JP5556529B2 - In-vehicle speech recognition device - Google Patents

Description

  The present invention relates to an in-vehicle voice recognition device that sends a control signal corresponding to a voice command that has been voice-recognized to a control target device.

  Conventionally, in an in-vehicle information terminal, many HMIs (Human Machine Interfaces) using a voice recognition function have been adopted in order to improve safety and convenience.

  As such an in-vehicle information terminal, for example, there is one that recognizes a voice command from a voice spoken by an occupant using a voice recognition function and performs an operation according to a recognition result (for example, see Patent Document 1). .

JP 2003-152848 A

  In an apparatus such as that described in Patent Document 1, the operations for voice commands that have been voice-recognized are uniformly defined. That is, there is a one-to-one relationship between a voice command that has been voice-recognized and an operation corresponding to this voice command.

  However, in the configuration in which the voice command recognized in this way and the operation corresponding to the voice command have a one-to-one relationship, only one operation can be performed with one voice command. There is a problem that a highly functional function cannot be realized.

  The present invention has been made in view of the above problems, and an object thereof is to realize a more convenient function.

In order to achieve the above object, according to the first aspect of the present invention, a voice command is recognized from a voice signal input via the voice signal input means, and a control signal corresponding to the voice command is sent to the control target device. One of the voice signal input means is a vehicle interior microphone, and one of the voice signal input means is a portable communication device connected via a communication interface, and the voice signal input A control signal sending means for distinguishing a voice signal input through the means according to a predetermined condition, sending a control signal according to a voice command according to the distinction of the voice signal to a control target device, and a voice signal Storage means for storing an information table that defines different control signals according to voice commands according to the distinction of voice signal input means to Output means, characterized by referring to the information table stored in the storage unit, that transmits to the control target device by varying a control signal corresponding to the voice command according to the distinction between the speech signal input means to which the audio signal is inputted It is said.

According to such a configuration, the audio signal input via the audio signal input unit is distinguished according to a predetermined condition, and the control target device is differentiated according to the audio command according to the audio signal distinction. Therefore, even for one voice command, different control signals can be sent to the control target device according to the distinction of the voice signal, and a more convenient function can be realized.
In addition, referring to the information table in which the control signal according to the voice command is defined differently according to the distinction of the voice signal input means to which the voice signal is inputted, the voice command according to the distinction of the voice signal input means to which the voice signal is inputted The control signal corresponding to the signal can be made different and sent to the control target device.
Further, one of the audio signal input means can be a vehicle interior microphone, and one of the audio signal input means can be a portable communication device connected via a communication interface.

In the invention according to claim 3, the audio signals are separately input via the plurality of audio signal input means, and the control signal sending means is input via the audio signal input means. The audio signal is distinguished for each audio signal input means to which the audio signal is input.

  In this way, the audio signal input through the audio signal input means is distinguished for each audio signal input means to which the audio signal is input, and control is performed by varying the control signal corresponding to the audio command according to the audio signal distinction. It can be sent to the target device.

According to a second aspect of the present invention, when it is determined by the occupant presence / absence determining means for determining the presence / absence of an occupant in the vehicle interior and the occupant presence / absence determining means, the microphone of the portable communication device is used. Control signal transmission prohibiting means for prohibiting transmission of a control signal in accordance with a voice command recognized by voice from a voice signal input via the voice signal.

According to such a configuration, when it is determined that there is an occupant in the vehicle compartment, transmission of a control signal according to a voice command recognized by voice from a voice signal input via the microphone of the portable communication device is prohibited. Therefore, it is possible to prevent the problem that the setting of the control target device is changed by the user of the portable communication device against the intention of the passenger in the passenger compartment.

It is a figure which shows the block configuration of the vehicle-mounted speech recognition apparatus which concerns on 1st Embodiment of this invention. It is a figure for demonstrating the structure of an information table. It is a flowchart which shows the process of the control part which concerns on 1st Embodiment of this invention. It is a figure which shows the block configuration of the vehicle-mounted speech recognition apparatus which concerns on 2nd Embodiment of this invention. It is a figure for demonstrating also about the structure of a database. It is a flowchart which shows the process of the control part which concerns on 2nd Embodiment of this invention.

