JP2008309865A - Voice recognition device and voice recognition method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To prevent that uttering to outside of a vehicle is incorrectly recognized as uttering to an in-vehicle device, by separating the uttering to the in-vehicle device from the uttering to outside of the vehicle. <P>SOLUTION: In the voice recognition device 10, a voice recognition processing section 13a uses a specified area dictionary 12b and an outside specified area dictionary 12c, and recognizes voice by determining whether or not an uttered vocabulary is included in the dictionaries, and performs command conversion. A command after the command conversion is output to a car navigation device 20 by a command output processing section 13c. Furthermore, the voice recognition processing section 13a determines whether or not a current position of a vehicle 1 is included in the specified area information 12a based on a GPS position information obtained from the car navigation device 20. When it is determined that it is included, only the outside specified area dictionary 12c is used so that voice recognition of the uttered vocabulary which is assumed to be uttered in the specified area may not be performed. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention provides speech recognition means for speech recognition of a vocabulary spoken by determining whether or not a vocabulary spoken by a vehicle occupant is included in a speech recognition target vocabulary stored in the vocabulary storage means. The present invention relates to a voice recognition device having a command conversion means for converting a result of voice recognition by the voice recognition means into a command and passing it to a vehicle-mounted device, and a voice recognition method using the voice recognition device. The present invention relates to a speech recognition apparatus and a speech recognition method that distinguish between utterances and utterances to the outside of the vehicle and prevent erroneous recognition of utterances to the outside of the vehicle as utterances to an in-vehicle device.

  In recent years, various ideas have been made for realizing a technology for recognizing a user's voice. If the user's voice can be recognized, it is possible for the user to perform various device operations by voice. Especially, in-vehicle devices are concerned about the influence of manual operation by the driver on the driving. The practical application of operation technology is eagerly desired.

  By the way, at present, an in-vehicle device equipped with a vehicle voice recognition function has a specification for recognizing a specific command uttered after pressing a talk switch. By using this talk switch, the in-vehicle device can recognize a specific command more accurately.

  However, since the pressing operation of pressing the talk switch causes a load on the driver other than the driving operation, continuous voice recognition without using the talk switch is expected to become mainstream in the future. .

  In the continuous voice recognition without using the talk switch, it is necessary to clearly recognize whether or not the driver's utterance is for the in-vehicle device. For this purpose, for example, in Patent Document 1, a vocabulary pattern for speech recognition is registered in association with vehicle position information, and at the time of speech recognition, speech recognition processing is performed and a recognition vocabulary based on the vehicle location information is registered. Conventional techniques for improving the accuracy of speech recognition by determining validity are disclosed.

  Further, in Patent Documents 2 to 4, when a specific keyword or silent part is recognized, speech recognition of an utterance is started, the capacity of a dictionary for speech recognition is reduced, and accuracy of speech recognition is improved. The prior art is disclosed.

  Patent Document 5 discloses a conventional technique for recognizing a speaker's posture, recognizing whether or not the utterance is directed to a vehicle-mounted device from the recognized speaker's posture, and improving the accuracy of speech recognition. Is disclosed.

  Patent Document 6 distinguishes between a recognition dictionary used for voice recognition when the vehicle is running and a dictionary used for voice recognition when the vehicle is stopped, and whether the vehicle is running or stopped. Conventional techniques for improving the accuracy of speech recognition by switching these dictionaries depending on whether they are in the middle or the like are disclosed.

WO2003 / 088209 specification JP 2002-132290 A JP 2003-228392 A Japanese Patent Laid-Open No. 2004-341033 JP 2000-267695 A Japanese Unexamined Patent Publication No. 2-193198

  However, in the conventional techniques represented by the above-mentioned Patent Documents 1 to 6, there are cases where the utterance is the same vocabulary and both the utterance to the in-vehicle device and the utterance to the outside of the vehicle. These cannot be distinguished from each other, and the utterance to the outside of the vehicle is misrecognized as the utterance to the in-vehicle device, which may reduce the accuracy of voice recognition.

