JP4498902B2 - Voice recognition device - Google Patents

Description

  The present invention relates to a voice recognition device used for operating a navigation system with voice, for example, and more particularly to a technique for improving the operability.

  2. Description of the Related Art Conventionally, a voice recognition device that recognizes voice spoken by a user and outputs a recognition result is known. As such a speech recognition device, for example, Patent Document 1 discloses a command receiving device for promoting the use of voice command input. This command accepting device can accept both manual operation and voice operation using a touch panel, remote control, etc. When a command capable of voice input is manually input via a switch device or the like, a voice recognition function is used. Since these commands can also be recognized, for example, an adaptive screen and a command are displayed on the display screen of the display device to guide (teach) that the voice command input can be performed. For example, if the menu screen is displayed by manual input operation, “You can operate by voice input. Please say “Menu screen” on the “Map screen”. Is displayed. If it is a voice guide, the same content is uttered from the speaker for guidance. If there is such guidance, it is expected that the user will adopt a simple command input method by voice input instead of an input method such as manual input that is considered to be relatively troublesome from the next time. .

  Patent Document 2 discloses a traveling position display device that can notify a notification device together with a map of the current position of a car or the like by a key input operation or voice input. The travel position display device includes voice command registration means for registering the conversion result of the voice recognition means as a voice command in association with the key event input from the key input means. Thus, the recognizable voice can be expanded by registering an arbitrary key operation sequence designated by the user in one voice command.

JP 2003-114698 A JP 2001-117586 A

  However, in the conventional voice recognition apparatus described above, it is determined only whether or not there is a voice command corresponding to the manual operation, and if the command input by the manual operation can be input by voice, the voice command is input to the user. Therefore, when it is necessary to input by a plurality of steps of manual operation until the purpose is achieved, a voice command is presented at each step in the middle stage until the purpose is achieved. As a result, there is a problem that a series of voice operation methods until the user achieves the purpose cannot be understood.

  Further, in the conventional voice recognition device, it is necessary to memorize a voice command by having a voice dialogue function in order to determine whether or not a voice command corresponding to a manual operation exists and present the voice command to the user. Even if not, a voice command is presented. As a result, there is a problem that the user remembers unnecessary voice commands.

  Furthermore, in the conventional speech recognition device, since all speech recognition dictionaries are always set to the valid state, there is a problem that the user remembers or misrecognizes a speech command that is ambiguously remembered and tried, and There is a problem that it cannot be determined whether or not the spoken voice command really exists.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to provide a voice recognition device with excellent operability that can present a voice command suitable for voice operation of a device to a user. is there.

A voice recognition apparatus according to the present invention includes a voice input means for inputting voice, a voice recognition means for recognizing a voice input from the voice input means as a voice command, a manual input means for performing manual input, and a manual input A state management table storing whether or not a state generated by manual operation of the means is a state where the intended purpose has been achieved, and a state generated by manual operation of the manual input means and input from the voice input means Manual input when the voice command correspondence table storing the correspondence with the voice command and the state sequentially generated by manual operation of the manual input means are compared with the contents of the state management table to determine that the objective has been achieved. If a voice command corresponding to the goal achievement state caused by manual operation of the means is present in the voice command correspondence table, the manual input means is set by the voice command. And a control means for generating a message indicating that can realize a function corresponding to the manual operation, and output means for outputting the messages generated by the control means, the control means measures the time the manual input means is not manually operated A timer unit for checking the state sequentially generated by manual operation of the manual input means against the contents of the state management table and determining that the target achievement state has been reached. When it is measured that no operation has been performed, if there is a voice command corresponding to a state caused by manual operation of the manual input means up to that time, the voice command corresponding to the manual input means can be manually operated by the voice command. A message indicating that the function corresponding to the operation can be realized is generated .

