JP6912985B2 - Speech recognition system and computer program - Google Patents

Description

The present invention relates to a voice recognition technique for recognizing a user's spoken voice.

As a voice recognition technology for recognizing a user's voice, a first voice recognition that recognizes a user's voice picked up by a microphone in a system equipped with an audio device that outputs the voice of audio content such as music from a speaker. In addition to the unit, a second voice recognition unit that performs voice recognition of the voice output from the audio device to the speaker is provided, and the result of voice recognition by the first voice recognition unit and the result of voice recognition by the second voice recognition unit. There is known a technique for invalidating the result of voice recognition by the first voice recognition unit when the above is the same (for example, Patent Document 1).

According to such a first technique, the voice recognition result of the first voice recognition unit, which is also performed for the output voice of the audio device that wraps around the microphone, is used as the voice recognition result of the user's uttered voice. It is possible to prevent misidentification.

In addition, as a voice recognition technology for recognizing a user's voice, in a system in which a set of commands that are candidates for voice input changes depending on a command input last time, the first voice recognition unit is used as a candidate for voice input this time. While performing voice recognition of the command in the set of commands, the second voice recognition unit performs voice recognition of the command in the set of commands that was a candidate for voice input last time, so that the user's command can be rephrased. Accepting techniques are also known (for example, Patent Document 2).

Jitsuto 2602342 International Publication No. 2011/016129

By the way, in a system for voice input of commands, it is preferable to be able to accept voice input of many commands at each time point. In particular, it is preferable that the voice input of a command requesting the execution of an urgent process for the user can be accepted at any time.

However, in general, the number of commands that can be practically recognized by a voice recognition unit is limited.
On the other hand, in a system including two voice recognition units, a first voice recognition unit and a second voice recognition unit, the first voice recognition unit and the second voice recognition unit are used as in the technique of Patent Document 1 described above. If the recognition unit performs voice recognition of commands in different sets of commands, it becomes possible to accept voice input of more commands at each time point than when only one voice recognition unit is used. Then, it becomes impossible to prevent the command from being erroneously recognized for the output voice of the audio device that wraps around the microphone by applying the technique of Patent Document 2 described above.

Therefore, according to the present invention, in a voice recognition system including two voice recognition units, a first voice recognition unit and a second voice recognition unit, a command or the like can be used for the output voice of an audio device that wraps around the microphone. The challenge is to increase the number of words that can accept voice input as much as possible while suppressing misidentification of words.

In order to achieve the above object, the present invention provides a voice recognition system that recognizes a voice uttered in a space in which the voice output from an audio source device is radiated from the speaker to the speaker, and a microphone and a word are used. A plurality of registered first speech recognition dictionaries and voices picked up by the microphone are input, and a word matching the input voice is detected as a target candidate from a plurality of words registered in the first voice recognition dictionary. A first voice recognition unit, a second voice recognition unit, a second voice recognition dictionary used by the second voice recognition unit, a recognition unit that recognizes a word spoken by a user, a first recognition mode, and a second. It is provided with a voice input receiving unit that accepts voice input of a word recognized by the recognition unit while selectively setting the recognition mode of. Here, in the second voice recognition dictionary used by the second voice recognition unit in the first recognition mode, the same plurality of words as the plurality of words registered in the first voice recognition dictionary are registered. In the first recognition mode, the second voice recognition unit inputs the voice output to the speaker by the audio source device, and inputs from a plurality of words registered in the second voice recognition dictionary. A word matching the voice is detected as a target candidate, and in the first recognition mode, the recognition unit serves as the target candidate during a predetermined period after the target candidate is detected by the second voice recognition unit. Even if the same word as the detected word is detected as the target candidate by the first voice recognition unit, the word detected as the target candidate by the first voice recognition unit is recognized as the word spoken by the user. While suppressing it, the word detected as the target candidate by the first voice recognition unit is recognized as the word spoken by the user. In addition, the second speech recognition dictionary used by the second speech recognition unit in the second recognition mode includes a plurality of words registered in the first speech recognition dictionary and a plurality of words that are at least partially different from the plurality of words registered in the first speech recognition dictionary. It is registered, and in the second recognition mode, the second voice recognition unit inputs the voice picked up by the microphone and inputs from a plurality of words registered in the second voice recognition dictionary. A word matching the voice is detected as a target candidate, and in the second recognition mode, the recognition unit detects the target candidate as the target candidate when the target candidate is detected by the first voice recognition unit. Is recognized as a word spoken by the user, and when the target candidate is detected by the second voice recognition unit, the word detected as the target candidate is recognized as the word spoken by the user.

According to the above-mentioned voice recognition system, the second voice recognition dictionary is used to perform voice recognition in a form that suppresses erroneous recognition of words registered in the first voice recognition dictionary by the output voice of the audio source device. Since it has a recognition mode of 2 and a second recognition mode in which the second voice recognition dictionary is used in parallel with the first voice recognition dictionary to perform voice recognition, it can be used as the first voice recognition dictionary based on the output voice of the audio source device. In a situation where erroneous recognition of registered words is likely to occur, the first recognition mode is set to perform voice recognition, and erroneous recognition of words registered in the first voice recognition dictionary by the output voice of the audio source device is unlikely to occur. Under circumstances, a second recognition mode can be set to increase the number of words that can be voice recognized. Therefore, it is possible to increase the number of words that can accept voice input while effectively suppressing erroneous recognition of words registered in the first voice recognition dictionary by the output voice of the audio source device.

More specifically, for example, in such a voice recognition system, the first recognition mode is set as the recognition mode in the voice input reception unit, and the standby voice recognition dictionary is a predetermined voice recognition dictionary. Is set in the first voice recognition dictionary and the second voice recognition dictionary, the standby process for accepting the voice input of the word recognized by the recognition unit is performed, and the voice input of the word is received by the standby process. The sequence execution process in which the word that receives the voice input is set as the first layer word may be started. Here, in the sequence execution process, the second recognition mode is set as the recognition mode, and at each time point, one of the first voice recognition dictionary and the second voice recognition dictionary is the standby voice recognition dictionary. While using the first speech recognition dictionary and the second speech recognition dictionary so that the sub-speech recognition dictionary is set to and the other is the main speech recognition dictionary, a speech recognition dictionary determined according to the first layer word is set. While updating the main voice recognition dictionary to a voice recognition dictionary determined according to the words registered in the main voice recognition dictionary that received the voice input, the main voice recognition dictionary was registered in the main voice recognition dictionary recognized by the recognition unit. This is a process of executing a sequence in which a word voice input is accepted once or a plurality of times. Further, the voice input receiving unit starts the standby process when the sequence execution process is completed, and during the execution of the sequence execution process, the voice input of the word registered in the standby voice recognition dictionary is performed. Is accepted, at least one of the start of the sequence execution process in which the word that received the voice input is set as the first layer word and the start of other processes associated with the word that received the voice input are performed. ..