  The present invention has been made to solve the above problems (problems), and distinguishes between an utterance to an in-vehicle device and an utterance to the outside of the vehicle, and an utterance to the outside of the vehicle as an utterance to the in-vehicle device. It is an object of the present invention to provide a voice recognition device and a voice recognition method that prevent erroneous recognition.

  In order to solve the above-described problems and achieve the object, the present invention determines whether a vocabulary spoken by a vehicle occupant is included in a speech recognition target vocabulary stored in a vocabulary storage means. A speech recognition apparatus comprising: speech recognition means for recognizing the spoken vocabulary; and command conversion means for converting a result of speech recognition by the speech recognition means into a command and transferring it to a vehicle-mounted device, The vocabulary storage means is a specific point vocabulary storage means for storing the speech recognition target vocabulary that the passenger is supposed to utter at a specific point in the voice recognition target vocabulary in association with the position information corresponding to the specific point. And a vocabulary storage device outside the specific location for storing the speech recognition target vocabulary other than the vocabulary stored in the specific location vocabulary storage device.

  Further, in the present invention according to the above-described invention, the position information storage unit that stores the specific point and the position information corresponding to the specific point in association with each other, and the position information storage unit includes the position information where the vehicle is positioned. It is further characterized by further comprising position information determination means for determining whether or not to be performed.

  Also, in the present invention according to the above invention, the voice recognition means utters the passenger when the position information determination means determines that the position information storage means includes the position information where the vehicle is located. The spoken vocabulary is voice-recognized by determining whether or not the determined vocabulary is included in the speech recognition target vocabulary stored in the off-specific vocabulary storage unit, and the vehicle is detected by the position information determination unit When it is not determined that the position information storing means includes the position information storage means, the vocabulary spoken by the passenger is stored in the specific point vocabulary storage means and the specific point outside vocabulary storage means. The spoken vocabulary is recognized by speech by determining whether or not it is included in the speech recognition target vocabulary.

  Also, in the present invention according to the above invention, the command conversion means determines that the utterance by the passenger is made by the voice recognition means when the position information storage means determines that the position information where the vehicle is located is included in the position information storage means. When the determined vocabulary is determined to be included in the speech recognition target vocabulary stored in the specific point vocabulary storage means, the result of speech recognition by the speech recognition means is converted into a command and received by the in-vehicle device. It is characterized by canceling the passing.

  In the present invention, in the above invention, when it is determined that the position information storage means includes the position information where the vehicle is located, the vocabulary spoken by the passenger by the voice recognition means is the specified When it is not determined that the vocabulary is included in the speech recognition target vocabulary stored in the point vocabulary storage unit, the spoken vocabulary is associated with the specific point corresponding to the position information in the specific point vocabulary storage unit. It further has a recognition vocabulary registration means for registration.

  Further, the present invention according to the above-mentioned invention further comprises speech direction determination means for determining whether or not the utterance direction of the vocabulary spoken by the passenger is the in-vehicle device direction, and the position information determination means is Whether the position information storage means includes the position information of the vehicle only when the utterance direction determination means determines that the utterance direction of the vocabulary spoken by the passenger is not the in-vehicle device direction. It is characterized by determining whether or not.

  The present invention also provides speech recognition for speech recognition of a vocabulary spoken by determining whether a vocabulary spoken by a vehicle occupant is included in a speech recognition target vocabulary stored in a speech recognition dictionary. And a command conversion step of converting a result of speech recognition in the speech recognition step into a command and transferring it to the in-vehicle device, wherein the speech recognition dictionary is a dictionary of the speech recognition target vocabulary. Among them, a specific point vocabulary dictionary that stores a speech recognition target vocabulary that the passenger is supposed to utter at a specific point in association with position information corresponding to the specific point, and a vocabulary other than the vocabulary stored in the specific point vocabulary dictionary A vocabulary dictionary outside the specific location that stores the speech recognition target vocabulary, and the speech recognition step determines whether the location information where the vehicle is located is location information corresponding to the specific location. When the position information determining step determines that the position information where the vehicle is located is position information corresponding to the specific location, the vocabulary spoken by the passenger is stored in the vocabulary outside the specific location. The spoken vocabulary is voice-recognized by determining whether or not the vocabulary is included in the speech recognition target vocabulary, and the position information where the vehicle is located is the position information corresponding to the specific point in the position information determining step. If it is not determined that the vocabulary spoken by the passenger is included in the speech recognition target vocabulary stored in the specific location vocabulary dictionary and the specific location vocabulary dictionary, The vocabulary spoken is recognized by voice.

