JP2000061137A - Cap-verses game device and navigation system with cap- verses game function provided with the same - Google Patents

Abstract

PROBLEM TO BE SOLVED: To smoothly perform 'cap-verses' by preventing the reduction of a voice recognizing rate in the case of performing the 'cap-verses' while using voice recognizing techniques. SOLUTION: When the 'cap-verses' is started, a recognition control part 11c sets a dictionary for cap-verses to a dictionary managing part 11d as the dictionary of a retrieving object. When input voices from a user are converted to time series audio data by an input signal processing part 11a and the same pattern is repeated more than three times in the first section of time series audio data, a recognizing processing part 11b corrects the audio data so that this pattern can appear just once in the beginning of audio data. Further more, the recognized result based on similarity between the pattern of these audio data and the pattern of a recognizing object word stored in the dictionary for cap-verses set by the dictionary managing part 11d is outputted to the recognition control part 11c. The word having final one sound of a preceding word as an initial letter is outputted to a control part 20 as the recognized result by the recognition control part 11c.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a tail picking game device using voice recognition technology and a navigation system with a tail picking game function.

[0002]

2. Description of the Related Art Heretofore, a voice recognition device which has been proposed and realized is effective, for example, in the case where a destination setting in a car navigation system can be inputted by voice.

By the way, a configuration is considered in which a user and a device use the voice recognition technique as described above to carry out a word play, "skipping". Such a device, for example, recognizes a voice input by a user, selects an appropriate word with the last letter of the recognized word as an acronym from a dictionary, and outputs the synthesized voice by repeating the operation. It is possible to configure Here, in the conventional general speech recognition method,
The voice pattern input by the user is compared with the voice pattern of the recognition target word stored in advance, and the one having a high degree of similarity is selected.

[0004]

[Problems to be Solved by the Invention] However, in the word "buttocking", for example, when trying to recall a word with the acronym "go", "gogogogo, gorilla"
Often repeats an acronym. At this time, a word such as "gogogogora" is not registered as a recognition target word in the dictionary, so that the voice input by the user cannot be recognized. That is, the user is "Gorilla"
However, there is a disadvantage that the device is not recognized.

The present invention has been made in order to eliminate the above-mentioned inconvenience, and prevents a decrease in the voice recognition rate when performing "tail removal" using voice recognition technology, and performs "bottom removal" smoothly. The purpose is to

[0006]

In the tail picking game device of the present invention, when the user inputs a voice, the voice recognition means recognizes the input voice.
For example, this recognition method is performed by pattern matching with a recognition target word previously stored in a dictionary. For example, the voice recognition means sets a word having the highest degree of similarity among the recognition target words or a word having a degree of similarity equal to or higher than a predetermined value as a recognition result.

Further, the utterance word selecting means selects, from the pre-stored utterance target words, a word having an acronym as the last one sound of the words recognized by the voice recognizing means, in accordance with the tailing rule. For example, when the voice recognition means recognizes the input voice from the user as the word "tori (bird)", the utterance word selection means causes the last one sound of this word to be an acronym. Is selected from the utterance target words.
For example, if “apple” is the utterance target word, it is selected.

Then, the utterance means outputs the word thus selected as a voice. It is also possible that the utterance means outputs voice based on the voice dictionary,
It is also conceivable to speak a word for each syllable in synthetic speech. The voice dictionary stores voice signals in units of words, and the voice output based on this is output in a tone closer to that of a human.

Of course, the game ends when a word whose final sound is "n" is used, but it is also conceivable to set other conditions for ending the game. Such a condition for ending the game may be fixedly set at the time of shipment of the device, or the menu may be configured to be displayed on the display unit, and the user may follow the menu displayed on the display unit. It may be set. For example, as a condition for ending the game, it is possible to use a previously used word, use a word including a predetermined prohibited character, or the like. In addition, since the game may continue for a long time, it is possible to end the game as a draw if a predetermined number of answers is exceeded.