(First embodiment)
FIG. 1 shows a block configuration of the in-vehicle speech recognition device according to the first embodiment of the present invention. The in-vehicle voice recognition device 1 includes a microphone 10, a communication unit 12, an in-vehicle communication control unit 13, a display 14, a speaker 15, and a control unit 16. The in-vehicle voice recognition device 1 is connected to an air conditioner ECU 2, an audio ECU 3, and the like via an in-vehicle LAN 5 connected to the in-vehicle communication control unit 13.

  The microphone 10 is for collecting sound in the vehicle interior and sends out an audio signal corresponding to the sound in the vehicle interior to the control unit 16.

  The communication unit 12 is for communicating with an external communication device via a wireless communication network. The in-vehicle voice recognition device 1 can communicate with the mobile phone 4 via the communication unit 12.

  The in-vehicle communication control unit 13 is for communicating with control target devices such as the air conditioner ECU 2 and the audio ECU 3 connected via the in-vehicle LAN 5.

  The control unit 16 can control controlled devices such as the air conditioner ECU 2 and the audio ECU 3 through the in-vehicle communication control unit 13.

  The display 14 has a display unit such as a liquid crystal, and displays a video corresponding to the video signal input from the control unit 16 on the display unit.

  The speaker 15 is attached to the vehicle interior and outputs sound corresponding to the sound signal input from the control unit 16.

  The control unit 16 is configured as a computer including a speech recognition engine 16a for performing speech recognition processing, a storage unit 16b including a ROM, a RAM, and a flash memory, and a CPU (not shown). The CPU performs various processes according to the program stored in the storage unit 16b.

  The control unit 16 recognizes a voice command from a voice signal input via the microphone 10 or the mobile phone 4, and sends a control signal corresponding to the voice command to the control target devices 2 and 3.

  An audio signal is input from the microphone 10 and an audio signal is input from the mobile phone 4 to the control unit 16 of the in-vehicle voice recognition device 1. The control unit 16 can recognize a voice command from a voice signal input from the microphone 10 or can recognize a voice command from a voice signal input from the mobile phone 4.

  The control unit 16 in the present embodiment distinguishes between the voice signal input from the microphone 10 and the voice signal input from the mobile phone 4, and the voice command recognized from the voice signal input from the microphone 10 and the mobile phone 4. Even if the voice command recognized from the input voice signal is the same, the control signal sent in response to the voice command recognized from the voice signal inputted from the microphone 10 and the voice signal inputted from the mobile phone 4 Thus, it is possible to make the control signal to be transmitted different in accordance with the voice command recognized from the voice.

  The storage unit 16b of the control unit 16 stores an information table that defines an operation according to the voice command according to the distinction of the voice signal, that is, a control signal transmitted according to the voice command.

  The control unit 16 refers to the information table stored in the storage unit 16b, and transmits a control signal sent in response to a voice command recognized from a voice signal input from the microphone 10, and a voice signal input from the mobile phone 4. The control signal to be sent out in response to the voice command recognized from the voice is sent out differently.

  The configuration of the information table is shown in FIG. As shown in this figure, the operation for each voice command is defined for each input interface in the information table.

  For example, regarding the operation for the voice command “air conditioner on”, when the input interface is the microphone 10, it is defined that “the air conditioner is turned on by blowing air from the blowing portion at the previous use”, and the input interface is the mobile phone 4. Stipulates that “the air conditioner is turned on by blowing air from all the air blowing locations”.

  Therefore, when “air conditioner on” is uttered toward the microphone 10 in the passenger compartment, a control signal is sent to the air conditioner ECU 2 so that the air conditioner is turned on by blowing air from the blowing area at the previous use of the vehicle. If the mobile phone 4 is used for communication connection to the in-vehicle speech recognition device 1 from a place away from the vehicle before getting into the vehicle and “air-conditioner is turned on” is uttered, the air conditioner is blown from all the air blowing locations of the vehicle. A control signal that turns on is sent to the air conditioner ECU 2.

  Further, regarding the operation for the voice command “information”, it is defined that “information such as traffic information and gasoline price” is provided when the input interface is the microphone 10, and “key is locked” when the input interface is the mobile phone 4. "Provides information such as whether it is hung and the open / closed state of the window".

  Therefore, when “information” is uttered toward the microphone 10 in the passenger compartment, a control signal for displaying information such as traffic information and gasoline price is sent to the audio ECU 3, for example, from a place away from the vehicle. When the mobile phone 4 is connected to the in-vehicle speech recognition apparatus 1 and utters “information”, information such as “whether it is locked”, “open / closed state of window”, etc. is notified to the mobile phone 4 Is done.