  Further, in the present invention according to the above invention, in the command conversion step, when it is determined that position information where the vehicle is located is included in the position information storage unit, the utterance by the passenger is performed by the voice recognition step. When it is determined that the recognized vocabulary is included in the speech recognition target vocabulary stored in the specific point vocabulary dictionary, the result of speech recognition by the speech recognition process is converted into a command and transferred to the in-vehicle device. It is characterized by canceling.

  According to the present invention, the vocabulary storage means includes the specific-point vocabulary storage means and the specific-point vocabulary storage means, so that the voice recognition means properly uses these words that are spoken out of the vehicle by the vehicle occupant. There is an effect that voice recognition is possible.

  In addition, according to the present invention, it is possible to determine whether or not the point where the vehicle is located is a specific point based on the position information where the vehicle is located.

  Further, according to the present invention, when the vehicle is located at a specific point, voice recognition is performed using the non-specific point vocabulary storage unit, and when the vehicle is not located at the specific point, the specific point vocabulary storage unit is used for voice recognition. Because it recognizes, when the vehicle is located at a specific point, the recognition target vocabulary included in the specific point vocabulary storage means is not recognized, and it becomes possible to prevent an erroneous command to the in-vehicle device due to erroneous voice recognition. There is an effect.

  Further, according to the present invention, when the vehicle is located at a specific point, voice recognition is performed using the specific point vocabulary storage means, and the vocabulary spoken by the passenger is included in the specific point vocabulary storage means. Since the recognition result is converted into a command and the transfer to the in-vehicle device is canceled, when the vehicle is located at a specific point, the in-vehicle device is not controlled based on the recognition target vocabulary included in the specific point vocabulary storage means, There is an effect that it is possible to prevent an erroneous command to the in-vehicle device due to erroneous voice recognition.

  Further, according to the present invention, when the vocabulary recognized by speech when the vehicle is located at a specific point is not included in the specific point vocabulary storage means, the vocabulary is registered in the specific point vocabulary storage means. The registered vocabulary in the vocabulary storage means is automatically increased, and the recognition accuracy of utterances outside the vehicle can be improved.

  In addition, according to the present invention, it is determined whether or not the position where the vehicle is located is a specific point only when the utterance direction is not the in-vehicle device direction. There is an effect that it can be recognized.

  Exemplary embodiments according to a speech recognition apparatus and speech recognition method of the present invention will be described below in detail with reference to the accompanying drawings.

  First, the configuration of the speech recognition apparatus according to the embodiment will be described. FIG. 1 is a functional block diagram illustrating the configuration of the speech recognition apparatus according to the embodiment. As shown in the figure, in the vehicle 1, the current position information of the vehicle is acquired from the voice recognition device 10 according to the embodiment and the GPS (Global Positioning System) via the CAN (Controller Area Network) 2, and the voice is obtained. A car navigation device 20 that is a control target based on a command based on the utterance content that is delivered to the recognition device 10 and voice-recognized is connected.

  The voice recognition device 10 includes a display unit 11a that is a display unit such as a display device having a predetermined display screen, a voice generation unit 11b that is a voice generation unit such as a speaker device that emits voice, and a volatile or nonvolatile memory. It has the memory | storage part 12 and the control part 13 which are means. The voice recognition device 10 is connected to a microphone 14 that inputs voice data detected from the outside to the voice recognition device 10. Since the microphone 14 is a stereo microphone having directivity, the direction in which the detected sound is generated can be specified.

