Data processing: speech signal processing – linguistics – language – Speech signal processing – Recognition
Reexamination Certificate
2000-09-11
2004-03-16
McFadden, Susan (Department: 2655)
Data processing: speech signal processing, linguistics, language
Speech signal processing
Recognition
C704S251000, C704S255000, C704S270000
Reexamination Certificate
active
06708150
ABSTRACT:
INCORPORATION BY REFERENCE
The disclosures of the following priority applications are herein incorporated by reference:
Japanese Patent Application No. 11-255982 filed Sep. 9, 1999
Japanese Patent Application No. 11-255983 filed Sep. 9, 1999
Japanese Patent Application No. 11-255984 filed Sep. 9, 1999
Japanese Patent Application No. 2000-53257 filed Feb. 29, 2000
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a voice recognition apparatus and a voice recognition navigation apparatus.
2. Description of the Related Art
There are car navigation apparatuses (hereafter referred to as navigation apparatuses) that display the current position of the vehicle, display a map over a wide area or in detail and provide guidance to the driver along the traveling direction over the remaining distance to the destination in the prior art. There are also voice recognition navigation apparatuses in the prior art having a function of enabling the driver engaged in driving to issue operating instructions by voice to improve driver safety (see Japanese Laid-Open Patent Publication No. 09-292255, for instance).
The voice recognition software program used in a voice recognition navigation apparatus normally judges that a speech has ended at a point in time at which there is no longer any speech after the start of a speech and calculates the correlation values between audio data obtained up to the point in time at which there is no longer any speech after the start of the speech and all the recognition words in the recognition dictionary. Then, the recognition word achieving the largest correlation value is judged to be the recognition results. Speech that needs to be recognized by a voice recognition navigation apparatus falls into various categories of words and phrases such as navigation commands (bird's eye view display, enlarge, reduce, etc.) used to issue instructions for various types of navigation operations, train stations, golf course names, hospital names and ski resort names.
Among these speeches, the golf course names, hospital names, ski resort names and the like tend to be longer than navigation commands and train station names, and are, therefore, extremely difficult to recognize.
In addition, the voice recognition software program normally calculates the correlation values between the audio data representing the speech made by the user (driver) after a TALK switch or the like is pressed, and the recognition words in the recognition dictionary. It then judges the recognition word achieving the largest correlation value to be the recognition results.
However, there is a problem in that the chance of erroneous recognition increases when the user starts his speech immediately after pressing the TALK switch.
Furthermore, the driver may become confused as to which instruction should be given to the navigation apparatus next and may utter a totally erroneous instruction speech. In such a case, too, the recognition word in the recognition dictionary achieving the largest correlation value is judged to be the instruction spoken by the driver and the navigation operation corresponding to that instruction is performed. For instance, let us consider a situation in which the driver, wishing to display a map, says “map” when there are only three recognition words, e.g., “audio,” “television” and “bird's eye view display” provided in the recognition dictionary. In such a case, if the correlation value between the audio data and “television” is the largest, the navigation apparatus displays the television screen. As a result, a navigation operation other than that instructed by the driver is executed to confuse the driver.
There is another problem in that an erroneous recognition may occur if the user pronounces a given word in a slightly different manner or if the user employs an alternative expression.
SUMMARY OF THE INVENTION
A first object of the present invention is to provide a voice recognition apparatus and a voice recognition navigation apparatus capable of recognizing long speeches with ease and a high degree of reliability.
A second object of the present invention is to provide a voice recognition apparatus and a voice recognition navigation apparatus capable of achieving a successful voice recognition in a reliable manner even when a speech starts immediately after the TALK switch is pressed or when the actual pronunciation is slightly different from the standard pronunciation.
A third object of the present invention is to provide a voice recognition apparatus and a voice recognition navigation apparatus with which it is possible to ensure that none of the recognition words in the recognition dictionary is recognized if a word which is not provided in the recognition dictionary is spoken.
A fourth object of the present invention is to provide a voice recognition apparatus and a voice recognition navigation apparatus capable of achieving a successful voice recognition with a high degree of reliability even when the user pronounces part of the word or phrase in a manner slightly differently from the standard or if the user chooses an alternative word or phrase, and a recognition word generating method that may be adopted in the voice recognition apparatus and the voice recognition navigation apparatus.
Another object of the present invention is to provide a recording medium and a data signal in which data used in the apparatuses and a control program for controlling the apparatuses are provided.
In order to attain the above object, a voice recognition apparatus according to the present invention, comprises: a voice input device; a storage device that stores a recognition word indicating a pronunciation of a word to undergo voice recognition; and a voice recognition processing device that performs voice recognition processing by comparing audio data obtained through the voice input device and voice recognition data created in correspondence to the recognition word, and the storage device stores both a first recognition word corresponding to a pronunciation of an entirety of the word to undergo voice recognition and a second recognition word corresponding to a pronunciation of only a starting portion of a predetermined length of the entirety of the word to undergo voice recognition as recognition words for the word to undergo voice recognition.
In this voice recognition apparatus, it is preferred that when the pronunciation of the entirety of the word to undergo voice recognition extends over a first predetermined length, the storage device stores the second recognition word corresponding to a pronunciation of only a starting portion of a second predetermined length of the entirety of the word to undergo voice recognition as a recognition word for the word to undergo voice recognition.
A voice recognition navigation apparatus according to the present invention, comprises: a voice input device; a storage device that stores a recognition word indicating a pronunciation of a word to undergo voice recognition; and a voice recognition processing device that performs voice recognition processing by comparing audio data obtained through the voice input device and voice recognition data created in correspondence to the recognition word; a map information storage device that stores map information; and a control device that engages in control for providing route guidance based upon, at least, recognition results obtained by the voice recognition processing device and the map information, and the storage device stores both a first recognition word corresponding to a pronunciation of an entirety of the word to undergo voice recognition and a second recognition word corresponding to a pronunciation of only a starting portion of a predetermined length of the entirety of the word to undergo voice recognition as recognition words for the word to undergo voice recognition.
Another voice recognition apparatus according to the present invention, comprises: a voice input device; a storage device that stores a recognition word indicating a pronunciation of a wor
Hirayama Yoshikazu
Kobayashi Yoshiyuki
Crowell & Moring LLP
McFadden Susan
Zanavi Informatics Corporation
LandOfFree
Speech recognition apparatus and speech recognition... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Speech recognition apparatus and speech recognition..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Speech recognition apparatus and speech recognition... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3216466