  Next, processing of the control unit 16 will be described with reference to FIG. The vehicle-mounted speech recognition device 1 operates in the normal mode when the vehicle ignition switch is turned on, and operates in the low power consumption mode when the vehicle ignition switch is turned off. When there is an operation of pressing a PTT (Push Talk Switch) for instructing the start of voice recognition or an incoming call from an external device, the control unit 16 performs the process shown in FIG. 3 regardless of the operation mode.

  First, the input interface is determined (S100). In the present embodiment, it is determined whether the microphone 10 or the mobile phone 4 is used as the input interface.

  Here, when the microphone 10 is used as the input interface, the determination in S100 is “microphone”, and then the microphone is set as the input interface (S102). Specifically, a flag indicating that the input interface is a microphone is set in a predetermined area of the RAM.

  Next, the guidance for prompting command utterance is reproduced (S104). For example, a guidance such as “Please speak a voice command” is output from the speaker 15.

  If the user utters according to this guidance, then a speech recognition process is performed (S106). For example, when the user speaks “air conditioner on”, a voice command “air conditioner on” is recognized.

  Next, a control signal corresponding to the recognition result is sent (S108). Specifically, referring to the flag set in S102, it is discriminated whether the microphone 10 or the mobile phone 4 is used as an input interface, and an audio signal is input by referring to the information table. The control signal corresponding to the voice command is made different according to the distinction of the input interface to be transmitted to the control target device.

  As described above, when the microphone 10 is used as the input interface, and the voice command “air conditioner on” is recognized in S106, the air conditioner is turned on by the air blowing from all the air blowing portions of the vehicle. A signal is sent to the air conditioner ECU 2.

  In addition, when the microphone 10 is used as an input interface and a voice command “information” is recognized in S106, a control signal for displaying information such as traffic information and gasoline price is sent to the audio ECU 3. This process is terminated.

  When the mobile phone 4 is used as an input interface, the determination in S100 is “mobile phone”, and then it is determined whether or not there is even one passenger (S110). Specifically, seating sensors for detecting seating of occupants on the seating surface of each seat of the vehicle are provided, and whether or not there is even one passenger based on signals output from these seating sensors. Determine whether.

  Here, if there is no passenger, the determination in S110 is NO, and then the mobile phone 4 is set as an input interface (S112). Specifically, a flag indicating that the input interface is the mobile phone 4 is set in a predetermined area of the RAM, and the process proceeds to S104.

  In this case, the guidance for prompting the command utterance is reproduced in S104. According to this guidance, for example, when the user utters “air conditioner on”, the voice command “air conditioner on” is recognized and the recognition result is obtained in S108. A corresponding control signal is sent out.

  Here, when the cellular phone 4 is used as an input interface, and the voice command “air conditioner on” is recognized in S106, a control signal that turns on the air conditioner by blowing air from all the blowing locations of the vehicle. Is sent to the air conditioner ECU 2.

  In addition, when the mobile phone 4 is used as an input interface and the voice command “Information” is recognized in S106, information such as “whether it is locked”, “open / closed state of window”, etc. is carried. The telephone 4 is notified.

  If the mobile phone 4 is used as the input interface and there is even one passenger, the determination in S110 is YES, and access by the mobile phone is prohibited (S114). Specifically, transmission of a control signal according to a voice command according to the voice collected by the microphone of the mobile phone 4 is prohibited. Thereby, it is prevented that the setting etc. of a control object apparatus will be changed without permission by the other person who owns the mobile telephone 4 against the will of the passenger in a vehicle interior.

  Next, the voice recognition end guidance is reproduced (S116), and this process is ended.

  According to the above-described configuration, the audio signal input via the audio signal input unit is distinguished according to a predetermined condition, and the control signal according to the audio command is changed according to the audio signal to the control target device. Since it is transmitted, even if it is one voice command, different control signals can be sent to the control target device according to the distinction of the voice signal, and a more convenient function can be realized.

  Specifically, the input audio signal is distinguished for each audio signal input means to which the audio signal is input, and the control signal corresponding to the audio command is differentiated according to the audio signal and sent to the control target device. Can do.

  In addition, referring to the information table in which the control signal according to the voice command is defined differently according to the distinction of the voice signal input means to which the voice signal is inputted, the voice command according to the distinction of the voice signal input means to which the voice signal is inputted The control signal corresponding to the signal can be made different and sent to the control target device.

