variables so as to be used further in the application and/or to be stored into a database or file.
ASR technology is widely applied to Service Provider and Enterprise solutions and brings process-optimization, cost reductions and value-add through applications such as:
- Automated Attendant
- Call Center Routing
- Call Center Integration
- Voice Activated Dialing
- Unified Messaging
- Virtual Assistant
- Directory Services
- Collect Calling
- Prepaid Calling
- Outbound Telemarketing
- ... many more
The ASR Module for OmniVox3D allows Service Providers and Enterprises to rapidly and cost effectively enhance the caller experience by adding speech capabilities to customer interaction and workforce applications.
The rich feature set of ASR commands available through OmniView provide a friendly GUI based application development environment for speech driven applications. Depending on capacity and network requirements, ASR may be implemented in a centralized or distributed configuration.
The combination of OmniVox3D Application Server Software with speech engines, as well as other modules and technologies, make APEX the solution of choice for those organizations requiring deployment of flexible, standards-based and true open architecture voice platforms.
|