According to the present invention, both manual operation and voice operation by voice command are possible, and when manual operation is performed a plurality of times in order to achieve the intended purpose, the manual input means sequentially performs manual operation. When the generated state becomes the goal achievement state, if there is a voice command corresponding to the goal achievement state, a message indicating that the function corresponding to the manual operation of the manual input means can be realized by the voice command is generated. Output. Accordingly, in the middle of reaching the goal achievement state, even if there is a voice command corresponding to the manual operation, it is not output. Therefore, the user can understand the voice command optimum for achieving the goal, and the series of steps until the goal is achieved. It becomes easy to understand the voice operation method. As a result, it is possible to provide a voice recognition device with excellent operability that can present a voice command suitable for voice operation of the device to the user. In addition, when the voice recognition device is not operated for a certain period of time while operating the remote control to achieve the purpose, the state is regarded as the purpose achievement state, and manual operation until reaching the state is performed. Since the corresponding voice command is presented to the user, the user can know the voice command corresponding to the manual operation for any state.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
Embodiment 1 FIG.
The voice recognition apparatus according to the first embodiment of the present invention does not present a voice command in the middle of manual input, but presents a voice command to the user when a goal achievement state is reached.

  FIG. 1 is a block diagram showing a configuration of a speech recognition apparatus according to Embodiment 1 of the present invention. The speech recognition apparatus includes a speech recognition processing device 1, a remote controller (hereinafter abbreviated as “remote control”) 2, a microphone (hereinafter abbreviated as “microphone”) 3, a monitor 4 and a speaker 5.

  The remote controller 2 is used for manually operating the voice recognition device. The remote controller 2 generates a key code corresponding to the key (not shown) in response to an operation of the key (not shown), and sends the key code to the voice recognition processing device 1 by, for example, infrared rays. As a device used for manual operation, other input means such as a touch panel and a push button switch can be used instead of the remote controller 2. The microphone 3 is used for operating the voice recognition apparatus by voice. The microphone 3 converts the voice uttered by the user into an electric signal and sends it to the voice recognition processing device 1 as an analog voice signal.

  The monitor 4 displays an image or a telop on its screen in accordance with the analog video signal sent from the voice recognition processing device 1. The monitor 4 can be composed of, for example, a liquid crystal display device or a CRT device. The speaker 5 outputs a sound in accordance with an analog sound signal sent from the sound recognition processing device 1.

  The voice recognition processing device 1 includes a key input unit 6, a voice input unit 7, a control unit 8, an operation history storage unit 9, a state management table 10, a voice command correspondence table 11, a voice guide generation unit 12, a screen telop generation unit 13, The voice output unit 14, the screen output unit 15, the voice recognition unit 16, the voice recognition dictionary 17, and the recognition dictionary management unit 18 are configured. The manual input means of the present invention corresponds to the key input means 6. The output means of the present invention corresponds to the audio output means 14 and the screen output means 15.

  The key input means 6 receives a key code sent from the remote controller 2 by infrared rays and converts it into a digital electric signal. The key code converted into a digital electric signal by the key input means 6 is sent to the control means 8. The voice input means 7 is composed of, for example, an A / D converter, inputs an analog voice signal generated by the microphone 3 when a person speaks, and digital voice data that can be handled by the voice recognition means 16. Convert to The voice data obtained by the conversion in the voice input means 7 is sent to the voice recognition means 16.

  The control unit 8 includes a key input unit 6, an operation history storage unit 9, a state management table 10, a voice command correspondence table 11, a voice guide generation unit 12, a screen telop generation unit 13, a voice recognition unit 16, and a recognition dictionary management unit 18. The entire voice recognition processing device 1 is controlled by transmitting and receiving data between the two. Details of the control means 8 will be described later.

  The operation history storage means 9 stores a history of key operation states sent from the remote controller 2 by manual operation of the user and a state history of the voice recognition device changed by processing executed in response to the key operations. To do. The operation history storage unit 9 is accessed by the control unit 8.

  The state management table 10 is used for managing the key operation state of the remote controller 2 and the state of the voice recognition device. FIG. 2 shows an example of information stored in the state management table 10. This state management table 10 stores various states (manual input / state) of the speech recognition apparatus and information indicating whether or not the state is an objective achievement state. Here, the purpose achievement state is a state in which a series of operation procedures for obtaining desired information is completed, for example, a state in which a destination is set, a state in which peripheral facilities are selected, a state in which a place or facility is displayed, and the like. be able to. In the example shown in FIG. 2, as the goal achievement status, “Destination set” telop is displayed when the destination setting is completed, or “Search phone key for searching phone number” is pressed when the phone number search ends. "Is defined. This state management table 10 is accessed by the control means 8.