  Further, one of the audio signal input means can be a vehicle interior microphone, and one of the audio signal input means can be a portable communication device connected via a communication interface.

  In addition, when it is determined that there is an occupant in the passenger compartment, transmission of a control signal according to a voice command recognized by voice from an audio signal input via the microphone of the portable communication terminal is prohibited. It is possible to prevent the problem that the setting of the control target device is changed by the user of the portable communication terminal against the intention of the passenger.

(Second Embodiment)
The configuration of the in-vehicle speech recognition device according to this embodiment is shown in FIG. The in-vehicle voice recognition device 1 includes a microphone 10, an in-vehicle communication control unit 13, a display 14, a speaker 15, and a control unit 16. The in-vehicle voice recognition device 1 can communicate with the mobile phone 4 via the communication unit 12. In addition, the same code | symbol is attached | subjected to the same part as the said 1st Embodiment, description is abbreviate | omitted and it demonstrates centering on a different part hereafter.

  In the present embodiment, an audio signal output from the microphone 10 is directly input to the control unit 16.

  In the first embodiment, the input voice signal is distinguished depending on whether it is input from the microphone 10 or input from the mobile phone 4, and the control signal corresponding to the voice command is determined according to this distinction. In this embodiment, the input audio signal is analyzed to distinguish voice tones, and the control signal sent to the control target device according to the voice command is different according to this distinction. It comes to let you.

  The control unit 16 according to the present embodiment analyzes a voice signal input via the microphone 10 to distinguish a voice tone, and stores and stores the distinguished voice tone and the setting contents of the control target device in association with each other. The stored database 16b is stored and stored in the storage unit 16b, the setting content set in the past with the same tone as the analyzed voice tone is specified with reference to the database, and the specified operation content is specified. A process of sending a control signal indicating a voice command indicating to the control target device is performed. This process will be described later in detail.

  FIG. 5 shows the configuration of the database. As shown in the figure, a database in which voice commands are classified into a plurality of tone patterns and associated with voice commands and operation histories is created for each user.

  The tone pattern represents the tone level of the voice. In the present embodiment, the tone pattern of the voice command is specified by analyzing the frequency of the voice command. The voice command “air conditioner on” is classified into three tone patterns A, B, and C, and the voice command and the operation history are associated with each other and accumulated and stored in the database. , B are classified into two tone patterns, and voice commands and operation histories are associated and stored in a database. Note that the tone decreases in the order of tone patterns A, B, and C.

  For example, when the user X utters the voice command “set temperature change” and “25 ° C.” indicating the input of the set temperature of the air conditioner following the voice command “air conditioner on” with a tone classified into the tone pattern A, As shown in FIG. 5, the number of registrations of the items of user X, tone pattern A, and operation history 25 ° C. is increased by one for the voice command “air conditioner ON”.

  Further, when the user X is in a poor physical condition and the voice command “set temperature change” and “28 ° C.” are uttered by the tone classified into the tone pattern B following the voice command “air conditioner ON”, it is shown in FIG. Thus, the number of registrations of the items of user X, tone pattern B, and operation history 28 ° C. is increased by one for the voice command “air conditioner ON”.

  When the user is Mr. Y, a database in which the voice command of the user Y and the operation history are associated is stored.

  For the voice command “audio on”, a database in which the voice command is classified into tone patterns A and B for each user and the voice command and the operation history are associated is created.

  Next, processing of the control unit 16 will be described. FIG. 6 shows a flowchart of the control unit 16. When a PTT (Push Talk Switch) for instructing the start of speech recognition is pressed, the control unit 16 according to the present embodiment performs the process illustrated in FIG.

  First, a guidance for prompting a user name is generated (S200). For example, a guidance such as “Please speak your user name” is output from the speaker 15.

  When the user name is uttered according to this guidance, a voice recognition process is performed (S202). For example, when the user utters “X”, the user name “X” is recognized by voice.

  Next, a user to be used is set according to the user name that has been recognized (S204). When the user name “X” is recognized by voice in S202, the user in use is registered as “X”.

  Next, guidance for prompting command utterance is reproduced (S206). For example, a guidance such as “Please speak a voice command” is output from the speaker 15.

  When the user utters according to this guidance, next, speech recognition processing is performed (S208). For example, when a voice command “air conditioner on” is uttered, the voice command “air conditioner on” is recognized.

  Next, the voice signal input from the microphone 10 is analyzed to detect the tone of the voice command (S210). Here, the tone pattern of the voice command “air conditioner on” is specified.

