|
|
 |
|
|
|
|
|
|
|
|
|
| Key Benefits |
- SpeakFreely
SpeakFreely allows applications to understand an exceptionally broad range of words and phrases without requiring highly complex grammar rules that would be prohibitively time consuming and expensive to develop using traditional recognition techniques. SpeakFreely can handle response to open-ended questions (for instance, How can I help you today?) while still delivering very accurate recognition results. It achieves this using statistical language models (SLMs) and statistical semantic models (SSMs) that determine what a caller means in addition to the words they said.
- Automatic improvement in deployment
To further augment its already impressive accuracy, OpenSpeech Recognizer automatically adapts recognizer parameters to optimize performance in each deployment. This "LEARN" capability tunes OSR based on characteristics of both the caller population (for instance, regional accents) and telephony channels (such as VoIP).
- Endpointing and Barge-in
OpenSpeech Recognizer is able to separate speech from background noise with remarkable precision due to its superior endpointing and speech detection algorithms. This capability contributes to OSR's high quality barge-in capabilities, allowing callers to interrupt the application at any time rather than waiting until the end of a prompt. OSR can be configured to wait until a valid phrase is recognized, allowing long message playback even while the caller may be speaking with others.
- Outstanding Performance and Scalability
OpenSpeech Recognizer provides industry-leading densities and scalability thanks to patented Finite State Transducer technology that reduces CPU and memory needs by up to 90%. OSR easily handles very large grammars, even those with over one million words, for applications such as our Nuance Automated Directory Assistance solution.
- The Largest Language Portfolio
OpenSpeech Recognizer supports 46 languages with exceptional accuracy, more than any other recognition solution in the world. Nuance also boasts the largest international Professional Services organization in the world, meaning that we are ideally placed to deploy best of breed speech-enabled applications across EMEA, Asia, and Australasia as well as in the Americas. The full list of languages supported is below.
| Language |
Dialect |
| Cantonese |
Hong Kong |
| Catalan |
Spain |
| Czech |
Czech Republic |
| Danish |
Denmark |
| Dutch |
Netherlands |
| English |
Australia
Great Britain
India
Singapore
USA |
| Euskera |
Spain |
| Finnish |
Finland |
| Flemish |
Belgium |
| French |
Belgium
Canada
France
Luxembourg
Switzerland |
| German |
Austria
Germany
Luxembourg
Switzerland |
| Greek |
Greece |
| Hebrew |
Israel |
|
|
|
| Language |
Dialect |
| Hungarian |
Hungary |
| Italian |
Italy
Switzerland |
| Japanese |
Japan |
| Korean |
Korea |
| Mandarin |
PRC
Taiwan |
| Norwegian |
Norway |
| Polish |
Poland |
| Portuguese |
Brazil
Portugal |
| Russian |
Russia |
| Slovak |
Slovakia |
| Slovene |
Slovenia |
| Spanish |
Argentina
Colombia
Spain
USA |
| Swedish |
Finland
Sweden |
| Turkish |
Turkey |
| Welsh |
Great Britain |
|
|
Note that OSR is able to support many dialects that are not listed here (for instance, New Zealand English), and we are able to support additional languages on demand. Contact your sales representative for more information.
- Choice and Investment Protection through Open, Standards-Based Solutions
OSR provides the industry's most comprehensive support of standards and VoiceXML IVR platforms. This protects your investment in speech technology. OSR is optimized for VoiceXML and supports SRGS (Speech Recognition Grammar Specification) grammars as defined by the W3C (cf. http://www.w3.org/TR/2003/PR-speech-grammar-20031218/), and MRCP via the Nuance Media Server (SWMS).
|
|
|
|
|
|
|
|
|
|