The number of users is not limited to one. For example, user A → user B → butt removal game device → user A
→ ... "Assault" between multiple users
It is also possible to do. In this case, when the voice recognizing means recognizes the input voice from the user B, the utterance word selecting means outputs the corresponding utterance word. It is conceivable that the number and order of users can be set, for example, according to the menu displayed on the display means as described above. If such settings are stored in, for example, a non-volatile memory device, it is not necessary to set each time, which is convenient.

In addition to the basic functions as described above, particularly in the tail picking game device of the present invention, when the above-described voice recognition means repeats the same sound at the beginning of the input voice from the user a predetermined number of times or more. Recognizes the input sound by considering the repeated sound as one or two sounds. This will be described.

[0012] In the case of "jiri-ri", for example, when trying to recall a word whose acronym is "go", the user often pronounces the acronym repeatedly, such as "gogogogo, gorilla". . However, since the word “gogogogora” is usually not registered in the dictionary, the word input by the user cannot be recognized. That is, there is an inconvenience that the user does not recognize the device even though the user intends to be "Gorilla".

Therefore, in the present invention, when the same sound is repeated a predetermined number of times or more at the beginning of the voice input by the user, for example, the sound of "go" is repeated, for example, "gogogogora". If so, "gogogogo" is regarded as one or two sounds and the word is recognized. For example, "gogogogora" entered by the user is compared with a dictionary as "gogora" or "gogora". As a result, it will be recognized as "Gorilla".

Here, when considering the case of two sounds, for example, "hahahahahahaha (mother)" and "mimimimi, mizumizuku (wooden roar)", "hahahahahaha" This is because it can be recognized as proper Japanese by treating it as two sounds, such as "→" haha "and" mimimimimimi "→" mimi ".

Further, how many times the "predetermined number of times" is set may be determined as follows. For example, "haha (mother)", "peach (peach)", "mimi (ear)"
Considering that there are relatively many words in which the same sound is repeated twice and relatively few words in which the same sound is repeated three times, it is conceivable that the above-mentioned predetermined number of times is three times. That is, when the same sound is repeated three times or more at the beginning of the voice, the sound is regarded as one or two sounds and the voice recognition is performed. In this way, when the same sound is repeated only twice, it is not processed, which is advantageous in that the processing load is reduced.

However, "go, gorilla", "ri, apple"
Considering that there is an answer pattern, it is possible to set the predetermined number of times to two. In this case, a word in which the same sound is repeated only twice, for example, a word such as “haha” or “peach” is also processed, and “haha” → “haha”, “peach” → “ Although processing such as "peach" is performed and the processing load increases, "Gorilla" is replaced with "Gorilla"
It is possible to recognize "Ringo apple" as "Apple".

With the voice recognition means as described above, according to the tail removal game device of the present invention, it is possible to prevent the voice recognition rate from being lowered when performing the "butt removal" and to smoothly perform the "butt removal". Considering to further increase the voice recognition rate, as described in claim 3, the voice recognition means, when recognizing an input voice from the user, selects or recognizes immediately before by the speech word selection means or the voice recognition means. It is conceivable to configure so that the last one sound of the written word is recognized as an acronym. That is, the voice input by the user following the word output by the utterance means is a word with the last one sound of the word output by the utterance means as an acronym, based on the rule of “sucking”. In addition, user A → user B → butt removal game device → user A → ...
In the case of performing a “bottom off” with a plurality of users as described above, the voice input from the user B is the last one sound of the word input from the user A and recognized by the voice recognition means. Is a word to do. Therefore, when recognizing the voice input by the user, the voice recognition means recognizes the voice as the word having the last one sound of the immediately preceding word as an acronym. For example, if you cannot clearly recognize whether the word entered by the user is "Wakayama (Wakayama)" or "Okayama (Okayama)", but the last word is "Kanagawa (Kanagawa)", "Wakayama (Wakayama)" It is just like recognizing.

This makes it possible to further improve the voice recognition rate at the time of performing the "bottom removal", and the user can perform the "bottom removal" more smoothly. If voice recognition is not possible or insufficient due to noise or the like, the result of voice recognition may be output and the user may be asked to confirm. For example, voice output such as "Wakayama (Wakayama)" or "Is Wakayama (Wakayama) okay?", Waits for a reply from the user for a predetermined time, and if there is a reply that the recognition is wrong , Prompt for re-input of voice.