Alternatively, in such a voice recognition system, the first recognition mode is set as the recognition mode in the voice input receiving unit, and the standby voice recognition dictionary, which is a predetermined voice recognition dictionary, is used for the first voice recognition. It is set in the dictionary and the second voice recognition dictionary, and the standby process for accepting the voice input of the word recognized by the recognition unit is performed. If the voice input of the word is accepted by the standby process, the voice input is accepted. It may be configured to start the sequence execution process in which the word is the first layer word. Here, in the sequence execution process, the first voice recognition dictionary and the first voice recognition dictionary in which the second recognition mode is set as the recognition mode and different voice recognition dictionaries are set according to the first layer word are set. 2 A process of executing a sequence of accepting voice input of a word recognized by the recognition unit once or a plurality of times while updating the voice recognition dictionary to different voice recognition dictionaries determined according to the words received as voice input. Is. Further, the voice input receiving unit starts the standby process when the sequence execution process is completed.

Alternatively, in such a voice recognition system, the first recognition mode is set as the recognition mode in the voice input receiving unit, and the standby voice recognition dictionary, which is a predetermined voice recognition dictionary, is used for the first voice recognition. It is set in the dictionary and the second voice recognition dictionary, and the standby process for accepting the voice input of the word recognized by the recognition unit is performed, and the voice output from the audio source device is emitted from the speaker by the standby process. If the voice input of the word is accepted while the voice input is being performed, the first sequence execution process in which the word for which the voice input is accepted is set as the first layer word is started, and the audio source is transmitted from the speaker by the standby process. If the voice input of a word is accepted when the voice output from the device is not radiated, the second sequence execution process in which the word receiving the voice input is set as the first layer word is configured to start. You may. Here, in the first sequence execution process, the first speech recognition dictionary and the second speech recognition dictionary in which the first recognition mode is set as the recognition mode and the same speech recognition dictionary determined according to the first layer word is set. It is a process of executing a sequence of accepting the voice input of the word recognized by the recognition unit once or a plurality of times while updating the voice recognition dictionary to the same voice recognition dictionary determined according to the word that received the voice input. In the second sequence execution process, the second recognition mode is set as the recognition mode, and at each time point, one of the first voice recognition dictionary and the second voice recognition dictionary uses the standby voice recognition dictionary. While using the first speech recognition dictionary and the second speech recognition dictionary so that the set sub-speech recognition dictionary and the other become the main speech recognition dictionary, a speech recognition dictionary determined according to the first layer word was set. While updating the main voice recognition dictionary to a voice recognition dictionary determined according to the words registered in the main voice recognition dictionary that received voice input, the words registered in the main voice recognition dictionary recognized by the recognition unit. This is a process of executing a sequence in which the voice input of the above is accepted once or a plurality of times. Further, the voice input receiving unit starts the standby process when the first sequence execution process is completed, and starts the standby process when the second sequence execution process is completed, and the second sequence execution process is completed. If the voice input of the word registered in the standby voice recognition dictionary is accepted during the execution of the sequence execution process, the start of the sequence execution process in which the word receiving the voice input is set as the first layer word. , At least one of the start of other processing associated with the word that received the voice input.

Alternatively, in such a voice recognition system, the first recognition mode is set as the recognition mode in the voice input receiving unit, and the standby voice recognition dictionary, which is a predetermined voice recognition dictionary, is used for the first voice recognition. It is set in the dictionary and the second voice recognition dictionary, and the standby process for accepting the voice input of the word recognized by the recognition unit is performed, and the voice output from the audio source device is emitted from the speaker by the standby process. If the voice input of the word is accepted while the voice input is being performed, the first sequence execution process in which the word for which the voice input is accepted is set as the first layer word is started, and the audio source is transmitted from the speaker by the standby process. If the voice input of a word is accepted when the voice output from the device is not radiated, the second sequence execution process in which the word receiving the voice input is set as the first layer word is configured to start. You may. Here, in the first sequence execution process, the first speech recognition dictionary and the second speech recognition dictionary in which the first recognition mode is set as the recognition mode and the same speech recognition dictionary determined according to the first layer word is set. It is a process of executing a sequence of accepting the voice input of the word recognized by the recognition unit once or a plurality of times while updating the voice recognition dictionary to the same voice recognition dictionary determined according to the word that received the voice input. In the second sequence execution process, the first speech recognition dictionary and the second speech recognition dictionary in which the second recognition mode is set as the recognition mode and different speech recognition dictionaries are set according to the first layer word are set. In the process of updating the voice recognition dictionary to different voice recognition dictionaries determined according to the word that received the voice input, and executing a sequence that accepts the voice input of the word recognized by the recognition unit once or multiple times. be. Further, the voice input receiving unit starts the standby process when the first sequence execution process is completed, and starts the standby process when the second sequence execution process is completed.

Alternatively, in the above voice recognition system, the voice input receiving unit sets the recognition mode to the first recognition mode when the voice output from the audio source device is emitted from the speaker, and the voice recognition mode is set to the first recognition mode. The recognition mode may be set to the second recognition mode when the sound output from the audio source device is not emitted from the speaker.

Here, the voice recognition system as described above may be a voice recognition system used for voice input in an in-vehicle system mounted on an automobile.

As described above, according to the present invention, in a voice recognition system including two voice recognition units, a first voice recognition unit and a second voice recognition unit, the output voice of the audio device that wraps around the microphone can be obtained. On the other hand, it is possible to increase the number of words that can accept voice input as much as possible while suppressing misidentification of words such as commands.

It is a block diagram which shows the structure of the information processing system which concerns on embodiment of this invention. It is a figure which shows the recognition data which concerns on embodiment of this invention. It is a flowchart which shows the voice input control processing which concerns on embodiment of this invention. It is a flowchart which shows the voice input setting process which concerns on embodiment of this invention. It is a figure which shows the voice recognition method of the voice recognition engine which concerns on embodiment of this invention. It is a flowchart which shows the audio cancel mode recognition process which concerns on embodiment of this invention. It is a flowchart which shows the parallel recognition mode recognition process which concerns on embodiment of this invention. It is a figure which shows the transition example of the display screen and a voice recognition dictionary which concerns on embodiment of this invention. It is a figure which shows the transition example of the display screen and a voice recognition dictionary which concerns on embodiment of this invention. It is a flowchart which shows other example of the voice input setting process which concerns on embodiment of this invention. It is a figure which shows the other structural example of the information processing system which concerns on embodiment of this invention.

Hereinafter, embodiments of the present invention will be described by taking application to an information processing system mounted on an automobile as an example.
FIG. 1 shows the configuration of the information processing system according to the present embodiment.
As shown in the figure, the information processing system includes a data processing unit 1, a dictionary DB 2, a microphone 3, a voice input unit 4, a speaker 5, an audio source 6 such as a radio receiver and a music player, a display device 7, and a plurality of cameras 8. It is equipped with other peripheral devices 9 such as a GPS receiver.

Here, the voice input unit 4 recognizes the voice of the user input from the microphone 3 and outputs the recognition result to the data output unit.
Further, the audio source 6 is a device that operates under the control of the data processing unit 1 and serves as a sound source for a radio receiver, a music player, or the like, and outputs the audio of the audio content to the speaker 5 and the audio input unit 4. Further, the speaker 5 radiates the sound input from the audio source 6 into the vehicle.

Further, the plurality of cameras 8 are a front camera for photographing the front of the automobile, a back camera for photographing the rear of the automobile, a side camera for photographing the side of the automobile, and the like.
Then, the data processing unit 1 can perform various processes while using the voice input unit 4 for voice input such as a command and the display device 7 for displaying the screen.
Further, the voice input unit 4 includes a first voice recognition engine 41, a first voice recognition dictionary 42, a second voice recognition engine 43, a second voice recognition dictionary 44, and a recognition adjustment unit 45.
Here, such an information processing system may be configured by using a computer provided with a CPU, a memory, a peripheral device, or the like. In this case, the data processing unit 1 and the voice input unit 4 described above may be used. It may be realized by the CPU executing a computer program.

