Telephonic communications – Audio message storage – retrieval – or synthesis
Patent
1994-03-21
1996-12-03
Zele, Krista M.
Telephonic communications
Audio message storage, retrieval, or synthesis
379 88, H04M 350
Patent
active
055816002
DESCRIPTION:
BRIEF SUMMARY
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a service platform, in particular to a platform for the provision of voice services which use speech technology. Typically such services will be interactive services, such as home banking, but the invention also finds application with non-interactive services.
2. Related Art
There is increasing pressure on the telecommunications industry to provide enhanced telephone services, to adapt more quickly to changing demands, and to provide increased value for money. One of the areas of technology which is key to achieving these goals is "speech technology", that is the technology of speech recognition, speaker recognition, speaker verification, speech synthesis, text-to-speech conversion, speech compression, etc. This technology can be used to provide improved customer service, replacing or assisting human operators, and or providing completely new services to customers. Typically, the various aspects of this technology are provided separately, often by different items of equipment. For some applications the equipment to provide the speech recognition, etc., will be associated with a main exchange in a telephone network (PSTN), while other applications equipment will be provided in, or in association with, customer's premises. Particularly for network-based equipment, there exists a need for equipment which can handle many calls simultaneously, preferably with the ability to provide multiple services simultaneously.
To some extent, this need has been met by equipment which provides interactive voice services and in which processors, typically digital signal processors (DSP)s are programmed to provide a particular service, the recognition or synthesis algorithms necessary for the provision of the particular service also being stored on or in close association with the programmed DSP. Such known apparatus is typically dedicated to the provision of one service, with the DSP of DSPs needed to provide the service being allocated as a whole to each incoming call as and when it is answered. This dedicated arrangement is simple and convenient, but it does mean that if any part of the service offering is very processor hungry, for example using a complex, speaker-independent recognition algorithm with a large vocabulary, the level of processor provisioning required to run that part of the service needs to be provided throughout the entire length of each call to that service, and this is clearly inefficient. While this inefficiency is supportable in small systems, systems which need to handle many calls are handicapped by the need to over provide processing resource. It would be desirable, therefore, if some means could be found to provide processing resource on an "as-needed" basis, while at the same time providing a reliable, quick and user-friendly service platform.
BRIEF SUMMARY OF THE INVENTION
At least one aspect of the present invention addresses the problem of the allocation of processing resource (hereinafter "resource allocation").
According to a first aspect, the present invention provides an interactive voice platform for the provision of automated speech based services over a telephone network, the platform comprising: the following functions: speech recognition, speaker recognition, speaker verification, speech synthesis; telephony channels of the network; said first processing means; speech bus; and the line interface means; means and holding one or more of the following: a program including a speech recognition algorithm, a program including a speech synthesis algorithm, a program including a speaker recognition or verification algorithm, templates to be used in conjunction with any one of the foregoing programs; and operable to effect the transfer of an instance of one or more of the programs or templates held in said memory means to said first processing means for use therein in the provision of the appropriate function during the answering of a call.
BRIEF DESCRIPTION OF THE DRAWINGS
Preferred embodiments of th
REFERENCES:
patent: 4949374 (1990-08-01), Ishii
patent: 5125024 (1992-06-01), Gokcen
patent: 5165095 (1992-11-01), Borcherding
patent: 5185781 (1993-02-01), Dowden
patent: 5193110 (1993-03-01), Jones
Rosenberg, "Sprint Expands Testing of Voice-Activated Card", Kansas City Star Business Section Apr. 7, 1992.
Enhanced Information Services, Technical Report, Texas Instruments Telecom Systems, pp. 1-20.
Texas Instruments Multiserve.TM. Media Server.
Texas Instruments Multimedia Processor.
"Tandem Computers and Texas Instruments Announce Cooperative Marketing and Development Agreement", Media Information, pp. 1-3.
"Sprint Announces Expanded Voice-Activated Fonacrd (am) Pilot Based on Impressive Test Results", pp. 1-7.
"TI Announces Multiserve.TM. Telecommunications Platform", Texas Instruments, News Release ITG-005, pp. 1-5.
Texas Instruments Voice Dialing Services, pp. 1-6.
Enhanced Information Services A Strategic Approach, Texas Instruments Telecom Systems, pp. 1-14.
The NVP 1000, Voice Processing System, Newbridge, pp. 1-6.
"NVP 1060, Technical Supplement", Draft, Newbridge.
Mulla et al, "Application of Speech Recognition and Synthesis to PABX Services", published May 6, 1995, pp. 273-280.
Conversant I Voice System: Architecture and Applications AT&T Journal, R. J. Perdue E. L. Rissanen pp. 34-47.
Elliston Dave M.
Forse Nicholas J. A.
Hughes Patrick M.
Watts Martin O.
Presson Thomas F.
Zele Krista M.
LandOfFree
Service platform does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Service platform, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Service platform will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-792132