  The storage unit 12 stores specific area information 12a, a specific area dictionary 12b, and a non-specific area dictionary 12c. These are stored in the storage unit 12 as a predetermined table. The specific area information 12a is obtained when the current position information of the vehicle 1 hits the position information stored in the GPS position information, which is GPS longitude and latitude information, as shown in FIG. It is a table which can acquire the address corresponding to position information, the category to which the position of the vehicle 1 belongs, and the specific area name to which the position of the vehicle 1 belongs.

  As shown in FIG. 3, the specific area dictionary 12b has a voice corresponding to a command for voice control of the car navigation device 20 that is assumed to be spoken by a passenger of the vehicle 1 in the specific area. This is a dictionary that stores a vocabulary corresponding to a recognition target vocabulary for each specific area name. The utterance assumption vocabulary stored in this dictionary is a vocabulary subject to speech recognition that may be uttered by a passenger of the vehicle 1 in a specific area, and the car navigation device 20 is controlled by performing speech recognition in the specific area. It is a vocabulary that should be recognized as a normal utterance, particularly as an utterance outside the vehicle, without being based on the intention to do.

  The specific area outside dictionary 12c is registered in the specific area dictionary 12b among the speech recognition target vocabulary to be converted into a command that is voice-recognized and passed to the car-navigation device 20 in order to control the car navigation device 20. It is a dictionary that stores all vocabulary that is not.

  The control unit 13 is responsible for overall control of the voice recognition device 10. As a characteristic functional configuration particularly related to the embodiment, the voice recognition processing unit 13 a, the speech direction determination processing unit 13 b, and the command output processing unit 13 c are used. And have. Other functional configurations are omitted.

  When the vocabulary spoken by the passenger of the vehicle 1 is detected by the microphone 14, the speech recognition processing unit 13 a accepts the detected utterance vocabulary for a while and temporarily stores it. The speech vocabulary is subjected to voice recognition processing according to the determination result of whether the position of the vehicle 1 is in the specific area and whether the utterance direction of the passenger of the vehicle 1 is the direction of the car navigation device 20. Note that the term “speech recognition” simply refers to a case where it is determined that the detected utterance vocabulary is included in the specific area information 12a or the specific area dictionary 12b, and the speech recognition process is a process of attempting speech recognition. .

  The utterance direction determination processing unit 13 b determines whether or not the utterance direction of the utterance vocabulary by the passenger of the vehicle 1 detected by the microphone 14 is directed toward the car navigation device 20. When it is determined that the utterance direction is toward the car navigation device 20, the speech recognition processing unit 13a uses the specific area dictionary 12b and the non-specific area dictionary 12c, and the utterance vocabulary is included in these dictionaries. The voice is recognized by determining whether or not it is, and command conversion is performed. The command converted command is output to the car navigation device 20 by the command output processing unit 13c.

  On the other hand, when it is not determined that the utterance direction of the utterance vocabulary by the passenger of the vehicle 1 is toward the car navigation device 20, the speech recognition processing unit 13a is based on the GPS position information acquired from the car navigation device 20, When it is determined whether or not the current position of the vehicle 1 is included in the specific area information 12a and is determined to be included, it is assumed that the utterance is made in the specific area using only the specific area outside dictionary 12c. Utterance vocabulary is not recognized. Thereby, it is possible to prevent an erroneous command to the in-vehicle device due to erroneous voice recognition in the specific area.

  Alternatively, it is assumed that the voice recognition processing unit 13a performs voice recognition using the specific area dictionary 12b and the specific area outside dictionary 12c. If the utterance direction determination processing unit 13b does not determine that the utterance direction is toward the car navigation device 20, the voice recognition processing unit 13a is based on the GPS position information acquired from the car navigation device 20, and Is determined to be included in the specific area information 12a. At this time, when it is determined that the current position of the vehicle 1 is included in the specific area information 12a, the speech recognition processing unit 13a further determines that the utterance vocabulary by the passenger of the vehicle 1 is included in the specific area dictionary 12b. In the case where the command is received, the command output processing unit 13c cancels the output of the voice recognition result command by the voice recognition processing unit 13a to the car navigation device 20. This makes it possible to prevent erroneous commands to the in-vehicle device due to incorrect voice recognition.