By the way, if such a tail picking game device is mounted on a vehicle and used as described in claim 4, it is possible to prevent an important problem such as frustration during a traffic jam and drowsy driving. Very effective. From the viewpoint of preventing drowsy driving, for example, as shown in claim 5, when there is no voice input from the user for a predetermined time, a predetermined message is output by voice.
It is desirable to configure the utterance means. In other words, if the voice input from the user is cut off during the process of rearing, it is judged that the driver is likely to fall asleep, and the user is asked "Do you not understand?" Or "Are you losing?" Message such as "is output. If there is still no response from the user, it is effective to output a message such as "What's wrong?". At this time, the volume of the voice should be increased.

In recent years, the number of vehicles equipped with a navigation device has increased, and as described above as a conventional technique, a navigation device capable of inputting a destination setting and the like by voice is prevalent. However,
The voice recognition function is used only very little during the use time of the navigation device, such as when setting a destination.
Therefore, it is conceivable to adopt the configuration shown in claim 6.

That is, the skirt removal game device is provided with a skirt removal game device and a navigation device having a voice recognition function, and the voice recognition means described above is configured by utilizing the configuration for voice recognition included in the navigation device. It is a navigation system with a game function. According to this system, the voice recognition function of the navigation device can be effectively used. In addition, the effect of preventing frustration during a traffic jam and drowsy driving can be naturally obtained.

Similarly, for example, when the navigation device has a structure for utterance, it is conceivable that the utterance means described above is configured by using the structure for utterance of the navigation device. By thus sharing a part of the configuration with the navigation device, the manufacturing cost can be suppressed as compared with the case where the tail picking game device is manufactured alone.

By the way, considering the configuration for preventing the dozing driving as described above by using the navigation system with a tail picking game function, for example, as shown in claim 6, the utterance means further includes a voice from the user. It is conceivable that a predetermined message is output by voice when there is no input for a predetermined time and when the vehicle speed detected by the navigation device exceeds a predetermined value.

The utterance means of the bouncing game device according to claim 5 is such that when there is no voice input from the user for a predetermined time,
The predetermined message is output by voice.
Therefore, this message is output even when the vehicle is stopped. On the other hand, in the navigation system with a tail picking game function of the present invention,
When there is no voice input from the user for a predetermined time and when the vehicle speed detected by the navigation device exceeds a predetermined value, a predetermined message is output as a voice.
That is, the message output such as when the vehicle is stopped disappears,
It is possible to output a message for preventing the dozing driving at a more appropriate timing.

When realized as a navigation system with a tail removal game function, it is possible to use the information of the navigation device for "butt removal". For example, as described in claim 8, a word associated with a region is prepared in the utterance target words, and the utterance word selecting means further relates to a region set / detected by the navigation device. It can be considered that a word is configured to be preferentially selected from the utterance target words. For example, when the destination set by the navigation device is Nagoya, words such as "Miso Katsuka" that are deeply related to the area of Nagoya are preferentially output. by this,
For example, a word that reflects the land pattern of the destination will be used for “swifting”, and for example, “switting” during driving can be made more enjoyable.

[0026]

BEST MODE FOR CARRYING OUT THE INVENTION An embodiment of the present invention will be described below with reference to the drawings. The present invention is not limited to the following embodiments,
It goes without saying that various forms can be implemented without departing from the spirit of the present invention.

FIG. 1 is a navigation system with a tail picking game function according to the present embodiment (hereinafter referred to as "system").
It is a functional block diagram of. The system of the present embodiment is mounted on a vehicle and used, and has a navigation function for guiding a route to a set destination and a butt removal game function for performing a word play, "ass removal" with a user. Is. The system of the present embodiment includes an input unit 10 and
Control unit 20, display control unit 30, and audio output control unit 40
The nonvolatile memory 50, the current position calculation unit 60, the vehicle speed detection unit 70, the map data search unit 80, and the guidance control unit 90 are provided. The input unit 10 also includes a voice recognition unit 11, an operation unit 12, and a light receiving unit 13 that receives an optical signal from a remote controller (not shown).
A display 31 using a CRT or liquid crystal is connected to the display control unit 30, and a speaker 41 is connected to the audio output control unit 40.