Next, as shown in FIG. 2, the dictionary DB 2 stores recognition data of a plurality of layers from the first layer recognition data to the third layer recognition data.
The recognition data of each layer is written by a voice recognition dictionary used for voice recognition, and a number (No.) and a word are registered for each of a plurality of words.
Further, a plurality of second layer recognition data can be provided corresponding to each word registered in the first layer recognition data, and the third layer recognition data is each registered in each second layer recognition data. Multiple pieces can be provided corresponding to the word. That is, the dictionary DB2 has a tree structure in which the recognition data of each layer is used as a node.

Then, in the second layer recognition data, when the word of the first layer recognition data corresponding to the second layer recognition data is recognized as the recognition result by the voice recognition in the voice input unit 4, then the voice input unit A plurality of words and their numbers (No.) which are recognition candidates for voice recognition in 4 are registered.

Further, in the third layer recognition data, when the word of the second layer recognition data corresponding to the third layer recognition data is recognized as the recognition result by the voice recognition in the voice input unit 4, then the voice input unit A plurality of words and their numbers (No.) which are recognition candidates for voice recognition in 4 are registered.

The dictionary DB2 has been described above.
In the above, the case where the recognition data of three layers from the first layer recognition data to the third layer recognition data is provided as the recognition data to be registered in the dictionary DB2 has been described, but the layer recognition data to be registered in the dictionary DB2 is. It may be recognition data of any number of layers of 2 or more.

By the way, the data processing unit 1 is provided with various functions such as a car navigation function and a music player function, and when activated, a predetermined information processing (for example, a guide map for car navigation is displayed on the display device 7 by the car navigation function). Information processing to display the information processing and information processing to output music from the audio source 6 by the music player function) are started.

Further, when the data processing unit 1 is activated, the voice input control process shown in FIG. 3 is started.
As shown in the figure, in the voice input control process, after starting the voice input setting process described later (step 302), the standby process is started (step 304).
Here, in the standby process, the first layer recognition data is set as the current recognition data, the word of the current recognition data is input from the voice input unit 4 as the recognition result, and the input of the recognition result word is accepted. ..

Further, the voice input unit 4 recognizes and recognizes a word spoken by the user from among the words registered in the first layer recognition data and the words registered in the recognition data set in the current recognition data. The word is output to the data processing unit 1 as a recognition result. Here, the details of the operation for realizing such recognition of the voice input unit 4 will be described later.

Then, if the data processing unit 1 accepts the input of the word of the first layer recognition data set in the current recognition data input as the recognition result by the standby process (step 306), the recognition that the input is accepted Processing is performed according to the resulting word (step 308). Further, it is checked whether or not the second layer recognition data corresponding to the word of the first layer recognition data of the recognition result exists (step 310), and if it exists, if the standby process is being executed (step 312). ), After the standby process is completed (step 320), the sequence execution process in which the word of the first layer recognition data of the recognition result is set as the first layer word is started (step 314).

On the other hand, when the second layer recognition data corresponding to the word of the first layer recognition data of the recognition result does not exist (step 310), the process returns to step 306 while the standby process is continued, and the next step is performed by the standby process. Wait for the reception of the input of the recognition result of.

Here, in the sequence execution process started in step 314, the second layer recognition data corresponding to the first layer word of the first layer recognition data is set as the current recognition data, and then the word recognition result of the current recognition data is set. For each reception of the recognition result of each time, the processing according to the word of the recognition result that received the input and the recognition of the hierarchy one level below the current recognition data corresponding to the word of the recognition result are received. The sequence to be executed while setting the data to the current recognition data, that is, updating the current recognition data is executed.

Then, when the sequence execution process is started (step 314), the input from the voice input unit 4 as the word recognition result of the first layer recognition data occurs (step 316), and the sequence execution process ends (step 316). Monitor step 318).

Then, when the input as the recognition result of the word of the first layer recognition data occurs during the execution of the sequence execution process (step 316), the sequence execution process is terminated (step 322), and the process returns to step 308. In the standby process in which the first layer recognition data is the current recognition data, the same operation as when the input of the word of the first layer recognition data is accepted is performed as the input of the word of the recognition result.

That is, while performing processing according to the word of the first layer recognition data of the recognition result that received the input during the execution of the sequence execution process (step 308), the recognition result that received the input during the execution of the sequence execution process is the first. If the second layer recognition data corresponding to the word of the first layer recognition data exists (step 310), the word of the first layer recognition data of the recognition result that received the input during the execution of the sequence execution process is referred to as the first layer word. The sequence execution process to be performed is started (step 314).

On the other hand, when the end of the sequence execution process occurs (step 318), the process returns to step 304 and the standby process is restarted.
The voice input control process that the data processing unit 1 starts at startup has been described above.
Next, the data processing unit 1 performs the voice input setting process started in step 302 of the voice input control process as follows.
FIG. 4 shows the procedure of this voice input setting process.
As shown in the figure, in the voice input setting process, the data processing unit 1 monitors the occurrence of the setting of the current recognition data by the standby process and the sequence execution process described above (step 402).
Then, when the setting of the current recognition data occurs (step 402), it is determined whether or not the recognition mode of the current recognition data is the first layer recognition data (step 404).
Then, if the current recognition data is the first layer recognition data (step 404), the first layer recognition data which is the current recognition data is set in the first speech recognition dictionary 42 and the second speech recognition dictionary 44 (step 406). ), The audio cancel mode is set in the recognition adjustment unit 45 as the recognition mode (step 408).

Then, the recognition adjustment unit 45 is instructed to start voice recognition (step 410), and the process returns to the monitoring of step 402.
On the other hand, if the current recognition data is not one-layer recognition data (step 404), the current recognition data is set in the first speech recognition dictionary 42 (step 412), and the parallel mode is set in the recognition adjustment unit 45 as the recognition mode (step 412). Step 414). In step 414, the second voice recognition dictionary 44 is not updated, and as a result, the second voice recognition dictionary 44 is maintained as the first layer recognition data.

Then, the recognition adjustment unit 45 is instructed to start voice recognition (step 410), and the process returns to the monitoring of step 402.
The voice input setting process performed by the data processing unit 1 has been described above.
Next, the operation of voice recognition performed by the first voice recognition engine 41 and the second voice recognition engine 43 will be described.
The first voice recognition engine 41 and the second voice recognition engine 43 calculate the score of each recognition candidate word stored in the voice recognition dictionary for the recognition target voice in parallel with the input of the recognition target voice.

That is, the first speech recognition engine 41 calculates the score of the word stored in the first speech recognition dictionary 42 for the recognition target speech in parallel with the input of the recognition target speech, and the second speech recognition engine 43 recognizes it. In parallel with the input of the target voice, the score of each word stored in the second voice recognition dictionary 44 for the recognition target voice is calculated.

The recognition target voice of the first voice recognition engine 41 is a voice input from the microphone 3. On the other hand, with respect to the second voice recognition engine 43, one of the voice input from the microphone 3 and the voice input from the audio source 6 can be selectively used as the recognition target voice of the second voice recognition engine 43.