  Note that the speech recognition processing unit 13a determines the GPS position acquired from the car navigation device 20 by the speech recognition processing unit 13a when the speech direction determination processing unit 13b does not determine that the speech direction is toward the car navigation device 20. When it is determined that the current position of the vehicle 1 is included in the specific area information 12a based on the information, the speech recognition processing unit 13a does not include the utterance vocabulary by the passenger of the vehicle 1 in the specific area dictionary 12b. When determined, the utterance vocabulary may be registered in the specific area dictionary 12b. As a result, the vocabulary spoken when the vehicle 1 is located in the specific area can be automatically registered in the specific area dictionary 12b, and erroneous recognition is prevented and recognized in speech recognition of the spoken vocabulary in the specific area. The accuracy can be improved.

  Next, the speech recognition process of an Example is demonstrated. FIG. 4 is a flowchart illustrating a speech recognition processing procedure according to the embodiment. As shown in the figure, first, the voice recognition processing unit 13a receives input of voice information from the microphone 14 (step S101). Subsequently, the voice recognition processing unit 13a receives an input of position information from the car navigation device 20 (step S102).

  Subsequently, the speech direction determination processing unit 13b determines whether or not the speech direction is a direction of the car navigation device 20 (step S103). When it is determined that the utterance direction is the direction of the car navigation device 20 (Yes at Step S103), the process proceeds to Step S104. On the other hand, when it is not determined that the utterance direction is the direction of the car navigation device 20 (No at Step S103), the process proceeds to Step S105.

  In step S104, the voice recognition processing unit 13a sets to use the specific area dictionary 12b and the specific area outside dictionary 12c for voice recognition.

  In step S105, the voice recognition processing unit 13a determines whether or not the current position of the vehicle 1 is a specific area based on the GPS position information acquired from the car navigation device 20. When it is determined that the current position of the vehicle 1 is a specific area (Yes at Step S105), the process proceeds to Step S106, and when it is not determined that the current position of the vehicle 1 is a specific area (No at Step S105), Control goes to step S104.

  In step S106, the voice recognition processing unit 13a excludes the specific area dictionary 12b for voice recognition and sets to use only the non-specific area dictionary 12c.

  Next, speech recognition processing according to another embodiment will be described. FIG. 5 is a flowchart showing a voice recognition processing procedure according to another embodiment. As shown in the figure, first, the voice recognition processing unit 13a sets to use the specific area dictionary 12b and the specific area outside dictionary 12c for voice recognition (step S111). Subsequently, the voice recognition processing unit 13a receives input of voice information from the microphone 14 (step S112). Subsequently, the voice recognition processing unit 13a receives an input of position information from the car navigation device 20 (step S113).

  Subsequently, the speech direction determination processing unit 13b determines whether or not the speech direction is a direction of the car navigation device 20 (step S114). When it is determined that the utterance direction is the direction of the car navigation device 20 (Yes in step S114), the voice recognition process ends. On the other hand, when it is not determined that the utterance direction is the direction of the car navigation device 20 (No in step S114), the process proceeds to step S115.

  In step S115, the voice recognition processing unit 13a determines whether or not the current position of the vehicle 1 is a specific area based on the GPS position information acquired from the car navigation device 20. When it is determined that the current position of the vehicle 1 is a specific area (Yes at Step S115), the process proceeds to Step S116, and when it is not determined that the current position of the vehicle 1 is a specific area (No at Step S115), The voice recognition process ends.

  In step S116, when the utterance vocabulary by the speech recognition processing unit 13a is included in the vocabulary of the specific area dictionary 12b, the command output processing unit 13c sends the command of the speech recognition result by the speech recognition processing unit 13a to the car navigation device 20. Canceling output to. This makes it possible to prevent erroneous commands to the in-vehicle device due to incorrect voice recognition.