  Next, it is determined whether or not there is an operation history corresponding to the tone pattern whose registration count exceeds a specified value (S212). Specifically, the voice command recognized in S208 with reference to the database and the preset temperature registered with the tone pattern specified in S210 and registered exceeding the specified value. It is determined whether or not there is.

  Here, if there is no voice command recognized in S208 and the set temperature registered with the tone pattern specified in S210 that exceeds the specified value, the determination in S212 is NO. Thus, the set temperature of the air conditioner is set to the standard temperature setting (S220). In this embodiment, 25 ° C. is the standard temperature.

  Next, it is determined whether there is an instruction to change the set temperature (S222). Specifically, it is determined whether or not there is an instruction to change the set temperature based on whether or not the voice command “set temperature change” has been recognized.

  Here, when the voice command “change in set temperature” is uttered and the voice command “change in set temperature” is voice-recognized, the determination in S222 is YES, and then the voice recognition process is performed (S224). Here, if the user X utters “27 ° C.”, the voice command “27 ° C.” is recognized.

  Next, the tone of the voice command uttered by the user X is detected (S226). Here, the tone pattern of the voice command “27 ° C.” is specified. Thus, apart from S210, the tone pattern of the voice command is specified again to increase the accuracy of the voice command tone detection.

  Next, the tone pattern and the set temperature are associated and registered in the database (S228). For example, when the voice command “27 ° C.” is recognized in S224 and the tone pattern is specified as A in S226, the tone pattern A and the set temperature “27 ° C.” are set to the user X. Associate and register in the database. If the same condition is already registered in the database, 1 is added to the number of registrations under the same condition.

Next, a control signal corresponding to the recognition result is transmitted to the air conditioner ECU 2 (S218). Here, a control signal is sent to the air conditioner ECU 2 so that the set temperature of the air conditioner becomes 27 ° C.

  If the voice command “change set temperature” is not uttered within a predetermined time, the determination in S222 is NO, and the tone pattern detected in S210 is registered in the database in association with the standard temperature (S230). For example, when the tone pattern of the voice command “air conditioner on” is specified as C in S210, the tone pattern is registered in the database in association with the standard temperature (25 ° C.).

  Next, a control signal corresponding to the recognition result is transmitted to the air conditioner ECU 2 (S218). Here, the set temperature is the standard temperature (25 ° C.), and a control signal for turning on the air conditioner is transmitted to the air conditioner ECU 2.

  In S208, the voice command “air conditioner on” by the user X is recognized as voice, and in S210, the tone pattern of the voice command “air conditioner on” is specified, and the voice command recognized in S208 in the database, If there is a set temperature registered with the tone pattern specified in S210 that exceeds the specified value, the determination in S212 is YES, and then the highest set temperature is set. (S213). That is, the highest set temperature among the registered values exceeding the specified value is set.

  Therefore, in the case of the database configuration shown in FIG. 5, when the user X utters the voice command “air conditioner ON” with a tone classified as tone pattern A, the preset temperature is 25 ° C. Set.

  Further, when the user X utters a voice command “air conditioner ON” with a tone classified into the tone pattern B, the set temperature is set to 28 ° C., which is the specified value or more and the most frequently registered, in the tone pattern B. .

  Next, it is determined whether there is a change in the set temperature (S214). Here, if the set temperature is changed by a user operation, the determination in S214 is YES, and the process proceeds to S224. If the set temperature is not changed, the determination in S214 is NO, and the database is then updated by associating the voice command with the tone pattern (S215). For example, for the user X, a voice command “air conditioner ON” and a tone pattern are associated with each other and registered in the database. If the same condition is already registered in the database, 1 is added to the number of registrations under the same condition.

  Next, a control signal corresponding to the recognition result is transmitted to the air conditioner ECU 2 (S218). When the user X utters a voice command “air conditioner on” with a tone classified as the tone pattern A, the control signal for turning on the air conditioner is transmitted to the air conditioner ECU 2 at a set temperature of 25 ° C. When the user X utters the voice command “air conditioner on” with a tone classified as the tone pattern B, the control signal for turning on the air conditioner is transmitted to the air conditioner ECU 2 with the set temperature set at 28 ° C.

  As described above, the audio signal specifying unit that analyzes the audio signal input through the audio signal input unit and specifies the audio feature is provided, and the audio signal input through the audio signal input unit is converted into the audio feature. It is possible to discriminate according to the voice feature specified by the specifying means, and to send the control signal according to the voice command differently according to the voice feature distinction and send it to the control target device.