Here, the score of each word registered in the speech recognition dictionary for the speech to be recognized represents a predicted value of the magnitude of the difference between the word and the phrase represented by the speech to be recognized and the word, and predicts a larger difference. The more you do, the higher your score.

More specifically, in the calculation of the score, after setting a predetermined initial value as the score, each time the sound of each voice section of the recognition target voice (for example, the voice section for each phonetic element) is input, Whether or not the sound of the voice section matches the pronunciation of the part corresponding to the voice section of each word registered in the voice recognition dictionary is calculated, and if they match, the score is reduced by a predetermined value and matched. If not, the score is increased by a predetermined value. The increase / decrease value of the score for each voice section of the voice to be recognized is, for example, the magnitude obtained by multiplying the ratio of the words in the voice section to the total voice section by the initial value of the score.

According to such voice recognition, when the recognition target voice is "aiueoka", the transition of the score calculated for the word "aiueo" is shown in FIG. 5a, and it is calculated for the word "aiueoka". As shown in FIG. 5b, the score with the word gradually decreases and the sound of the recognition target voice that does not match the word is input while the sound of the recognition target voice that matches the word is input. The ward score will gradually increase while it is being played.

That is, for example, as shown in FIG. 5a, the recognition target voice "aiueoka", the word "aiueo", and the score gradually decrease during the period in which the recognition target voice "aiueo" sound is input, and then gradually decrease. , Increases when the "ka" of the recognition target voice is input.

Similarly, as shown in FIG. 5b, the recognition target voice "aiueoka", the word "aiai", and the score gradually decrease during the period in which the recognition target voice "ai" sound is input. After that, the period during which the recognition target voice "Eoka" is input gradually increases.

By the way, in the first voice recognition engine 41 and the second voice recognition engine 43, the score of the recognition target voice calculated as described above and one of the words is equal to or less than the preset threshold value Th. If so, the hit of the word whose score is equal to or less than the threshold value Th is detected, and the number (No.) of the hit word is notified to the recognition adjustment unit 45 as hit data.

That is, for example, in the case shown in FIG. 5a, the score for the word "aiueo" is equal to or less than the threshold value Th immediately before the "e" of the recognition target voice "aiueoka" is input. , The hit of the word "aiueo" is detected.

On the other hand, in the case shown in FIG. 5b, since the score for the word "Ai Aio" does not fall below the threshold value Th, the hit of this word "Ai Aio" is not detected.
Next, the operation of the recognition adjustment unit 45 of the voice input unit 4 will be described.
If the data processing unit 1 instructs the start of voice recognition by the voice input setting process shown in FIG. 4, the recognition adjustment unit 45 executes the audio cancel mode recognition process when the audio cancel mode is set. When the parallel recognition mode is set, the parallel recognition mode recognition process is executed.

First, the audio cancel mode recognition process performed by the recognition adjustment unit 45 when the audio cancel mode is set will be described.
FIG. 6 shows the procedure of the audio cancel mode recognition process shown.
As shown in the figure, in this audio cancel mode recognition process, a predetermined value Th1 is set in the first voice recognition engine 41 as the above-mentioned threshold value Th, and a predetermined value Th2 is set in the second voice recognition engine 43 as described above. The threshold value Th is set (step 602). Here, as Th1 and Th2, values such that Th2> Th1 are used.

Then, next, the recognition target voice of the second voice recognition engine 43 is set to the voice input from the audio source 6 (step 604).
Then, the hit data notification is generated from the first voice recognition engine 41 (step 606), the hit data notification is generated from the second voice recognition engine 43 (step 608), and the timer timeout occurs (step 610). ) And monitor.

Then, if the notification of the hit data from the second voice recognition engine 43 occurs (step 608), the mask flag is set (step 616), and the number indicated by the hit data notified from the second voice recognition engine 43 (step 616). No.) is set as the adjustment word number (step 618). Then, the timer described above is started by setting a predetermined timeout time (step 620), and the process returns to the monitoring of steps 606, 608, and 610.

On the other hand, when the timer time-out occurs (step 610), the mask flag is cleared (step 612) and the adjustment word number is cleared (step 614). Then, the process returns to the monitoring of steps 606, 608, and 610.

Further, when the notification of the hit data from the first voice recognition engine 41 occurs (step 606), it is checked whether or not the mask flag is set (step 622), and if the mask flag is not set, the first step. 1 The word of the number (No.) indicated by the hit data from the first voice recognition engine 41 of the voice recognition dictionary 42 is calculated as the word to be the recognition result, and the calculated recognition result is output to the data processing unit 1 (step). 626). Then, the audio cancel mode recognition process is terminated.

On the other hand, in step 622, when the mask flag is set, it is checked whether the number (No.) indicated by the hit data from the first speech recognition engine 41 matches the adjusted word number (step). 624) If they match, the process returns to the monitoring of steps 606, 608, and 610.

On the other hand, when the number (No.) indicated by the hit data from the first voice recognition engine 41 does not match the adjustment word number (step 624), the first voice recognition engine 41 of the first voice recognition dictionary 42 The word of the number (No.) indicated by the hit data of is calculated as a word to be a recognition result, and the calculated recognition result is output to the data processing unit 1 (step 626). Then, the audio cancel mode recognition process is terminated.

The audio cancel mode recognition process performed by the recognition adjustment unit 45 when the audio cancel mode is set has been described above.
Here, according to such an audio cancel mode recognition process, the word for which the second voice recognition engine 43 detects a hit with respect to the voice input from the audio source 6 is subsequently the first voice recognition engine 41 for a certain period of time. Even if a hit is detected in, it is not recognized as a recognition result.

Further, when the voice output by the audio source 6 wraps around the microphone 3 and a word hit is detected by the first voice recognition engine 41 for the voice, the threshold Th of the word is the first voice recognition. Since the second voice recognition engine 43 is set larger than the engine 41 and the voice of the audio source 6 input to the second voice recognition engine 43 has better voice quality, the second voice is set before that. The recognition engine 43 detects the hit of the word.

Therefore, even if the voice output by the audio source 6 wraps around the microphone 3 and a word hit is detected by the first voice recognition engine 41 for the voice, the word is recognized as a recognition result by the data processing unit. It is suppressed that it is output to 1.

Next, the parallel recognition mode recognition process performed by the recognition adjustment unit 45 when the parallel recognition mode is set will be described.
FIG. 7 shows the procedure of this parallel recognition mode recognition process.
As shown in the figure, in this parallel recognition mode recognition process, the above-mentioned value Th1 is set as the threshold value Th in both the first voice recognition engine 41 and the second voice recognition engine 43 (step 702).

Further, the recognition target voice of the second voice recognition engine 43 is set to the voice input from the microphone 3 (step 704).
Then, the generation of the hit data notification from the first voice recognition engine 41 (step 706) and the generation of the hit data notification from the second voice recognition engine 43 (step 708) are monitored.

Then, when the notification of the hit data from the first voice recognition engine 41 occurs (step 706), the number (No.) indicated by the hit data from the first voice recognition engine 41 of the first voice recognition dictionary 42. The word is calculated as a word to be the recognition result, and the calculated recognition result is output to the data processing unit 1 (step 710). Then, the parallel recognition mode recognition process is terminated.