The input unit 10 is for a user to give an instruction to the system, and enables voice input through the voice recognition unit 11 and operation input through the operation unit 12 or the light receiving unit 13. ing. The operation unit 12 includes an operator such as a push button switch, and when a user operates the operator, the operation unit 12 outputs a signal corresponding to the operator to the control unit 20. A remote controller (not shown) is also provided with an operator such as a push button switch, and when the user operates the operator of the remote controller, the light receiving unit 13
Receives an optical signal transmitted from the remote controller corresponding to the operator and outputs a signal corresponding to the optical signal to the control unit 20.

The non-volatile memory 50 can retain the stored data even when the ignition switch of the vehicle is turned off and the power supply to the system is stopped. The current position calculation unit 60 performs calculation for grasping the current position of the vehicle. The vehicle speed detector 70
The vehicle speed is detected based on the pulse signal from the vehicle speed sensor. Further, the map data search unit searches the map data stored in advance when the destination is set by the user via the voice recognition unit 11 of the input unit 10, and the current position calculation unit 60
A route connecting the current position calculated by and the set destination is searched. Then, the guidance control unit 90 guides the route searched by the map data search unit 80. The message for this guidance is output as voice from the speaker 41 by the voice output control unit 40.
The map for guidance is displayed on the display 31 by the display control unit 30. Since a configuration for realizing such a navigation function is well known, detailed description will be omitted.

Now, the voice recognition unit 11 of the input unit 10 will be described with reference to FIG. FIG. 2 is a functional block diagram of the voice recognition unit 11. The voice recognition unit 11 includes an input signal processing unit 11a, a recognition processing unit 11b, a recognition control unit 11c, and a dictionary management unit 11d.

The input signal processing section 11a A / D-converts the voice input by the user and converts it into voice data which is data-processable waveform data in the frequency direction, for example.
The recognition processing unit 11b outputs, as a recognition result, a recognition target word having a similarity of a predetermined value or more from among the recognition target words in the dictionary set by the dictionary management unit 11d described later. Note that the recognition result output here is not limited to one since the similarity may be equal to or greater than a predetermined value. This recognition result is output to the control unit 20 via the recognition control unit 11c. The recognition control unit 11c specifies the dictionary to be used based on the information from the control unit 20. The dictionary management unit 11d
The dictionary designated by the recognition control unit 11c is set. The dictionary will be described later.

The characteristic part of the present system is that the voice recognition unit 11 described above can be used to allow the user to play a word-playing game, "skipping". Therefore, next, the tail removal execution processing executed in the system of the present embodiment will be described in detail based on the flowchart of FIG. This process is a process that is executed when the user gives an instruction to perform tail removal via the input unit 10.

First, in step S100, game contents are set. This process is for the user to set the rules and the like when playing the tail picking game. In the present embodiment, it is possible to set the ending condition (condition when "losing"), genre, device level, number of users, name, and order. Here, the display control unit 10 causes the display 31 to display a menu for performing these settings. The user makes settings via the input unit 10 according to the menu displayed on the display 31. The game content set by the user is stored in the non-volatile memory 50. Since the non-volatile memory 50 is used, the game content stored in the non-volatile memory 50 may be selected when the tail removal execution processing is executed again after the tail removal execution processing is completed. it can. As a result, the trouble of setting the game every time is eliminated, and the convenience of the user can be improved.

Here, the contents of the game set by the user will be described. (1) Termination condition If a word with "n" at the end is used, it will naturally be "lost", but in addition to this, prohibited sounds and prohibited words other than "n" can be set. With this setting, the player loses when using this prohibited sound or prohibited phrase. Further, both a setting of "losing" when a forbidden sound and a prohibited word are used at the end and a setting of "losing" when the forbidden sound and a prohibited word are included somewhere in the word can be selected.