  Next, the recognition vocabulary registration process of an Example is demonstrated. FIG. 6 is a flowchart illustrating a recognized vocabulary registration processing procedure according to the embodiment. The recognition vocabulary registration processing procedure of FIG. 6 is the same as the speech recognition processing procedure of the other embodiment shown in FIG. On the other hand, in the recognition vocabulary registration processing procedure of FIG. 6, step S115a which is a determination processing is inserted after step S115 of the speech recognition processing procedure of the other embodiment shown in FIG. 5, and the determination result is affirmative. If the determination result is negative, the process proceeds to step S116a.

  In step S115a, the speech recognition processing unit 13a determines whether or not there is a registered speech vocabulary that has been speech-recognized in the specific area dictionary 12b. When it is determined that there is a registered speech utterance vocabulary in the specific area dictionary 12b (Yes in step S115a), the command output processing unit 13c uses the command of the speech recognition result by the speech recognition processing unit 13a as a car navigation system. The output to the apparatus 20 is canceled (step S116). On the other hand, if it is not determined that the speech-recognized utterance vocabulary is registered in the specific area dictionary 12b (No in step S115a), the speech-recognized vocabulary is registered in the specific area dictionary 12b (step S116a). Through the above processing, the registered vocabulary of the specific area vocabulary dictionary 12b can be automatically increased, and the recognition accuracy of utterances outside the vehicle can be improved.

  As mentioned above, although the Example of this invention was described, this invention is not limited to this, In the range of the technical idea described in the claim, even if it implements in a various different Example, it is. It ’s good. Moreover, the effect described in the Example is not limited to this.

  In addition, among the processes described in the above embodiment, all or part of the processes described as being automatically performed can be manually performed, or the processes described as being manually performed can be performed. All or a part can be automatically performed by a known method. In addition, the processing procedure, control procedure, specific name, information including various data and parameters shown in the above embodiment can be arbitrarily changed unless otherwise specified.

  Each component of each illustrated device is functionally conceptual and does not necessarily need to be physically configured as illustrated. In other words, the specific form of distribution / integration of each device is not limited to that shown in the figure, and all or a part thereof may be functionally or physically distributed or arbitrarily distributed in arbitrary units according to various loads or usage conditions. Can be integrated and configured.

  Furthermore, each or all of the processing functions performed in each device are entirely or partially a CPU (Central Processing Unit) (or a microcomputer such as an MPU (Micro Processing Unit) or MCU (Micro Controller Unit)) and It may be realized by a program that is analyzed and executed by the CPU (or a microcomputer such as MPU or MCU), or may be realized as hardware by wired logic.

  The present invention is useful when it is desired to distinguish between an utterance to the in-vehicle device and an utterance to the outside of the vehicle and to prevent erroneous recognition of the utterance to the outside of the vehicle as an utterance to the on-vehicle device.

It is a functional block diagram which shows the structure of the speech recognition apparatus concerning an Example. It is a figure which shows the example of a specific point position information storage table. It is a figure which shows the example of a specific point utterance vocabulary table. It is a flowchart which shows the speech recognition process sequence of an Example. It is a flowchart which shows the speech recognition process sequence of another Example. It is a flowchart which shows the recognition vocabulary registration process sequence of an Example.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Vehicle 10 Voice recognition apparatus 11a Display part 11b Voice generation part 12 Storage part 12a Specific area information 12b Specific area dictionary 12c Non-specific area dictionary 13 Control part 13a Speech recognition process part 13b Speech direction determination process part 13c Command output process part 14 Microphone 20 Car navigation system

Claims (8)