  The voice command correspondence table 11 stores a list of voice commands corresponding to the key operation state of the remote controller 2 and the state of the voice recognition device. FIG. 3 shows an example of information stored in the voice command correspondence table 11. The voice command correspondence table 11 stores various states (manual input / state) of the voice recognition device and voice commands corresponding to the states when there are voice commands for realizing the states. . This voice command correspondence table is referred to by the control means 8.

  The voice guide generation unit 12 generates voice output data to be presented to the user by voice through the speaker 5 in response to an instruction from the control unit 8. The voice output data generated by the voice guide generation unit 12 is sent to the voice output unit 14. The screen telop generation unit 13 generates screen output data to be presented to the user as a telop via the monitor 4 in accordance with an instruction from the control unit 8. The screen output data generated by the screen telop generation means 13 is sent to the screen output means 15.

  The audio output means 14 is composed of, for example, a D / A converter, converts the audio output data sent from the audio guide generation means 12 into an analog audio signal, and sends it to the speaker 5. Thereby, a message corresponding to the voice output data, that is, a voice guide is output from the speaker 5. The screen output means 15 is composed of, for example, a D / A converter, converts the screen output data sent from the screen telop generation means 13 into an analog video signal, and sends it to the monitor 4. As a result, a message corresponding to the screen output data, that is, an image including a screen telop is displayed on the monitor 4.

  The speech recognition means 16 performs speech recognition processing with reference to the speech recognition dictionary 17. The voice recognition means 16 analyzes the voice data sent from the voice input means 7 and performs a matching process for comparing the analysis result with the vocabulary acquired from the voice recognition dictionary 17 via the recognition dictionary management means 18. . The recognition result obtained by this matching process is sent to the control means 8. The speech recognition dictionary 17 includes a plurality of recognition dictionaries classified according to vocabulary types, for example, an “address dictionary” including vocabularies (prefecture names, municipalities, etc.) used for addresses, and vocabularies used for facility names (Tokyo Tower). Etc.), a “telephone number dictionary” including vocabulary (numbers) used for telephone numbers, and the like. Each recognition dictionary has a unique dictionary number.

  The recognition dictionary management means 18 manages a plurality of recognition dictionaries constituting the speech recognition dictionary 17 using a dictionary management table 19. FIG. 4 shows an example of information stored in the dictionary management table 19. The dictionary management table 19 stores “recognized vocabulary information”, “dictionary number”, “dictionary state”, and “last use date” as dictionary information. The recognition vocabulary information includes a plurality of vocabularies (voice commands) included in the recognition dictionary. The dictionary number is a unique number assigned to each recognition dictionary. The dictionary state is either a valid state or an invalid state. If the dictionary state is valid, the voice command corresponding to the recognition dictionary can be recognized, but if the dictionary state is invalid, it cannot be recognized. The last use date and time is the date when the recognized vocabulary (voice command) is used, and is updated each time the recognition dictionary is referenced.

  This recognition dictionary management means 18 changes the dictionary state of the dictionary management table 19 in response to a dictionary switching request sent from the control means 8, and selects one recognition dictionary to be recognized. The recognition dictionary management means 18 reads the contents of the selected recognition dictionary and sends them to the voice recognition means 16 when voice recognition is performed.

  Next, details of the control means 8 will be described. The control unit 8 includes a state management unit 81, a state confirmation unit 82, a voice command acquisition unit 83, a timer unit 84, a recognition dictionary state change unit 85, a recognition dictionary invalidation unit 86, and a current time acquisition unit 87.

  When a key code is sent from the key input unit 6, the state management unit 81 changes the state of the voice recognition device so as to correspond to the key code, and stores the changed state in the operation history storage unit 9. To do. The state confirmation unit 82 acquires the current state from the operation history storage unit 9 and compares the current state with information stored in the state management table 10 to determine whether or not the state has been achieved. Then, if the objective is achieved, the current state is sent to the voice command acquisition unit 83 to instruct acquisition of the voice command. On the other hand, if the objective is not achieved, the current state is sent to the timer unit 84 to start measuring time.