If the previously used words are used,
It is possible to select whether or not to lose each of the cases where the homonyms are used. In the present embodiment, the above-described settings can be made as the end condition. However, for example, an answer time limit can be set such that if you do not answer within 5 seconds, you will be "losed". Alternatively, for the purpose of preventing the “deepening” from continuing, it may be possible to set “draw” when the exchanges are performed a predetermined number of times or more. (2) Genres such as person name, food, car type, place name, etc. can be selected. For example, when a person's name is designated as the genre, the person's name is used as "Noricho Oda"->"GamoUjisato"->"HideyoshiToyotomi". If you say a word that does not fit the selected genre, you will lose. (3) Level on the device side The level on the device side refers to the strength of the game on the device side when performing tail removal. In this embodiment, a plurality of opponent characters are prepared, and the strength of each character is set. Therefore, the user selects the level on the device side by selecting the character.
The strength of the character is determined mainly by the number of usable vocabularies and the probability of actively selecting a word that meets the above-mentioned termination condition. The former is realized by setting a usable word and an unusable word for each character in the utterance target words stored in the tail picking dictionary, as described later. At this time, it is conceivable to increase or decrease the usable vocabulary of a specific genre to bring out the characteristics of each character. In the latter case, for example, the final sound of the ending is "n".
That is, the word that becomes will be deliberately selected with a certain probability from the dictionary for tailing. If this probability is relatively high, it will be weak, and if it is relatively low, it will be strong. (4) User name, number of people, order This is set when more than one person enjoys butt removal. For example, in the case of "assaulting" with two people and this system,
It is possible to set a game progression order of user A → user B → system → user A → ... In addition, the user A can set the name “Mr. XX” and the user B can set the name “Mr. XX”.

By enabling the various settings as described above, the user can further enjoy the "tailing". Returning to the description of the flowchart of FIG. When the game settings have been made as described above, in S110, "starting" is started. Here, one note is randomly determined. The user will say the first word with this one sound as an acronym.

In subsequent S120, a voice input request is made.
This processing requests the user to input a voice. Here, the speaker 4 is controlled by the audio output control unit 40.
A message for prompting the user to input a voice is output from 1. For example, it shows the last sound of the word just before, and says, "Next is'pa '. Please.". If there is no word immediately before, that is, immediately after the start of "cutting", the one sound determined in S110 is indicated.

Therefore, in S130, it is determined whether or not a voice input is made. This judgment is made by the voice recognition unit 11. If it is determined that voice input is made here (S
130: YES), and proceeds to S135. On the other hand, if it is determined that no voice is input (S130: N
O), and shifts to S230.

At S135, voice recognition processing is performed. This process is a process in the voice recognition unit 11 of the input unit 10. This voice recognition process will be described later. In S140, it is determined whether the recognition result is wrong. Here, the voice output control unit 40 outputs a message to the effect that the recognition result is confirmed from the speaker 41, then waits for a predetermined time, and within the predetermined time, the user inputs that the recognition result is wrong. If yes, affirmative decision is made.
On the other hand, if no input is made by the user within the predetermined time, a negative determination is made. If it is determined that the recognition result is incorrect (S140: YES), the process from S120 is repeated. On the other hand, when it is determined that the recognition result is correct (S1
40: NO), and shifts to S150.

In S150, it is determined whether or not there is a special instruction. This is a process of determining whether or not the recognition result from the voice control unit 11 is “interruption”. If it is determined that there is a special instruction here (S150: YES), S
Move to 160. On the other hand, when it is determined that there is no special instruction (S150: NO), the process proceeds to S180.

In S160, the fact that the process is interrupted is displayed. Here, the display control unit 30 displays an indication on the display 31 that the display has been interrupted. In the present embodiment, S
In 150, it is configured to judge an instruction of "interruption" as a special instruction. For example, a "help" instruction for asking the system for help when an appropriate word cannot be conceived,
You may judge the instruction | indication of "pass" which turns the right of reply to the next, without answering.

