Data processing: speech signal processing – linguistics – language – Speech signal processing – Application
Reexamination Certificate
1999-11-15
2002-09-03
Dorvil, Richemond (Department: 2654)
Data processing: speech signal processing, linguistics, language
Speech signal processing
Application
C704S501000, C704S503000, C704S201000
Reexamination Certificate
active
06446042
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to mobile communications, and specifically to the elimination of speech drop-outs for certain voice transmissions.
BACKGROUND OF THE INVENTION
Effective voice recognition technology can reduce the need for keypads and large displays. This is important when considering portable devices which are intended to connect to the world-wide communications network known as the internet. The problem is that current voice recognition technology, which is suitable for use on portable, battery-powered devices, fails to achieve needed speed or accuracy. The solution, because such products are intended to connect wirelessly to a network, is to install voice recognition hardware and software on network-based servers which a user can dial into.
Server-based recognition systems are in widespread use in wired telephone networks for such tasks as directory assistance and simple data look-up, and work well as long as the caller is using a wired telephone. Problems develop, however, when a digital wireless, e.g., cellular or PCS, telephone is used. This is because speech processing algorithms in use by all major wireless standards, such as GSM, IS-136, IS-95 and PDC, do not provide for error-free transmission. This results in signal corruption, which appear as muted “blocks” of speech, on the order of 20 ms each. To improve the perceived voice quality at the receiving end, these same systems often perform some form of extrapolation or smoothing operation to make the corruption less noticeable to the human auditory system. Unfortunately, tests have established that the underlying corruption and the follow-on extrapolation or smoothing renders the received speech nearly imperceptible to high-performance server-based speech recognition systems. Prior art systems and methods do not offer a meaningful solution to the aforementioned problem, however, a number of attempts have been made to provide speech recognition systems and GSM communications, although very little work has been done to combine the two fields of art.
U.S. Pat. No. 4,058,838, granted Nov. 15, 1987 to Crager et al., for Packet-switched facsimile communications system, describes full duplex communications between a number of communications devices, using a store-and-forward protocol.
U.S. Pat. No. 4,624,008, granted Nov. 18, 1986 to Vensko et al., for Apparatus for automatic speech recognition, describes a technique for recognizing sentence end based on pause length.
U.S. Pat. No. 4,649,567, granted Mar. 10, 1987 to Childress, for Dispatch overdialing for inter-group and other added calling/called access to communications channels in a trunked radio communications system, describes as system enabling radio transceivers, already operating with a trunked system, to communicate with additional transceivers.
U.S. Pat. No. 4,975,957, granted Dec. 4, 1990, to Ichikawa et al., for Character voice communication system, describes the extraction of parameters at the handset and the transmission of codewords as data to a base station which reconstructs the speech, and focuses on transmission of parameters as a bandwidth-saving strategy, and the algorithm presented, assuming error-free codeword transmission, will likely result in significant voice quality degradation.
U.S. Pat. No. 5,406,617, granted Apr. 11, 1995, to Bauer, for Cordless telephone/entry intercom system, describes a radio-based intercom system wherein the base station acts as a repeater for the wireless system components.
U.S. Pat. No. 5,432,883, granted Jul. 11, 1995, to Yoshihara, for Voice coding apparatus with synthesized speech LPC [linear prediction coefficients] code book, describes a system for coding speech based on LPC and error minimization.
U.S. Pat. No. 5,515,375, granted May 7, 1996, to DeClerck, for Method and apparatus for multiplexing fixed length message data and variably coded speech, describes a voice coding techniques wherein a variable rate vocal encoder receives and encodes speech.
U.S. Pat. No. 5,570,389, granted Oct. 29, 1996, to Rossi, for Method for reliable exchange of modem handshaking information over a cellular radio carrier, describes a technique for sending an initial FSK-encoded modem handshake.
U.S. Pat. No. 5,600,649, granted Feb. 4, 1997, to Sharma et al., for Digital simultaneous voice and data modem, describes a system incorporating a PC for system control, and which allows voice communication, voice mail, EMail, facsimile management, and other communications functions.
U.S. Pat. No. 5,684,791, granted Nov. 4, 1997, to Raychaudhuri et al., for Data link control protocols for wireless A TM access channels, describes on-demand available bit-rate data burst transmission in a time division multiple access channel to confirm data accuracy.
U.S. Pat. No. 5,737,716, granted Apr. 7, 1998, to Bergstrom et al., for Method and apparatus for encoding speech using neural network technology for speech classification, describes a neural network VRS which operates in single or multi stages.
U.S. Pat. No. 5,754,734, granted May 19, 1998 to Emeott et al., for Method of transmitting voice coding information using cyclic redundancy check bits, describes a techniques for prioritizing encoded speech packets prior to error checking. After error checking, the packets are interleaved for transmission.
SUMMARY OF THE INVENTION
A speech encoding system for use with a digital cellular communication device and a receiving station, includes a mechanism for determining whether a voice communications packet needs to be treated as a data communications packet; a voice recognition mechanism for receiving instructions by voice command; and a control mechanism for responding to said voice command and controlling a controlled entity.
A method for encoding a voice command generated on a digital cellular communication device and transmitted over a wireless communication network to a receiving station for controlling a controllable entity includes recognizing a voice command; determining whether the voice command needs to be treated as a data communications packet; encoding the voice command; connecting the voice command to a voice recognition mechanism; and controlling a controlled entity with the voice command.
An object of the invention is to provide error-free voice transmission for providing voice control of a controlled entity.
Another object of the invention is to provide a voice recognition system for use with a digital cellular phone system.
REFERENCES:
patent: 4058838 (1977-11-01), Crager et al.
patent: 4624008 (1986-11-01), Vensko et al.
patent: 4649567 (1987-03-01), Childress
patent: 4975957 (1990-12-01), Ichikawa et al.
patent: 5406617 (1995-04-01), Bauer
patent: 5432883 (1995-07-01), Yoshihara
patent: 5515375 (1996-05-01), DeClerck
patent: 5570389 (1996-10-01), Rossi
patent: 5592586 (1997-01-01), Maitra et al.
patent: 5600649 (1997-02-01), Sharma et al.
patent: 5615297 (1997-03-01), Davis
patent: 5684791 (1997-11-01), Raychaudhuri et al.
patent: 5737716 (1998-04-01), Bergstrom et al.
patent: 5754734 (1998-05-01), Emeott et al.
patent: 5970457 (1999-10-01), Brant et al.
patent: 5999898 (1999-12-01), Richter
patent: 6188978 (2001-02-01), Harada
patent: 6226533 (2001-05-01), Akahane
patent: 01-066416 (1989-03-01), None
patent: 10-363571 (1998-12-01), None
Shimokoshi et al (“A Study of Voice/Non-Voice Discrimination Method using Neural Networks for Integrated Packet Switching System”, ISCAS 1989).
Detlef Michael John
Ishii Atsushi
Dorvil Richemond
Nolan Daniel A.
Sharp Laboratories of America Inc.
Varitz, P.C. Robert D.
LandOfFree
Method and apparatus for encoding speech in a communications... 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 apparatus for encoding speech in a communications..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for encoding speech in a communications... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2828683