  When the current state is sent from the state confirmation unit 82 or the timer unit 84, the voice command acquisition unit 83 acquires a voice command corresponding to the current state from the voice command correspondence table 11. Then, the voice command acquired from the voice command correspondence table 11 is sent to the voice guide generation unit 12 and the screen telop generation unit 13. Further, the voice command acquisition unit 83 sends the voice command acquired from the voice command correspondence table 11 to the recognition dictionary state change unit 85.

  The timer unit 84 measures the duration time when there is no key input from the remote controller 2. The timer unit 84 monitors the key input means 6 even when the timer is operating. If there is a key input during the timer operation, the measurement is stopped. On the other hand, if there is no key input during the timer operation, the current state received from the state confirmation unit 82 is sent to the voice command acquisition unit 83.

  The recognition dictionary state change unit 85 specifies a recognition dictionary that includes the voice command sent from the voice command acquisition unit 83, and sends a dictionary switching request to the recognition dictionary management unit 18 so as to switch to the specified recognition dictionary. The recognition dictionary invalidation unit 86 sends a request for changing the unused dictionary to the invalid state after the speech recognition apparatus is activated, to the recognition dictionary management means 18. The current time acquisition unit 87 acquires the current time (year / month / day) from a clock (not shown) and sends it to the recognition dictionary management means 18.

  Next, the operation of the speech recognition apparatus according to Embodiment 1 of the present invention configured as described above will be described. In the following description, a case where the speech recognition apparatus is applied to a car navigation system (hereinafter abbreviated as “car navigation”) will be described as an example.

  First, a manual operation procedure for setting a XX park searched using the facility search function of the car navigation system as a destination will be described with reference to FIG. In this case, as shown in FIG. 5, the [Menu] key is pressed (Step 1) → [Set destination] is selected (Step 2) → [Set from name] is selected (Step 3) → [XX Park ] (Step 4)-> select [list] (step 5)-> select name (step 6)-> press [execute] key (step 7).

  In the above-described manual operation procedure, for example, there is a voice command “menu screen” corresponding to “press the [menu] key” in step 1, and if “menu screen” is spoken, the same operation as “pressing the [menu] key” is performed. Although it can be realized, since it is an operation in the middle of achieving the purpose, no voice command is presented. In the speech recognition apparatus according to the first embodiment of the present invention, for the above-described manual operation, for example, “For voice,“ A message such as “You can operate by speaking“ ○ Park ”→“ Set as destination ”” is presented to the user.

  Next, the operation of the speech recognition apparatus for realizing the above-described function will be described with reference to the flowchart shown in FIG.

  When the voice recognition device is activated by turning on the power, it is first checked whether or not there is a key input (step ST1). Specifically, the state management unit 81 of the control unit 8 checks whether a key code is sent from the key input unit 6. If it is determined in step ST1 that there is no key input, a standby state is entered while step ST1 is repeatedly executed at a constant cycle. When it is determined that there is a key input in a standby state by repeatedly executing step ST1, a manual input command C1 is acquired (step ST2).

  Next, the current state is acquired from the state management unit 81 and is set to the initial state Q1 (step ST3). Next, the initial state Q1 acquired in step ST3 is stored in the operation history storage means 9 (step ST4), and processing for the received manual input command is performed (step ST5). For example, when the manual input command indicates pressing of the menu key, a process for displaying a menu screen is performed as a process corresponding thereto. As described above, since the state is changed by processing for the command, the current state is acquired from the state management table 10 and is set to the current state Q2 (step ST6). Then, the manual input command C1 and the current state Q2 are stored in the operation history storing means 9 (step ST7).

  Next, it is checked whether or not the current state Q2 is a goal achievement state (step ST8). That is, the state confirmation unit 82 acquires the current state Q2 from the operation history storage unit 9, and refers to the information (see FIG. 2) stored in the state management table 10, so that the current state Q2 is the goal achievement state. (Whether the goal achievement state is ○ or ×). If it is determined in step ST8 that the objective is not achieved, the sequence returns to step ST1 in order to wait for the next manual input. On the other hand, if it is determined in step ST8 that the objective is achieved, the state confirmation unit 82 passes the current state Q2 to the voice command acquisition unit 83 and determines whether or not it is necessary to present the voice command. Therefore, the process proceeds to step ST9.