Speech recognition means for recognizing the spoken vocabulary by determining whether or not a vocabulary spoken by a passenger of the vehicle is included in a speech recognition target vocabulary stored in the vocabulary storage means; and the speech recognition A speech recognition device having command conversion means for converting a result of speech recognition by the means into a command and passing it to the in-vehicle device,
The vocabulary storage means stores the speech recognition target vocabulary that is assumed to be spoken by the passenger at a specific location in the speech recognition target vocabulary in association with the position information corresponding to the specific location. And a non-specific point vocabulary storage unit for storing the speech recognition target vocabulary other than the vocabulary stored in the specific point vocabulary storage unit. Position information storage means for storing the specific point and position information corresponding to the specific point in association with each other;
The voice recognition device according to claim 1, further comprising: position information determination means for determining whether or not position information on which the vehicle is located is included in the position information storage means.   The speech recognition means stores the vocabulary spoken by the passenger when the position information determination means determines that the position information where the vehicle is located is included in the position information storage means. The spoken vocabulary is voice-recognized by determining whether it is included in the speech recognition target vocabulary stored in the means, and the position information determining means stores the position information where the vehicle is located in the position information storage Whether or not a vocabulary spoken by the passenger is included in the speech recognition target vocabulary stored in the specific point vocabulary storage unit and the non-specific point vocabulary storage unit when it is not determined to be included in the unit The speech recognition apparatus according to claim 2, wherein the spoken vocabulary is recognized by speech by determining whether or not the speech is spoken.   When it is determined that the position information storage means includes the position information where the vehicle is located, the command conversion means converts the vocabulary spoken by the passenger by the voice recognition means into the specific point vocabulary storage means. When it is determined that the vocabulary is included in the speech recognition target vocabulary stored in the memory, the result of speech recognition performed by the speech recognition means is converted into a command and transferred to the in-vehicle device is canceled. The speech recognition apparatus according to claim 2 or 3.   When it is determined that the position information storage means includes the position information where the vehicle is located, the vocabulary spoken by the passenger by the voice recognition means is stored in the specific point vocabulary storage means A recognition vocabulary registration unit that registers the spoken vocabulary in the specific point vocabulary storage unit in association with the specific point corresponding to the position information when it is not determined to be included in the speech recognition target vocabulary; The voice recognition device according to claim 4. Utterance direction determination means for determining whether the utterance direction of the spoken vocabulary by the passenger is the in-vehicle device direction;
The position information determination means determines that the position information of the vehicle is the position information only when the utterance direction determination means determines that the utterance direction of the spoken vocabulary by the passenger is not the in-vehicle device direction. 6. The speech recognition apparatus according to claim 2, wherein it is determined whether or not it is included in a storage unit. A speech recognition step for recognizing the spoken vocabulary by determining whether or not a vocabulary spoken by a vehicle occupant is included in a speech recognition target vocabulary stored in the speech recognition dictionary; and the speech recognition A voice recognition method including a command conversion step of converting a result of voice recognition by the step into a command and transferring it to the in-vehicle device,
The voice recognition dictionary is a specific point vocabulary dictionary that stores a voice recognition target vocabulary that the passenger is supposed to utter at a specific point in the voice recognition target vocabulary in association with position information corresponding to the specific point. A vocabulary dictionary outside the specific point that stores the speech recognition target vocabulary other than the vocabulary stored in the specific point vocabulary dictionary,
In the voice recognition step, the position information where the vehicle is located by the position information determination step for determining whether the position information where the vehicle is located is position information corresponding to the specific point is position information corresponding to the specific point. The vocabulary spoken by the passenger is included in the speech recognition target vocabulary stored in the off-site vocabulary dictionary to determine whether the spoken vocabulary The vocabulary spoken by the passenger when the position information determining step recognizes that the position information where the vehicle is located is not the position information corresponding to the specific point, and the specific point vocabulary dictionary and Speech recognition of the spoken vocabulary by determining whether it is included in the speech recognition target vocabulary stored in the vocabulary outside the specific location Speech recognition method according to claim.   In the command conversion step, when it is determined that the position information storage means includes the position information where the vehicle is located, the vocabulary spoken by the passenger in the voice recognition step is added to the specific point vocabulary dictionary. When it is determined that the vocabulary is included in the stored speech recognition target vocabulary, the result of speech recognition by the speech recognition step is converted into a command and the transfer to the in-vehicle device is canceled. Item 8. The speech recognition method according to Item 7.

Images