In S170, it is determined whether or not there is an instruction to restart the tail removal. This process is for determining whether or not there is a restart instruction input from the user via the input unit 10 after the process is temporarily stopped. If it is determined that the restart instruction has been issued (S170: YES), the process proceeds to S120 and the subsequent processes are repeated. On the other hand, when it is determined that there is no restart instruction (S170: NO), this determination process is repeated. If there is no restart instruction, the minimum necessary power is supplied to the CPU to perform power saving management.

In S180, it is determined whether or not the word recognized in S135 matches the ending condition set in S100. If it is determined that the end condition is met (S180: YES), the end process is executed in S220, and then the main tail removal execution process is ended. On the other hand, if it is determined that the termination conditions are not met (S1
80: NO), and proceeds to S190.

In S190, it is determined whether it is the user's turn next. This determination is made when a plurality of users are set in S100. For example, if the order of progression of the game is user A → user B → system, if the next is the order of user B, an affirmative determination is made here. If it is determined that the next is the user's turn (S190: YES), the process from S120 is repeated. On the other hand, when it is determined that the next is not the user's turn (S190: NO), the process proceeds to S200.

In S200, a speech word is selected. This process is performed from the utterance target words stored in the tail picking dictionary,
The word whose first letter of the word recognized in S130 is an acronym is selected. Here, the tail removing dictionary will be described. The tail removing dictionary is a file created in the format shown in FIG. As shown in FIG. 5, genres, area information, and usable characters are set in correspondence with words. When the tail genre is set to "food" in S100, the word classified as food is searched as the utterance target word.

Also, the number of characters that can be used is set in each word. In FIG. 5, the character "1"
Is "Miso Katsura", "Gorilla", "Tori", "Apple"
It indicates that all of the words illustrated as can be used. The character "2" can use "gorira" and "apple" in the exemplified words, and the character "3" can use only "apple" in the exemplified words. Therefore, a word that can be used by the character selected as the opponent in S100 is searched for as the utterance target word.

After that, one word is selected from the searched utterance target words. In the present embodiment, the utterance target word is stored in association with the area information. For example, regional information “Nagoya” is stored for “Miso Katsura”. Then, the word related to the current position of the vehicle or the set destination which is repeatedly calculated by the current position calculation unit 60 is preferentially selected from the searched utterance target words.
As a result, the word reflecting the land pattern that the vehicle is traveling or intended for is used for the “tail removal”, and for example, the “back removal” during driving can be made more enjoyable.

In S210, it is determined whether or not the word selected in S200 meets the ending condition. As described above, the probability of selecting a word that matches the ending condition is set for each character, and the word that matches the ending condition is positively selected with that probability. Here, an affirmative decision is made when a word that matches the ending condition is selected in S200. If the ending condition is met (S210: YES), the ending process is executed in S220, and then the main tail removal executing process is ended. On the other hand, if the termination conditions are not met (S210: N
O), the process from S120 is repeated.

The ending process of S220 is to output by voice which of the user and the device (the character selected as the opponent) wins and which loses. That is, the voice output control unit 40 outputs the voice that conveys the win or loss from the speaker 41. If an affirmative decision is made in the processing of S180, it means "the user has lost" and S21
If the affirmative judgment is made in the processing of 0, "the system loses"
Becomes After that, the main tail removal execution process ends.

In S230, which is executed when no voice is input in S130, it is determined whether a predetermined time has elapsed. If it is determined that the predetermined time has elapsed (S230: YES), that is, if there is no response from the user for the predetermined time, the doze driving prevention process is called in S240. On the other hand, if the predetermined time has not elapsed (S230: NO), the process proceeds to the determination process of S130.

Next, the drowsy driving prevention process called in S240 will be described with reference to the flowchart of FIG. First, in step S300, the voice output control unit 40 outputs a predetermined message from the speaker 41. For example, a message such as "Don't you know?" Or "Are you sure you want to lose?" Is output.

In S310, it is determined whether or not there is a response from the user within a predetermined time such as 3 seconds. If there is a response from the user here (S310:
YES), the process returns to the tail removal execution process of FIG. 3, and the process from S120 is repeated. On the other hand, when there is no response from the user (S310: NO), the process proceeds to S320.