  In step ST9, it is checked whether there is a voice command corresponding to manual input. Specifically, whether or not there is a voice command corresponding to the manual operation stored in the operation history storage means 9 is a voice command correspondence table 11 in which a list of the manual operation and the corresponding voice command is described. (See FIG. 3). If it is determined in step ST9 that there is no voice command corresponding to the manual operation, the sequence proceeds to step ST13.

  On the other hand, if it is determined in step ST9 that there is a voice command corresponding to manual input, the voice command is acquired from the voice command correspondence table 11 (step ST10). That is, the voice command acquisition unit 83 acquires a voice command corresponding to the current state Q2 from the voice command correspondence table 11. Next, a voice guide is generated by the voice guide generation means 12, and a screen telop is generated by the screen telop generation means 13 (step ST11). Next, the voice guide generated by the voice output unit 14 and the screen telop generated by the screen output unit 15 are respectively sent to the speaker 5 and the monitor 4 and presented to the user (step ST12). Thereafter, the operation history is initialized (step ST13). Specifically, the operation history such as the initial state Q1, the current state Q2, and the manual input command C1 is initialized.

  As described above, according to the voice recognition device according to the first embodiment of the present invention, both manual operation and voice operation using a voice command are possible, and a plurality of times are performed in order to achieve the intended purpose. When manual operation is performed, when it is determined that the states sequentially generated by manual operation of the remote controller 2 have reached the objective achievement state by referring to the state management table 10, the voice command corresponding to the objective achievement state Is present in the voice command correspondence table 11, a message (voice guide and screen telop) indicating that a function corresponding to manual operation of the remote controller 2 can be realized by the voice command is generated and output. Accordingly, in the middle of reaching the goal achievement state, even if there is a voice command corresponding to the manual operation, it is not output. Therefore, the user can understand the voice command optimum for achieving the goal, and the series of steps until the goal is achieved. It becomes easy to understand the voice operation method. As a result, it is possible to provide a voice recognition device with excellent operability that can present a voice command suitable for voice operation of the device to the user.

Embodiment 2. FIG.
The speech recognition apparatus according to the second embodiment of the present invention is regarded as having achieved the objective when it has not been operated for a certain period of time while operating the remote controller 2 to achieve the objective. A message is presented to the user.

  The configuration of the speech recognition apparatus according to Embodiment 2 of the present invention is the same as the configuration of the speech recognition apparatus according to Embodiment 1 described above.

  Next, the operation of the speech recognition apparatus according to Embodiment 2 of the present invention will be described with reference to the flowchart shown in FIG. Note that steps in which the same processing as in the speech recognition apparatus according to Embodiment 1 is performed are denoted by the same reference numerals as those used in the flowchart of FIG.

  If it is determined in step ST8 that the objective is achieved, the process proceeds to step ST9 to determine whether or not it is necessary to present a voice command as described above. On the other hand, if it is determined in step ST8 that the objective is not achieved, first, the timer unit 84 is initialized (step ST14). Specifically, the count value T of the timer unit 84 is set to zero. Next, the timer unit 84 starts counting (step ST15). That is, the state confirmation unit 82 passes the current state Q2 to the timer unit 84 and starts counting.

  Next, it is checked whether there is a key input (step ST16). The process in step ST16 is the same as the process in step ST1. If it is determined in step ST16 that there is a key input, it is recognized that the next operation for achieving the purpose has been performed, and the sequence returns to step ST2. On the other hand, if it is determined in step ST16 that there is no key input, it is then checked whether the count value T has reached 5 seconds (sec) or more (step ST17). Here, if it is determined that the count value T is not 5 seconds or more, the sequence returns to step ST16, and the above-described processing is repeated. In the course of this repeated execution, if it is determined in step ST17 that the count value T has reached 5 seconds or more, the current state is recognized as the user's goal achievement state, and the sequence proceeds to step ST9.

