Speech Recognition Software SolutionKey Features
NSCServer - Runs on the host computer and performs resource management activities as well as managing the communication between the speech-driven application and the speech recognition engine. NSC Speech Recognition Engine - Runs on the speech recognition server and implements NSC's proprietary speech recognition technology on the host CPU. NSC develops and implements two speech recognition engines: NaturalASR and NaturalKWS. API - Application Programming Interface between the speech-driven application and the speech recognition engine.. Development Tools - Provided for the development of speech applications on various target platforms. Software ArchitectureNSC’s software product offering supports MRCP (Media Resource Control Protocol), an open standard allowing for distributed IVR (Interactive Voice Response) platforms to invoke, via API and other control, a specialized vendor-specific speech server. The protocol includes an API that can interface with speech recognition engines such as NSC’s speech engines. IVR platforms supporting MRCP can directly plug into NSC’s speech recognition solution. Customers can seamlessly upgrade to future NSC recognition engine versions via the MRCP interface that’s in place. NSC’s current offering supports MRCPv1/ MRCPv2. MRCP is a key component of NSC’s speech recognition software architecture for IVR applications. NSC’s keyword spotting software readily integrates with leading speech analytics solutions and is also an integral part of NSC’s MSS product. Applications interface with NSC’s speech recognition solution via MRCP. MRCP interfaces with the NSCServer, the server side of the client/server speech recognition architecture. NSCServer can access two types of speech recognition resources: (1) NaturalKWS – NSC’s Keyword spotting engine and (2) NaturalASR – NSC’s Grammar-based speech recognition engine. Grammars are provided by the application via SRGS (Speech Recognition Grammar Specification), a standard grammar specification markup language, part of the MRCP protocol. Each speech recognition resource can run any of the supported languages. NSC Software-Based Speech Recognition EnginesNaturalASR – A High-Accuracy Speech Recognition Engine for NSC’s speech recognition software offering targeted at Telephony Speech-Driven Services in the Call Center, Telco, and Enterprise markets. Its main focus is speech enabling IVR (Interactive Voice Response) applications. NaturalKWS– A robust, real-time keyword spotting engine, part of NSC’s speech recognition software offering, designed for locating pre-defined words in real-time and in off-line calls targeting the Security and Intelligence, Call Center and Online Media markets. NaturalKWS is also the speech recognition engine that drives NSC’s MSS product. NaturalASRNSC NaturalASR - A High-Accuracy Speech Recognition Engine for NSC’s speech recognition software offering targeted at Telephony Speech-Driven Services. NSC's telephony speech recognition engine, part of NSC’s speech recognition software offering, designed for speech enabling services in the Call Center, Telco, and Enterprise markets. Its main focus is IVR (Interactive Voice Response) applications. NaturalASR speech enables services such as: Call center automation, information access, automated attendant, value-added services, call routing, and voice controlled voicemail. NSC NaturalASR has the following main features:
NaturalKWSNSC NaturalKWS - A robust, real-time keyword spotting engine for NSC’s speech recognition software offering driving Audio Surveillance, Speech Analytics, and Data Mining applications. It is the keyword spotting engine for MSS, NSC’s speech-driven media search engine. NSC NaturalKWS is the Keyword Spotting (KWS) engine for NSC’s speech recognition software offering designed for locating words out of a pre-defined list in spoken conversations. It operates in real time or on recorded calls, and is targeted for the Security and Intelligence and Call Center markets. NaturalKWS is used for speech analytics and call surveillance in the Security and Intelligence market as well as for speech analytics gauged at mining business information and monitoring agent performance in the Call Center market. In addition, NaturalKWS can be used in any market segment, service or application where there is a need for analysis of a large number of speech calls or recordings. NaturalKWS implements a unique word spotting algorithm - a one-stage approach. The approach searches for keywords in a one-stage operation. The recognition is phoneme-based - NSC NaturalKWS looks for the complete word in the speech stream based on a target sequence of phonemes that represent the keyword or key phrase. The unique approach by NSC operates directly on the incoming speech in one stage rather than the usual two-stage system that translates the speech into text (phonemes or words) and then performs a text search for the keyword. NSC NaturalKWS has the following main features:
|