On the other hand, when the notification of the hit data from the second voice recognition engine 43 occurs (step 708), the number (No.) indicated by the hit data from the second voice recognition engine 43 of the second voice recognition dictionary 44. The word is calculated as a word to be the recognition result, and the calculated recognition result is output to the data processing unit 1 (step 712). Then, the parallel recognition mode recognition process is terminated.

The parallel recognition mode recognition process performed by the recognition adjustment unit 45 when the parallel recognition mode is set has been described above.
According to such a parallel recognition mode recognition process, it becomes possible to perform voice recognition for both the words registered in the first voice recognition dictionary 42 and the words registered in the second voice recognition dictionary 44.

Here, an operation example of the above information processing apparatus is shown in FIG.
The activated data processing unit 1 starts information processing for displaying a guide map for car navigation on the display device 7 and information processing for outputting music from the audio source 6.
Further, when the data processing unit 1 is activated, the voice input setting process is started, and the first layer recognition data is set as the current recognition data, so that the first layer recognition data is as shown in FIGS. 8a2 and 8a3. Is set in both the first voice recognition dictionary 42 and the second voice recognition dictionary 44, and the audio cancel mode is set in the recognition adjustment unit 45.

As a result, in the recognition adjustment unit 45 of the voice input unit 4, the voice recognition of the word registered in the first voice recognition dictionary 42 by the audio cancel mode recognition process causes erroneous recognition by the output voice of the audio source 6. 2 It is performed while suppressing using the voice recognition dictionary 44.

Next, in this state, when the user utters the word "Mokukichisete" registered in the first speech recognition dictionary 42, the recognition adjustment unit 45 uses the word "Mokukichisete" as the recognition result. The calculation is performed, and the word "Mokukichisetei" is output from the voice input unit 4 to the data processing unit 1 as a recognition result.

When the data processing unit 1 receives the word "Mokukichiset" as the recognition result, in response to this, the data processing unit 1 responds to the second layer recognition corresponding to the word "Mokukichiset" of the first layer recognition data. Set the data to the current recognition data. Then, as shown in FIG. 8b1, the display screen of the display device 7 is changed to a screen including a list of words registered in the second layer recognition data set in the current recognition data.

Further, the data processing unit 1 sets the second layer recognition data set in the current recognition data in the first voice recognition dictionary 42 as shown in FIG. 8b2. On the other hand, as shown in FIG. 8b3, the second speech recognition dictionary 44 is maintained as the first layer recognition data. Then, the data processing unit 1 sets the parallel recognition mode in the recognition adjustment unit 45.

As a result, the recognition adjustment unit 45 of the voice input unit 4 registers the word of the second layer recognition data registered in the first voice recognition dictionary 42 and the second voice recognition dictionary 44 by the parallel recognition mode recognition process. Speech recognition of both words of the first layer recognition data is performed.

Next, in this state, when the user utters the word "Chikaku no Ramenya" registered in the first speech recognition dictionary 42, the recognition adjustment unit 45 calculates the word "Chikaku no Ramenya" as the recognition result. , The word "Chikaku no Ramenya" is output from the voice input unit 4 to the data processing unit 1 as a recognition result.

When the data processing unit 1 accepts the word "Chikaku no Ramenya" as the recognition result, in response to this, the word "Chikaku no Ramenya" of the second layer recognition data set in the current recognition data is used. The corresponding third layer recognition data is set as the current recognition data. Then, a ramen shop near the current position is searched by the car navigation function provided in the data processing unit 1, and as shown in FIG. 8c1, the display screen of the display device 7 is set as the current recognition data in the third layer. The list of words registered in the recognition data is changed to 5 words from "I go to the first" to "I go to the tsutsume" registered in the third layer recognition data set in the current recognition data in the list. Change the display of the five ramen shops near the searched current position to a screen that includes the corresponding form.

Further, the data processing unit 1 sets the third layer recognition data set in the current recognition data in the first voice recognition dictionary 42 as shown in FIG. 8c2. On the other hand, as shown in FIG. 8c3, the second speech recognition dictionary 44 is maintained as the first layer recognition data. Then, the data processing unit 1 sets the parallel recognition mode in the recognition adjustment unit 45.

As a result, the recognition adjustment unit 45 of the voice input unit 4 registers the word of the third layer recognition data registered in the first voice recognition dictionary 42 and the second voice recognition dictionary 44 by the parallel recognition mode recognition process. Speech recognition of both words of the first layer recognition data is performed.

Then, in this state, when the user utters the word "first go" registered in the first speech recognition dictionary 42, the recognition adjustment unit 45 calculates the word "first go" as a recognition result. The word "first go" is output from the voice input unit 4 to the data processing unit 1 as a recognition result. Then, when the data processing unit 1 accepts the word "go to the first" as the recognition result, the data processing unit 1 sets the ramen shop shown on the screen of FIG. 8c1 as the destination in association with the word "go to the first". , The car navigation function provided in the data processing unit 1 starts processing for route guidance to the destination.

On the other hand, in a state where the second layer recognition data of FIGS. 8b1, b2, b3 is set in the first speech recognition dictionary 42, the word "word" of the first layer recognition data registered in the second speech recognition dictionary 44 by the user. When "Bakukamera" is spoken, the recognition adjustment unit 45 calculates the word "Bakukamera" as a recognition result, and the word "Bakukamera" is recognized as a recognition result from the voice input unit 4 to the data processing unit 1. It is output.

Here, the data processing unit 1 according to the present embodiment is a camera 8 corresponding to when words such as "Furonto Kamera", "Saido Kamera", and "Bakku Kamera" are input as recognition results. It is assumed that the process of displaying the captured image on the display device 7 is performed.

In this case, if the data processing unit 1 accepts the word "bakkamera" as the recognition result, the data processing unit 1 currently recognizes the second layer recognition data corresponding to the word "bakkamera" of the first layer recognition data. Set in the data.

Further, as shown in FIG. 9d1, the data processing unit 1 registers an image of the rear of the vehicle taken by a back camera provided as one of the cameras 8 in the second layer recognition data set as the current recognition data. Display with a list of words. Note that FIGS. 9a1, a2, and a3 are the same as those in FIGS. 8a1, a2, and a3, and FIGS. 9b1, b2, and b3 are the same as those in FIGS. 8b1, b2, and b3.

Further, the data processing unit 1 sets the second layer recognition data set in the current recognition data in the first speech recognition dictionary 42 as shown in FIG. 9d2. On the other hand, as shown in FIG. 9d3, the second speech recognition dictionary 44 is maintained as the first layer recognition data. Then, the data processing unit 1 sets the parallel recognition mode in the recognition adjustment unit 45.

As a result, the recognition adjustment unit 45 of the voice input unit 4 registers the word of the second layer recognition data registered in the first voice recognition dictionary 42 and the second voice recognition dictionary 44 by the parallel recognition mode recognition process. Speech recognition of both words of the first layer recognition data is performed.

By the way, when the parallel recognition mode is set as described above, the false recognition by the output voice of the audio source 6 is not suppressed. However, the parallel recognition mode is set when the layer recognition data of the layer other than the first layer recognition data is set as the active recognition data, and the layer recognition data of the layer other than the first layer recognition data is set. When it is set to the current recognition data, it is when the data processing unit 1 is performing the sequence execution process described above and the user is continuously performing a series of hierarchical voice inputs.

