Data processing: speech signal processing – linguistics – language – Speech signal processing – Application
Reexamination Certificate
2001-01-31
2004-04-27
McFadden, Susan (Department: 2655)
Data processing: speech signal processing, linguistics, language
Speech signal processing
Application
Reexamination Certificate
active
06728677
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to the field of speech recognition or other speech processing fields such as speaker verification or text-to-speech processing. In particular the present invention discloses a system and method for dynamically improving the performance of speech recognition or other speech processing systems.
BACKGROUND OF THE INVENTION
Speech recognition systems are currently in use for responding to various forms of commerce via a telephone network. One example of such a system is utilized in conjunction with a stock brokerage. According to this system, a caller can provide their account number, obtain a quotation for the price of a particular stock issue, purchase or sell a particular number of shares at market price or a predetermined target price among other types of transactions. Natural language systems can also be used to respond to such things as requests for telephone directory assistance.
These types of speech recognition systems are typically deployed to handle a maximum call capacity for peak periods. This means that the hardware supporting the system provides enough memory, processing power and bandwidth to handle calls with a predetermined level of accuracy. For example, a company may deploy a speech recognition system that may handle 10,000 callers because at noon the company has that many calls. However, at all other times the system is not used to its full potential because only 5,000 callers are in the system.
Speech recognition systems typically are configurable, within limits, as to the amount of processing power, memory, network bandwidth, and other system resources that they may consume. Often, memory, speed, and accuracy can be traded off against each other. For instance, a particular configuration of one system may use less CPU resources than another, typically at the cost of lower average speech recognition accuracy. System configuration is often done ahead of time, resulting in a particular resource/performance tradeoff for the particular deployment.
SUMMARY OF THE INVENTION
The present invention introduces a system and method for dynamically improving speech recognition or other speech processing systems by estimating the utilization of resources in the system; and improving the performance of the system according to the availability of resources.
REFERENCES:
patent: 6119087 (2000-09-01), Kuhn et al.
patent: 6345279 (2002-02-01), Li et al.
patent: 6505161 (2003-01-01), Brems
patent: 6542600 (2003-04-01), Munson et al.
Douglas A. Reynolds, Automatic Speaker Recognition Using Gaussian Mixture Speaker Models, vol. 8, No. 2, 1995, The Lincoln Laboratory Journal, Pp. 173-192.
Kannan Ashvin
Leggetter Christopher
Murveit Hy
Schuster Michael
Blakely , Sokoloff, Taylor & Zafman LLP
McFadden Susan
Nuance Communications
LandOfFree
Method and system for dynamically improving performance of... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and system for dynamically improving performance of..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and system for dynamically improving performance of... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3212212