Recognizer features
Automated speech recognition (ASR) is the foundation of an efficient contact center.
Nuance Recognizer is the software at the core of Nuance's contact-center automation solutions designed to work seamlessly in both on-premise and cloud-based environments. Built upon years of experience across six different product lines, Recognizer is available around the world in a large number of different languages.
Nuance Recognizer understands and interprets spoken and touchtone input to help deliver seamless self-service. Recognizer is grammar-based, which means it requires grammars to interpret a caller's utterance. Through advancements in semantic and language modeling, Recognizer can understand and interpret a broad range of words, phrases, and full sentences, and thus react to multiple pieces of information in a caller’s unconstrained request, promoting conversational speech.
Speech applications make requests for recognition, and the requests are delivered by the voice browser (using MRCP) to Nuance Speech Server, and then to Recognizer. Speech Server uses the Nuance recognition server to handle speech recognition of very large grammars such as those used by Directory Assistance. These large grammars are preloaded on specific recognition servers during initialization to avoid delays while handling a call.
- When the recognition server starts up, it loads any pre-assigned grammars and is ready to handle recognition requests.
- Speech Server relays recognition requests to Recognizer and returns recognizer responses to the voice browser.
- Speech Server receives the user’s speech from the RTP stream or inside SPEECHDATA messages, and sends it to Recognizer.
- Recognizer performs recognition on the audio and detects the end of speech.
- Recognizer returns results through Speech Server to the browser.
The recognizer supports a large number of languages. The language packs are packaged separately from Nuance Speech Suite, are available for download from Nuance Network.