Therefore, the period from the setting of the parallel recognition mode to the voice input by the user's utterance is short, and during this period, it is the same as the word set in the first voice recognition dictionary 42 and the second voice recognition dictionary 44. Word audio is rarely output from audio source 6.

Therefore, when the layer recognition data of the layer other than the first layer recognition data is set as the working recognition data, there may be a practical problem even if the false recognition by the output voice of the audio source 6 is not suppressed. do not have.

When the first layer recognition data is set as the current recognition data, the data processing unit 1 is in the state of performing the standby process described above, and after setting the first layer recognition data as the current recognition data. The period until voice input by the user's utterance occurs is indefinite. Therefore, during this period, there is a high possibility that the voice of the same word as the word set in the first voice recognition dictionary 42 is output from the audio source 6, so that the second voice of erroneous recognition by the output voice of the audio source 6 is not small. It is necessary to perform deterrence using the recognition dictionary 44.

On the other hand, while maintaining the second voice recognition dictionary 44 in the first layer recognition data as described above, the first voice recognition dictionary 42 is updated and the parallel recognition mode is set, so that the words recognized by the voice can be obtained. While updating the word to be recognized next time, the word of the first layer recognition data can be constantly recognized by voice.

In addition, a word representing a command requesting execution of an urgent process for the user is often registered in the first layer recognition data.
For example, the data processing unit 1 according to the present embodiment corresponds to when words such as "front camera", "saido camera", and "back camera" are input as recognition results as described above. The process of displaying the image taken by the camera 8 on the display device 7 is performed, and these "Furonto Kamera", "Saido Kamera", and "Bat" registered in the first layer recognition data are performed. A word such as "Kukamera" is a word of a command instructing the user to display a captured image of the camera 8 for confirming the surrounding situation, and therefore corresponds to a word representing a command requesting execution of an urgent process.

Therefore, according to the present embodiment, it becomes possible to accept the voice input of the command requesting the execution of the processing that is urgent for the user at an arbitrary time point.
The embodiment of the present invention has been described above.
By the way, in the above embodiment, the voice input setting process shown in FIG. 10 may be performed instead of the voice input setting process shown in FIG.
That is, in this voice input setting process, the data processing unit 1 monitors the occurrence of the setting of the current recognition data (step 1002).
Then, when the setting of the current recognition data occurs (step 1002), it is checked whether or not the recognition mode of the current recognition data is the first layer recognition data (step 1004).
Then, if the current recognition data is the first layer recognition data (step 1004), the first layer recognition data which is the current recognition data is set in the first voice recognition dictionary 42 and the second voice recognition dictionary 44 (step 1006). ), The audio cancel mode is set in the recognition adjustment unit 45 as the recognition mode (step 1008).

Then, the recognition adjustment unit 45 is instructed to start voice recognition (step 1010), and the process returns to the monitoring of step 1002.
On the other hand, if the current recognition data is not the one-layer recognition data, it is examined whether or not the data processing unit 1 is currently outputting audio from the audio source 6 to the speaker 5 (step 1012).

Then, if the voice is output from the audio source 6 to the speaker 5 (step 1012), the current recognition data is set in the first voice recognition dictionary 42 and the second voice recognition dictionary 44 (step 1006), and the recognition mode is set. The audio cancel mode is set in the recognition adjustment unit 45 (step 1008).

Then, the recognition adjustment unit 45 is instructed to start voice recognition (step 1010), and the process returns to the monitoring of step 1002.
If no voice is output from the audio source 6 to the speaker 5 (step 1012), the current recognition data is set in the first voice recognition dictionary 42, and the first layer recognition data is set in the second voice recognition dictionary 44. , (Step 1014), the parallel recognition mode is set in the recognition adjustment unit 45 as the recognition mode (step 1016).

Then, the recognition adjustment unit 45 is instructed to start voice recognition (step 1010), and the process returns to the monitoring of step 1002.
According to such an audio input setting process, when the audio is output from the audio source 6 to the speaker 5, the audio is output from the audio source 6 to the speaker 5 while erroneously recognizing the output audio of the audio source 6. When there is no such thing, that is, when there is no erroneous recognition due to the output voice of the audio source 6, the number of words that can be voice-recognized can be increased.

Further, in the above embodiment, the first layer recognition data is maintained as the second speech recognition dictionary 44 in the parallel recognition mode, but in the parallel recognition mode, the same as the first speech recognition dictionary 42. The contents of the second voice recognition dictionary 44 may be switched.

That is, for example, as shown in FIG. 11, the recognition data of each layer other than the first layer recognition data includes the main recognition data and the sub-recognition data in which different sets of words are registered, and FIG. In step 412 of the voice input setting process shown in FIG. 10 and step 1014 of the voice input setting process shown in FIG. 10, the main recognition data of the current recognition data is set in the first voice recognition dictionary 42, and the sub-recognition data of the current recognition data is set. May be set in the second speech recognition dictionary 44.

By doing so, it becomes possible to more flexibly set the words that can be voice-recognized by the parallel recognition mode recognition process.

1 ... Data processing unit, 2 ... Dictionary DB, 3 ... Microphone, 4 ... Voice input unit, 5 ... Speaker, 6 ... Audio source, 7 ... Display device, 8 ... Camera, 9 ... Peripheral device, 41 ... First voice recognition Engine, 42 ... 1st voice recognition dictionary, 43 ... 2nd voice recognition engine, 44 ... 2nd voice recognition dictionary, 45 ... recognition adjustment unit.

Claims (14)