In S320, it is determined whether the vehicle speed is less than or equal to a predetermined value. This process is to determine whether the vehicle speed detected by the vehicle speed detector 70 is equal to or lower than a predetermined value. When it is determined that the vehicle speed is equal to or lower than the predetermined value (S320: YES), the process proceeds to S330. On the other hand, when it is determined that the vehicle speed exceeds the predetermined value (S320:
No), because there is no response even while driving,
Assuming that there is a possibility of drowsy driving, the volume is increased in S350, and the process proceeds to S300. In S300, a message such as "What happened?" Is output.

In S330 and subsequent S340, the same interrupting process and restart instruction determining process as in S160 and S170 described above are performed. Next, the voice recognition process of S130 described above will be described. The recognition control unit 11c shown in FIG.
When the tail removal is started, the tail management dictionary is designated to the dictionary management unit 11d, and the dictionary management unit 11d sets the tail management dictionary as a search target. As a result, the recognition processing section 11b carries out voice recognition based on the tail picking dictionary. This tail picking dictionary is the same as the dictionary in which the utterance target word is stored.

When the input voice from the user is converted into time-series voice data by the input signal processor 11a, the recognition processor 11b first applies the same pattern to the first portion of the time-series voice data. Determine if there are more than 3 repetitions. And the same pattern is repeated 3 times.
If it is more than once, the voice data is modified so that this pattern appears only once at the beginning of the voice data. This focuses on the fact that when a user recalls a word with the acronym "go", for example, the word "gogogo, gorilla" may be pronounced repeatedly. The voice data indicating "gogora" is corrected to the voice data indicating "gogora". It should be noted that the reason why the determination is made three times or more is that the same sound such as “haha (mother)”, “peach (peach)”, and “mimi (ear)” is repeated twice. This is because there are relatively many words and relatively few words in which the same sound is repeated three times.

After that, the recognition processing unit 11b calculates the similarity between the pattern of the voice data and the pattern of the recognition target word stored in the tailing dictionary set by the dictionary management unit 11d, and the similarity is predetermined. The recognition target word having a value or more is output to the recognition control unit 11c as a recognition result. When there are a plurality of words as the recognition result, the recognition control unit 11c outputs to the control unit 20 the recognition result of the last one sound of the immediately preceding word or the word with the one sound determined in S110 as an acronym. For example, when there are two words output from the recognition processing unit 11b such as "Wakayama (Wakayama)" and "Okayama (Okayama)", the last one sound of the immediately preceding word or the determined one sound is "wa". If it is, it means that "Wakayama (Wakayama)" is the recognition result.

Next, the effect of the system of this embodiment will be described. The system according to the present embodiment enables the word play “sushiri” with the user. In the “sushiri”, as described above, for example, a word with the acronym “go” is recalled. When you do, "go, go, gora"
The acronym is often pronounced repeatedly. Therefore, in this embodiment, before performing pattern matching between the voice data based on the input voice and the recognition target word stored in the tailing dictionary, when the same pattern is repeated a predetermined number of times in the first part of the voice data, Modifies the audio data so that this pattern appears only once. As a result, for example, after the voice data indicating “gogogogora” has been corrected to voice data indicating “gogora”, pattern matching with the recognition target word in the dictionary is performed.
When you take "butt removal", you can prevent the voice recognition rate from decreasing and you can enjoy the smooth adjustment.

Further, in the system of the present embodiment, the recognition control section 11c, when there are a plurality of recognition results from the recognition processing section 11b, makes the last one sound of the immediately preceding word or the determined one sound an acronym. The word that does is the recognition result. by this,
It is possible to further improve the voice recognition rate when executing “tail removal”.

Furthermore, by adding such a tail picking game function to the navigation system mounted on the vehicle, it is possible to prevent the frustration and the dozing driving during the traffic jam. To prevent this drowsy driving, when the response from the user is interrupted during the bottoming process (S in Fig. 3).
230: YES), a voice message is output (S300 in FIG. 4), and if there is still no response (S3 in FIG. 4)
10: NO), the volume is increased (S35 in FIG. 3).
0), output a voice message (S30 in FIG. 3)
0). As a result, the prevention of drowsy driving can be effectively realized. Even if there is no response from the user (S310: NO in FIG. 3), when the vehicle is stopped (S320: YES in FIG. 3), the voice message is not output. did. As a result, it is possible to output the doze-driving prevention message at an appropriate timing.