  In addition, a database is created in which the audio features specified by the audio feature specifying means are associated with the setting contents of the control target device. The database is referred to, and the same classification as the audio features specified by the audio feature specifying means is made. A control signal indicating a voice command indicating the setting content of the control target device associated with the feature is transmitted to the control target device. That is, a database that automatically associates the features of the voice command that the user has uttered in the past with the settings of the control target device set at that time is automatically created. It is possible to send a control signal indicating a voice command indicating the setting contents of the control target device associated with the feature of the same category as the specified voice feature to the control target device.

  Moreover, even if the occupant who utters a voice command changes to another occupant, according to the configuration described above, a database is created for each occupant, and information on the occupant who utters a voice command is read from the database and analyzed. Since a control signal indicating a voice command indicating the setting contents of the control target device associated with the same classification as that of the feature is sent to the control target device, the voice of the occupant is confused with the voice of another occupant so as to cause a malfunction. It is possible to prevent this.

  In addition, this invention is not limited to the said embodiment, Based on the meaning of this invention, it can implement with a various form.

  For example, in the above-described embodiment, the air conditioner ECU 2 and the audio ECU 3 have been described as control target devices. However, these ECUs are not limited, and for example, another ECU or the mobile phone 4 may be the control target device.

  In the first embodiment, the configuration is shown in which one of the audio signal input means is a microphone of a mobile phone connected via a communication interface. However, the present invention is not limited to a mobile phone. Type communication device.

  Moreover, although the structure which communicates with the mobile telephone 4 via a wireless communication network was shown in the said 1st Embodiment, you may make it communicate with the mobile telephone 4 via interfaces, such as Bluetooth, for example.

  In the second embodiment, the input audio signal is distinguished according to the audio tone. However, the input audio signal is not limited to the audio tone, and various audio signals such as an audio volume, a combination of audio volume and tone, and the like are used. You may make it distinguish according to the characteristic.

  The correspondence relationship between the configuration of the above embodiment and the configuration of the claims will be described. The microphone 10 and the microphone of the mobile phone 4 correspond to audio signal input means, and the air conditioner ECU 2 and the audio ECU 3 correspond to control target devices. S100 to S108 and S218 correspond to control signal transmission means, the storage unit 16b corresponds to storage means, S110 corresponds to passenger presence determination means, S114 corresponds to control signal transmission prohibition means, and S210 and S226. Corresponds to voice feature specifying means, S222 to S230 correspond to database creation processing means, and S200 to S204 correspond to occupant specifying means.

1 On-vehicle speech recognition device 2 Air conditioner ECU
3 Audio ECU
4 Mobile phone 10 Microphone 12 Communication unit 13 In-vehicle communication control unit 14 Display 15 Speaker 16 Control unit 16a Speech recognition engine 16b Storage unit

Claims (3)

An in-vehicle voice recognition device that recognizes a voice command from a voice signal input via a voice signal input unit and sends a control signal corresponding to the voice command to a control target device,
One of the audio signal input means is a vehicle interior microphone, and one of the audio signal input means is a portable communication device connected via a communication interface,
Control for distinguishing the voice signal input through the voice signal input means according to a predetermined condition, and sending a control signal according to the voice command to the control target device according to the distinction of the voice signal A signal transmission means ;
Storage means for storing an information table defining different control signals according to the voice command according to the distinction of the voice signal input means to which the voice signal is input,
The control signal sending means refers to the information table stored in the storage means, and changes the control signal according to the voice command according to the distinction of the voice signal input means to which the voice signal is inputted. A vehicle-mounted speech recognition apparatus characterized by being sent to a control target device. Occupant presence / absence determining means for determining the presence / absence of an occupant in the passenger compartment;
When it is determined by the occupant presence / absence determination means that there is an occupant in the passenger compartment, a control signal corresponding to the voice command recognized by the voice signal input from the microphone signal of the portable communication device is transmitted. The vehicle-mounted speech recognition apparatus according to claim 1 , further comprising a control signal transmission prohibiting unit that prohibits the signal. The audio signals are input separately via a plurality of the audio signal input means,
Said control signal transmitting means, said voice signal inputted through the sound signal input means, to claim 1 or 2, characterized in that to distinguish for each of the audio signal input means for the audio signal is input The vehicle-mounted speech recognition apparatus described.

Images