It is a voice recognition system that recognizes the voice uttered in the space where the voice output from the audio source device is radiated from the speaker to the speaker.
With a microphone
The first speech recognition dictionary in which multiple words are registered, and
A first voice recognition unit that inputs a voice picked up by the microphone and detects a word matching the input voice as a target candidate from a plurality of words registered in the first voice recognition dictionary.
The second voice recognition unit and
The second voice recognition dictionary used by the second voice recognition unit,
A recognition unit that recognizes words spoken by the user,
It has a voice input receiving unit that receives voice input of a word recognized by the recognition unit while selectively setting a first recognition mode and a second recognition mode.
In the second voice recognition dictionary used by the second voice recognition unit in the first recognition mode, the same plurality of words as the plurality of words registered in the first voice recognition dictionary are registered.
In the first recognition mode, the second voice recognition unit inputs the voice output to the speaker by the audio source device, and the input voice is selected from a plurality of words registered in the second voice recognition dictionary. Detects words that match the target as target candidates,
In the first recognition mode, the recognition unit recognizes the same words as the words detected as the target candidates during the predetermined period after the target candidate is detected by the second voice recognition unit. Even if the target is detected by the unit, the target is prevented from being recognized as a word spoken by the user by the first voice recognition unit. Recognize the word detected as a candidate as the word spoken by the user and
In the second voice recognition dictionary used by the second voice recognition unit in the second recognition mode, a plurality of words registered in the first voice recognition dictionary and a plurality of words that are at least partially different from the plurality of words are registered. And
In the second recognition mode, the second voice recognition unit inputs the voice picked up by the microphone, and among a plurality of words registered in the second voice recognition dictionary, a word matching the input voice. Is detected as a target candidate,
In the second recognition mode, when the target candidate is detected by the first voice recognition unit, the recognition unit recognizes the word detected as the target candidate as the word spoken by the user, and the second recognition unit. (2) A voice recognition system characterized in that when the target candidate is detected by the voice recognition unit, the word detected as the target candidate is recognized as the word spoken by the user. The voice recognition system according to claim 1.
The voice input reception unit is
The first recognition mode is set as the recognition mode, the standby voice recognition dictionary, which is a predetermined voice recognition dictionary, is set in the first voice recognition dictionary and the second voice recognition dictionary, and the recognition unit recognizes them. Performs standby processing to accept voice input of words,
When the voice input of the word is accepted by the standby process, the sequence execution process in which the word for which the voice input is accepted is set as the first layer word is started.
In the sequence execution process, the second recognition mode is set as the recognition mode, and at each time point, one of the first voice recognition dictionary and the second voice recognition dictionary sets the standby voice recognition dictionary. While using the first speech recognition dictionary and the second speech recognition dictionary so that the sub-speech recognition dictionary and the other become the main speech recognition dictionary, the main speech recognition dictionary determined according to the first layer word is set. While updating the voice recognition dictionary to a voice recognition dictionary determined according to the words registered in the main voice recognition dictionary that received voice input, the voice of the words registered in the main voice recognition dictionary recognized by the recognition unit. It is a process to execute a sequence that accepts input once or multiple times.
When the sequence execution process is completed, the standby process is started, and the standby process is started.
If the voice input of the word registered in the standby voice recognition dictionary is accepted during the execution of the sequence execution process, the sequence execution process in which the word receiving the voice input is set as the first layer word is started. A voice recognition system, characterized in that it performs at least one of the above and the start of other processing associated with the word that has received the voice input. The voice recognition system according to claim 1.
The voice input reception unit is
The first recognition mode is set as the recognition mode, the standby voice recognition dictionary, which is a predetermined voice recognition dictionary, is set in the first voice recognition dictionary and the second voice recognition dictionary, and the recognition unit recognizes them. Performs standby processing to accept voice input of words,
When the voice input of the word is accepted by the standby process, the sequence execution process in which the word for which the voice input is accepted is set as the first layer word is started.
In the sequence execution process, the first voice recognition dictionary and the second voice recognition are set by setting the second recognition mode as the recognition mode and setting different voice recognition dictionaries determined according to the first layer word. It is a process of executing a sequence of accepting the voice input of the word recognized by the recognition unit once or a plurality of times while updating the dictionary to a different voice recognition dictionary determined according to the word received as the voice input.
A voice recognition system characterized in that the standby process is started when the sequence execution process is completed. The voice recognition system according to claim 1.
The voice input reception unit is
The first recognition mode is set as the recognition mode, the standby voice recognition dictionary, which is a predetermined voice recognition dictionary, is set in the first voice recognition dictionary and the second voice recognition dictionary, and the recognition unit recognizes them. Performs standby processing to accept voice input of words,
If the voice input of a word is accepted when the voice output from the audio source device is radiated from the speaker by the standby process, the word that has received the voice input is set as the first layer word. The first sequence execution process is started,
If the voice input of a word is accepted when the voice output from the audio source device is not radiated from the speaker by the standby process, the word that has received the voice input is set as the first layer word. The second sequence execution process is started,
In the first sequence execution process, the first voice recognition mode and the second voice recognition dictionary in which the first recognition mode is set as the recognition mode and the same voice recognition dictionary determined according to the first layer word is set are set. Is a process of executing a sequence of accepting the voice input of the word recognized by the recognition unit once or a plurality of times while updating the same voice recognition dictionary determined according to the word that received the voice input.
In the second sequence execution process, the second recognition mode is set as the recognition mode, and at each time point, one of the first voice recognition dictionary and the second voice recognition dictionary uses the standby voice recognition dictionary. While using the first speech recognition dictionary and the second speech recognition dictionary so that the set sub-speech recognition dictionary and the other become the main speech recognition dictionary, a speech recognition dictionary determined according to the first layer word was set. While updating the main voice recognition dictionary to a voice recognition dictionary determined according to the words registered in the main voice recognition dictionary that received voice input, the words registered in the main voice recognition dictionary recognized by the recognition unit. It is a process to execute a sequence that accepts the voice input of
When the first sequence execution process is completed, the standby process is started, and the standby process is started.
When the second sequence execution process is completed, the standby process is started, and the standby process is started.
If the voice input of the word registered in the standby voice recognition dictionary is received during the execution of the second sequence execution process, the word receiving the voice input is set as the first layer word. A voice recognition system, characterized in that it starts at least one of the start of processing and the start of other processing associated with the word that received the voice input. The voice recognition system according to claim 1.
The voice input reception unit is
The first recognition mode is set as the recognition mode, the standby voice recognition dictionary, which is a predetermined voice recognition dictionary, is set in the first voice recognition dictionary and the second voice recognition dictionary, and the recognition unit recognizes them. Performs standby processing to accept voice input of words,
If the voice input of a word is accepted when the voice output from the audio source device is radiated from the speaker by the standby process, the word that has received the voice input is set as the first layer word. The first sequence execution process is started,
If the voice input of a word is accepted when the voice output from the audio source device is not radiated from the speaker by the standby process, the word that has received the voice input is set as the first layer word. The second sequence execution process is started,
In the first sequence execution process, the first voice recognition mode and the second voice recognition dictionary in which the first recognition mode is set as the recognition mode and the same voice recognition dictionary determined according to the first layer word is set are set. Is a process of executing a sequence of accepting the voice input of the word recognized by the recognition unit once or a plurality of times while updating the same voice recognition dictionary determined according to the word that received the voice input.
In the second sequence execution process, the first voice recognition dictionary and the second voice recognition dictionary in which the second recognition mode is set as the recognition mode and different voice recognition dictionaries are set according to the first layer word are set. In the process of executing a sequence of accepting the voice input of the word recognized by the recognition unit once or multiple times while updating the voice recognition dictionary to a different voice recognition dictionary determined according to the word that received the voice input. can be,
When the first sequence execution process is completed, the standby process is started, and the standby process is started.
A voice recognition system characterized in that the standby process is started when the second sequence execution process is completed. The voice recognition system according to claim 1.
The voice input receiving unit sets the recognition mode to the first recognition mode when the voice output from the audio source device is emitted from the speaker, and is output from the audio source device from the speaker. A voice recognition system characterized in that the recognition mode is set to the second recognition mode when no sound is emitted. The voice recognition system according to claim 1, 2, 3, 4, 5 or 6.