  As described above, according to the speech recognition apparatus according to the second embodiment of the present invention, when the remote controller 2 is operated for a certain time during the operation of the remote controller 2 to achieve the object, As a goal achievement state, the voice command corresponding to the manual operation up to that state is presented to the user, so that the user knows the voice command corresponding to the manual operation for any state. Can do.

Embodiment 3 FIG.
When the voice recognition apparatus according to the third embodiment of the present invention has a voice dialogue function, only the necessary voice commands are presented to the user.

  The configuration of the speech recognition apparatus according to Embodiment 3 of the present invention is the same as that of the speech recognition apparatus according to Embodiment 1 described above. However, the control means 8 shall have a voice interaction function. In addition, since the structure and operation | movement for implement | achieving a voice interaction function are known, description is abbreviate | omitted.

  Next, the operation of the speech recognition apparatus according to Embodiment 3 of the present invention will be described. In the following description, a case where the voice recognition apparatus is applied to a car navigation system and a telephone number is searched will be described as an example. The operation of this speech recognition apparatus is the same as the operation described with reference to the flowchart of FIG.

  When searching for a phone number by manual operation, as shown in FIG. 8, press the [Menu] key (step 1) → select [Find a place / facility] (step 2) → select [Set from phone number] (Step 3) → [Phone number] is input (Step 4) → [Search] key is pressed (Step 5). Then, when the [Search] key is pressed in step 5, the pressing of the [Search] key in the telephone number search represents a goal achievement state as shown in FIG. As shown in FIG. 3, the voice command corresponding to the press is “telephone number search”. For example, a message such as “voice can be operated by speaking“ telephone number search ”” is displayed to the user. Presented.

  Therefore, when searching for a telephone number by voice operation next time, as shown in FIG. 9, “telephone number search” and utterance (step 1) → “speak phone number” guidance (step 2) → phone number Speech (Step 3) → Guidance (Step 4) “Speak with search or correction” → Procedure of voice dialogue “Search” and speech (Step 5).

  As described above, since the voice command “search” corresponding to pressing of the [search] key exists in step 5 of the manual operation, the same operation as “pressing the [search] key” can be realized by speaking “search”. In this voice recognition device, it is only necessary to input according to the voice guidance “Speak or search,” so that phone number search can be started without knowing the vocabulary to be spoken (“search”) and the timing of the utterance. .

  As described above, according to the voice recognition device according to the third embodiment of the present invention, when the voice recognition device has the voice recognition function, only the voice command necessary for starting the voice dialogue is presented to the user. Therefore, the user does not need to learn unnecessary voice commands.

Embodiment 4 FIG.
In the speech recognition apparatus according to Embodiment 4 of the present invention, only the recognition dictionary corresponding to the speech command presented to the user is enabled.

  The configuration of the speech recognition apparatus according to Embodiment 4 of the present invention is the same as that of the speech recognition apparatus according to Embodiment 1 described above.

  Next, the operation of the speech recognition apparatus according to Embodiment 4 of the present invention will be described with reference to the flowchart shown in FIG. Note that steps in which the same processing as in the speech recognition apparatus according to Embodiment 1 is performed are denoted by the same reference numerals as those used in the flowchart of FIG. In the initial state, it is assumed that all of the plurality of recognition dictionaries constituting the speech recognition dictionary 17 are in an invalid state.

  In step ST12, the voice guide generated by the voice output unit 14 and the screen telop generated by the screen output unit 15 are respectively sent to the speaker 5 and the monitor 4 and presented to the user. A corresponding recognition dictionary is detected from the recognition dictionary management means 18 (step ST18). Next, the state of the recognition dictionary is acquired (step ST19). Next, it is checked whether or not the recognition dictionary detected in step ST18 is in an invalid state (step ST20). If it is determined in step ST20 that the state is not invalid, that is, the state is valid, the sequence proceeds to step ST13.

  If it is determined in step ST20 that the voice command is invalid, the recognition dictionary status corresponding to the voice command is changed from the invalid status to the valid status so that the voice command presented to the user can be used. (Step ST21). Then, the information of the recognition dictionary changed to the valid state is initialized (step ST22). Specifically, the last use date and time of the dictionary management table 19 is changed to the date when the recognition dictionary is enabled, that is, the time (year / month / day) acquired by the current time acquisition unit 87. Thereafter, the sequence proceeds to step ST13.