In addition, recently, a navigation device capable of inputting a destination setting and the like by voice has become widespread.
The voice recognition function was used very little during the usage time of the navigation device, such as when setting a destination. Therefore, the system according to the present embodiment uses the voice recognition unit 11 for setting the destination to perform “bottom cutting”. In this way, since the configuration of the conventional navigation device is partially used, it is possible to suppress the manufacturing cost for realizing the tail picking game function.

The voice recognition unit 11 described above corresponds to a "voice recognition unit", the control unit 20 corresponds to a "speech word selection unit", and the voice output control unit 40 corresponds to a "speech unit".

[Brief description of drawings]

FIG. 1 is a functional block diagram of a navigation system with a tail picking game function according to an embodiment.

FIG. 2 is a functional block diagram of a voice recognition unit of the navigation system with a tail picking game function according to the embodiment.

FIG. 3 is a flowchart showing a tail removal execution process.

FIG. 4 is a flowchart showing a drowsiness driving prevention process called from the tail removal execution process.

FIG. 5 is an explanatory diagram illustrating the configuration of a tail removing dictionary.

[Explanation of symbols]

10 ... Input part 11 ... Voice recognition part 11a ... Input signal processing part 11b ... Recognition processing part 11c ... Recognition control part 11d ... Dictionary management part 12 ... Operation part 13 ... Light receiving part 20 ... Control part 30 ... Display control part 31 ... Display 40 ... Voice output control unit 41 ... Speaker 50 ... Non-volatile memory 60 ... Current position calculation unit 70 ... Vehicle speed detection unit 80 ... Map data search unit 90 ... Guidance control unit

─────────────────────────────────────────────────── ─── Continued Front Page (51) Int.Cl. ⁷ Identification Code FI Theme Coat (Reference) G10L 11/00 G10L 3/00 531C 551H 9/00

Claims (8)

[Claims] 1. A voice recognition means for recognizing the inputted voice when a voice input is made by a user, and a word having an acronym as the last one sound of the word recognized by the voice recognition means. A fluttering game device provided with utterance word selecting means for selecting from the stored utterance target words and utterance means for outputting the word selected by the utterance word selecting means as voice In the case where the same sound at the beginning of the voice input by the user is repeated a predetermined number of times or more, the voice recognition unit regards the repeated sound as one or two sounds and inputs the same. An assault-taking game device, which is configured to recognize a voice that is generated. 2. The tail removing game device according to claim 1, wherein the predetermined number of times is two or three times. 3. The tail picking game device according to claim 1, wherein the voice recognition unit recognizes an input voice from the user, immediately before the speech target word selection unit or the voice recognition unit. A tail picking game device, which is configured to recognize the last one sound of the selected or recognized words as an acronym. 4. The bottom removing game device according to claim 1, wherein the rear removing game device is mounted on a vehicle for use. 5. The tail picking game device according to claim 4, wherein the utterance means is further configured to output a predetermined message by voice when voice input from the user is not made for a predetermined time. Butt removal game device. 6. A navigation system with a tail picking game function, comprising the tail picking game device according to claim 4 and a navigation device having a voice recognition function, wherein the voice recognition means comprises a voice included in the navigation device. A navigation system with a tail picking game function, characterized by being configured using a configuration for recognition. 7. The navigation system with a tail picking game function according to claim 6, wherein the utterance means is further detected when there is no voice input from the user for a predetermined time, and is detected by the navigation device. A navigation system with a tail picking game function, which is configured to output a predetermined message by voice when the vehicle speed exceeds a predetermined value. 8. The navigation system with a tail picking game function according to claim 6 or 7, wherein a word associated with a region is prepared in the utterance target word, and the utterance word selecting means is provided. Further, the navigation system with a tail picking game function is configured to preferentially select a word related to a region set / detected by the navigation device from the utterance target words.