The voice recognition system is a voice recognition system used for voice input in an in-vehicle system mounted on an automobile. A computer program that is read and executed by a computer equipped with a microphone placed in a space where the sound output from the audio source device is radiated from the speaker.
The computer program uses the computer,
The first speech recognition dictionary in which multiple words are registered, and
A first voice recognition unit that inputs a voice picked up by the microphone and detects a word matching the input voice as a target candidate from a plurality of words registered in the first voice recognition dictionary.
The second voice recognition unit and
The second voice recognition dictionary used by the second voice recognition unit,
A recognition unit that recognizes words spoken by the user,
A computer program that functions as a voice input receiving unit that receives voice input of a word recognized by the recognition unit while selectively setting a first recognition mode and a second recognition mode.
In the second voice recognition dictionary used by the second voice recognition unit in the first recognition mode, the same plurality of words as the plurality of words registered in the first voice recognition dictionary are registered.
In the first recognition mode, the second voice recognition unit inputs the voice output to the speaker by the audio source device, and the input voice is selected from a plurality of words registered in the second voice recognition dictionary. Detects words that match the target as target candidates,
In the first recognition mode, the recognition unit recognizes the same words as the words detected as the target candidates during the predetermined period after the target candidate is detected by the second voice recognition unit. Even if the target is detected by the unit, the target is prevented from being recognized as a word spoken by the user by the first voice recognition unit. Recognize the word detected as a candidate as the word spoken by the user and
In the second voice recognition dictionary used by the second voice recognition unit in the second recognition mode, a plurality of words registered in the first voice recognition dictionary and a plurality of words that are at least partially different from the plurality of words are registered. And
In the second recognition mode, the second voice recognition unit inputs the voice picked up by the microphone, and among a plurality of words registered in the second voice recognition dictionary, a word matching the input voice. Is detected as a target candidate,
In the second recognition mode, when the target candidate is detected by the first voice recognition unit, the recognition unit recognizes the word detected as the target candidate as the word spoken by the user, and the second recognition unit. (2) A computer program characterized in that when the target candidate is detected by the voice recognition unit, the word detected as the target candidate is recognized as the word spoken by the user. The computer program according to claim 8.
The voice input reception unit is
The first recognition mode is set as the recognition mode, the standby voice recognition dictionary, which is a predetermined voice recognition dictionary, is set in the first voice recognition dictionary and the second voice recognition dictionary, and the recognition unit recognizes them. Performs standby processing to accept voice input of words,
When the voice input of the word is accepted by the standby process, the sequence execution process in which the word for which the voice input is accepted is set as the first layer word is started.
In the sequence execution process, the second recognition mode is set as the recognition mode, and at each time point, one of the first voice recognition dictionary and the second voice recognition dictionary sets the standby voice recognition dictionary. While using the first speech recognition dictionary and the second speech recognition dictionary so that the sub-speech recognition dictionary and the other become the main speech recognition dictionary, the main speech recognition dictionary determined according to the first layer word is set. While updating the voice recognition dictionary to a voice recognition dictionary determined according to the words registered in the main voice recognition dictionary that received voice input, the voice of the words registered in the main voice recognition dictionary recognized by the recognition unit. It is a process to execute a sequence that accepts input once or multiple times.
When the sequence execution process is completed, the standby process is started, and the standby process is started.
If the voice input of the word registered in the standby voice recognition dictionary is accepted during the execution of the sequence execution process, the sequence execution process in which the word receiving the voice input is set as the first layer word is started. A computer program characterized by performing at least one of the operation and the start of other processing associated with the word that received the voice input. The computer program according to claim 8.
The voice input reception unit is
The first recognition mode is set as the recognition mode, the standby voice recognition dictionary, which is a predetermined voice recognition dictionary, is set in the first voice recognition dictionary and the second voice recognition dictionary, and the recognition unit recognizes them. Performs standby processing to accept voice input of words,
When the voice input of the word is accepted by the standby process, the sequence execution process in which the word for which the voice input is accepted is set as the first layer word is started.
In the sequence execution process, the first voice recognition dictionary and the second voice recognition are set by setting the second recognition mode as the recognition mode and setting different voice recognition dictionaries determined according to the first layer word. It is a process of executing a sequence of accepting the voice input of the word recognized by the recognition unit once or a plurality of times while updating the dictionary to a different voice recognition dictionary determined according to the word received as the voice input.
A computer program characterized in that the standby process is started when the sequence execution process is completed. The computer program according to claim 8.
The voice input reception unit is
The first recognition mode is set as the recognition mode, the standby voice recognition dictionary, which is a predetermined voice recognition dictionary, is set in the first voice recognition dictionary and the second voice recognition dictionary, and the recognition unit recognizes them. Performs standby processing to accept voice input of words,
If the voice input of a word is accepted when the voice output from the audio source device is radiated from the speaker by the standby process, the word that has received the voice input is set as the first layer word. The first sequence execution process is started,
If the voice input of a word is accepted when the voice output from the audio source device is not radiated from the speaker by the standby process, the word that has received the voice input is set as the first layer word. The second sequence execution process is started,
In the first sequence execution process, the first voice recognition mode and the second voice recognition dictionary in which the first recognition mode is set as the recognition mode and the same voice recognition dictionary determined according to the first layer word is set are set. Is a process of executing a sequence of accepting the voice input of the word recognized by the recognition unit once or a plurality of times while updating the same voice recognition dictionary determined according to the word that received the voice input.
In the second sequence execution process, the second recognition mode is set as the recognition mode, and at each time point, one of the first voice recognition dictionary and the second voice recognition dictionary uses the standby voice recognition dictionary. While using the first speech recognition dictionary and the second speech recognition dictionary so that the set sub-speech recognition dictionary and the other become the main speech recognition dictionary, a speech recognition dictionary determined according to the first layer word was set. While updating the main voice recognition dictionary to a voice recognition dictionary determined according to the words registered in the main voice recognition dictionary that received voice input, the words registered in the main voice recognition dictionary recognized by the recognition unit. It is a process to execute a sequence that accepts the voice input of
When the first sequence execution process is completed, the standby process is started, and the standby process is started.
When the second sequence execution process is completed, the standby process is started, and the standby process is started.
If the voice input of the word registered in the standby voice recognition dictionary is received during the execution of the second sequence execution process, the word receiving the voice input is set as the first layer word. A computer program characterized in that it starts at least one of the start of the other processing associated with the word that received the voice input. The computer program according to claim 8.
The voice input reception unit is
The first recognition mode is set as the recognition mode, the standby voice recognition dictionary, which is a predetermined voice recognition dictionary, is set in the first voice recognition dictionary and the second voice recognition dictionary, and the recognition unit recognizes them. Performs standby processing to accept voice input of words,
If the voice input of a word is accepted when the voice output from the audio source device is radiated from the speaker by the standby process, the word that has received the voice input is set as the first layer word. The first sequence execution process is started,
If the voice input of a word is accepted when the voice output from the audio source device is not radiated from the speaker by the standby process, the word that has received the voice input is set as the first layer word. The second sequence execution process is started,
In the first sequence execution process, the first voice recognition mode and the second voice recognition dictionary in which the first recognition mode is set as the recognition mode and the same voice recognition dictionary determined according to the first layer word is set are set. Is a process of executing a sequence of accepting the voice input of the word recognized by the recognition unit once or a plurality of times while updating the same voice recognition dictionary determined according to the word that received the voice input.
In the second sequence execution process, the first voice recognition dictionary and the second voice recognition dictionary in which the second recognition mode is set as the recognition mode and different voice recognition dictionaries are set according to the first layer word are set. In the process of executing a sequence of accepting the voice input of the word recognized by the recognition unit once or multiple times while updating the voice recognition dictionary to a different voice recognition dictionary determined according to the word that received the voice input. can be,
When the first sequence execution process is completed, the standby process is started, and the standby process is started.
A computer program characterized in that the standby process is started when the second sequence execution process is completed. The computer program according to claim 8.
The audio input receiving unit sets the recognition mode to the first recognition mode when the sound output from the audio source device is emitted from the speaker, and is output from the audio source device from the speaker. A computer program characterized in that the recognition mode is set to the second recognition mode when no sound is emitted. The computer program according to claim 8, 9, 10, 11, 12 or 13.
The computer is a computer program characterized in that it is a computer mounted on an automobile.

Images