  As described above, according to the voice recognition device according to the fourth embodiment of the present invention, in the initial state, all the recognition dictionaries for recognizing the voice command are set in the invalid state and correspond to the voice command presented to the user. Since only the recognition dictionary to be activated is enabled and can be used thereafter, the recognition dictionary in the invalid state cannot be used, and the user can use the voice command with a high recognition rate. In addition, since the voice input command that can be used becomes clear to the user, if the voice command spoken by the user is not recognized, it cannot be determined whether the voice command spoken does not exist or has been misrecognized. The occurrence of the situation can be avoided.

Embodiment 5 FIG.
The speech recognition apparatus according to Embodiment 5 of the present invention is such that, in the speech recognition apparatus according to Embodiment 4, when the activated recognition dictionary is not used for a certain period of time, the speech recognition apparatus is disabled. .

  The configuration of the speech recognition apparatus according to Embodiment 5 of the present invention is the same as that of the speech recognition apparatus according to Embodiment 1 described above.

  Next, the operation of the speech recognition apparatus according to Embodiment 5 of the present invention will be described with reference to the flowchart shown in FIG. Note that steps in which the same processing as that of the speech recognition apparatus according to the fourth embodiment is performed are denoted by the same reference numerals as those used in the flowchart of FIG.

  When the speech recognition apparatus is activated by turning on the power, first, a recognition dictionary that is unused for a long period (for example, for six months) is extracted (step ST23). That is, the dictionary information of each recognition dictionary is acquired from the dictionary management table 19 of the recognition dictionary management means 18, and a recognition dictionary that has not been used for a long time is extracted by referring to the last use date and time included in this dictionary information. Next, the state of the extracted recognition dictionary is changed from the valid state to the invalid state (step ST24). That is, the recognition dictionary invalidation unit 86 of the control unit 8 sends a recognition dictionary state change request that instructs the recognition dictionary management unit 18 to change the extracted recognition dictionary to an invalid state. Thereby, when the recognition dictionary made into the valid state is not used for a certain period, it is made invalid and cannot be used. Thereafter, the sequence proceeds to step ST1.

  As described above, according to the speech recognition apparatus according to the fifth embodiment of the present invention, the recognition dictionary that has been validated is invalidated when it is not used for a certain period of time. The recognition rate for voice commands increases.

It is a block diagram which shows the structure of the speech recognition apparatus which concerns on Embodiment 1 of this invention. It is a figure which shows the memory content of the state management table used with the speech recognition apparatus which concerns on Embodiment 1 of this invention. It is a figure which shows the memory content of the audio | voice command corresponding table used with the speech recognition apparatus which concerns on Embodiment 1 of this invention. It is a figure which shows the memory content of the dictionary management table used with the speech recognition apparatus which concerns on Embodiment 1 of this invention. It is a figure for demonstrating the procedure of the manual operation of the speech recognition apparatus which concerns on Embodiment 1 of this invention. It is a flowchart for demonstrating operation | movement of the speech recognition apparatus which concerns on Embodiment 1 of this invention. It is a flowchart for demonstrating operation | movement of the speech recognition apparatus which concerns on Embodiment 2 of this invention. It is a figure for demonstrating the procedure of the manual operation of the speech recognition apparatus which concerns on Embodiment 3 of this invention. It is a figure for demonstrating the procedure of the voice operation of the speech recognition apparatus which concerns on Embodiment 3 of this invention. It is a flowchart for demonstrating operation | movement of the speech recognition apparatus which concerns on Embodiment 4 of this invention. It is a flowchart for demonstrating operation | movement of the speech recognition apparatus which concerns on Embodiment 5 of this invention.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 Voice recognition processing apparatus, 2 Remote control, 3 Microphone, 4 Monitor, 5 Speaker, 6 Key input means, 7 Voice input means, 8 Control means, 9 Operation history storage means, 10 State management table, 11 Voice command correspondence table, 12 Voice guide generation means, 13 screen telop generation means, 14 voice output means, 15 screen output means, 16 voice recognition means, 17 voice recognition dictionary, 18 recognition dictionary management means, 19 dictionary management table, 81 status management section, 82 status check Part, 83 voice command acquisition part, 84 timer part, 85 recognition dictionary state change part, 86 recognition dictionary invalidation part, 87 current time acquisition part.

Claims (4)

Voice input means for inputting voice;
Voice recognition means for recognizing the voice input from the voice input means as a voice command;
Manual input means for inputting manually,
A state management table storing whether or not the state generated by manual operation of the manual input means is a state where the intended purpose has been achieved;
A voice command correspondence table storing a correspondence relationship between a state generated by manual operation of the manual input means and a voice command input from the voice input means;
Corresponds to the goal achievement state caused by manual operation of the manual input means when it is judged that the goal achievement state has been reached by comparing the state sequentially generated by manual operation of the manual input means with the contents of the state management table Control means for generating a message indicating that the function corresponding to the manual operation of the manual input means can be realized by the voice command if the voice command to be present exists in the voice command correspondence table;
Output means for outputting a message generated by the control means
And
The control means includes
A timer section for measuring a time during which the manual input means is not manually operated,
The manual input means has not been operated by the timer means for a certain period of time before determining that the objective achievement state has been reached by comparing the states sequentially generated by manual operation of the manual input means with the contents of the state management table. If the voice command corresponding to the state generated by the manual operation of the manual input means is present in the voice command correspondence table up to that time, the manual operation of the manual input means is performed according to the voice command. features and to Ruoto voice recognition system to generate a message indicating that can realize a function corresponding to the. Voice input means for inputting voice;
Voice recognition means for recognizing the voice input from the voice input means as a voice command;
Manual input means for inputting manually,
A state management table storing whether or not the state generated by manual operation of the manual input means is a state where the intended purpose has been achieved;
A voice command correspondence table storing a correspondence relationship between a state generated by manual operation of the manual input means and a voice command input from the voice input means;
Corresponds to the goal achievement state caused by manual operation of the manual input means when it is judged that the goal achievement state has been reached by comparing the state sequentially generated by manual operation of the manual input means with the contents of the state management table Control means for generating a message indicating that the function corresponding to the manual operation of the manual input means can be realized by the voice command if the voice command to be present exists in the voice command correspondence table;
Output means for outputting a message generated by the control means
And
The control means has a voice interaction function,
When it is determined that it is now the purpose achieving state sequentially resulting state by manual operation of said manual input means against the contents of the state management table, it is possible to realize a function corresponding to the manual operation of the pre-Symbol manual input means features and to Ruoto voice recognition system to generate a message to be presented only voice commands necessary to initiate the voice conversation. Voice input means for inputting voice;
Voice recognition means for recognizing the voice input from the voice input means as a voice command;
Manual input means for inputting manually,
A state management table storing whether or not the state generated by manual operation of the manual input means is a state where the intended purpose has been achieved;
A voice command correspondence table storing a correspondence relationship between a state generated by manual operation of the manual input means and a voice command input from the voice input means;
Corresponds to the goal achievement state caused by manual operation of the manual input means when it is judged that the goal achievement state has been reached by comparing the state sequentially generated by manual operation of the manual input means with the contents of the state management table Control means for generating a message indicating that the function corresponding to the manual operation of the manual input means can be realized by the voice command if the voice command to be present exists in the voice command correspondence table;
Output means for outputting a message generated by the control means
And
A plurality of recognition dictionaries used for recognizing speech by the speech recognition means,
A recognition dictionary managing means for managing each of the plurality of recognition dictionaries by setting them in a valid state or invalid state,
The control unit is configured to change only the recognition dictionary corresponding to the voice command included in the generated message to the valid state from the plurality of recognition dictionaries that are set to the invalid state in the initial state. features and to Ruoto voice recognition device further comprising a recognition dictionary state changing unit that instructs the. The control means includes a recognition dictionary invalidation unit that instructs the recognition dictionary management means to change the recognition dictionary to an invalid state when the recognition dictionary set to the valid state has not been used for a certain period of time. The speech recognition apparatus according to claim 3 .

Images