close

Вход

Забыли?

вход по аккаунту

JP2005184461

код для вставкиСкачать
Patent Translate
Powered by EPO and Google
Notice
This translation is machine-generated. It cannot be guaranteed that it is intelligible, accurate,
complete, reliable or fit for specific purposes. Critical decisions, such as commercially relevant or
financial decisions, should not be based on machine-translation output.
DESCRIPTION JP2005184461
PROBLEM TO BE SOLVED: To provide a microphone device capable of setting a microphone to
the same position or direction each time to a user and maintaining a good voice recognition rate
and a constant quality of collected voice. SOLUTION: An attachment state of a microphone 2 is
detected by a position sensor 23 (23a, 23b, 23c) for detecting an attachment state of the
microphone based on an operation instruction by a user and a position adjustment motor 21
(21a, 21b, 21c). At the same time, the attachment state is stored in the storage device 33 for
each user, and when the user is selected thereafter, the microphone is driven to be in the stored
attachment state. In addition, when the attachment state of the microphone is changed according
to a predetermined movement pattern and the attachment state in which the sound collection of
the microphone is maximized is stored in the storage device for each user, and then the user is
selected. The microphone is driven to be in the stored attachment state. [Selected figure] Figure 1
Microphone control device and microphone control method
[0001]
The present invention relates to, for example, a microphone control device and a microphone
control method provided in a cabin of a vehicle.
[0002]
Conventionally, a microphone is provided in an instrument panel of a vehicle, etc., and the voice
recognition is performed by collecting the voice uttered by the speaker with the microphone and
03-05-2019
1
comparing the collected voice with a dictionary stored in advance. Is known (Patent Document
1).
Unexamined-Japanese-Patent No. 2003-140687
[0003]
As for the microphone for the vehicle, it is preferable to tune the mounting position and
directivity according to the user, but it is complicated to adjust them finely for each user and ship
the vehicle. It is decided to place, install and choose a position that withstands use. However,
since the volume differs for each user, and the position of the user's mouth differs with respect to
the mounting position and mounting direction of the microphone, the microphone can not
properly pick up the voice of the speaker, and the voice recognition rate decreases. There was a
problem that the quality of the collected sound was degraded.
[0004]
In order to solve the above-mentioned problems, the present invention can set a microphone to
the same position or direction every time for each user, and can maintain a good voice
recognition rate and quality of collected voice, and a microphone control device It aims at
providing a microphone control method.
[0005]
Therefore, the present invention changes the mounting state of the microphone whose mounting
state can be changed based on the operation instruction by the user, stores the mounting state
for each user, and then selects the user When it was, it was made to drive a microphone so that it
might be in the memorized attachment state.
In addition, the attachment state of the microphone whose attachment state can be changed is
changed according to a predetermined movement pattern, and is stored for each user in the
attachment state where the sound collection of the microphone is maximum, and then the user is
selected In this case, the microphone was driven so as to be in the stored attachment state.
03-05-2019
2
[0006]
According to the present invention, the mounting state of the microphone can be reproduced for
each user in an appropriately adjusted position, and a good voice recognition rate or quality of
collected voice can be maintained.
[0007]
Hereinafter, embodiments of the present invention will be described.
FIG. 1 is a block diagram of a speech recognition apparatus to which the present invention is
applied. The voice recognition apparatus comprises a voice recognition control unit (hereinafter
referred to as voice recognition C / U) 1 and a microphone 2, a display 3, a speaker 4 connected
thereto, a selection decision switch 5 as an operation unit, a voice recognition start switch 6a,
voice recognition end switch 6b, user selection switch 7, microphone position adjustment switch
8, registration switch 9, volume indicator 10, and automatic adjustment switch 11. The voice
recognition C / U1 further includes position adjustment motors 21 (21a, 21b, 21c) for driving
the positions of the microphones 2 so as to be changeable, and position sensors 23 (23a, 23b,.
23c) is connected.
[0008]
A detailed description of each configuration is given below. The voice recognition C / U 1
comprises a central processing unit (hereinafter referred to as CPU) 31 and a read only memory
(hereinafter referred to as ROM) 32a connected thereto, a random access memory (hereinafter
referred to as RAM) 32b, and a storage device 33. A / D converter (A / D) 34a that converts an
analog signal from the microphone 2 into a digital signal and inputs it to the CPU 31 converts a
digital audio output signal from the CPU 31 into an analog signal and outputs it to the speaker 4
It comprises a D / A converter (D / A) 34b.
[0009]
The ROM 32 a stores and stores programs operated by the CPU 31 and preset data required for
the programs. The RAM 32 b temporarily stores data being processed by the CPU 31. The
storage device 33 is, for example, a hard disk drive, and includes a dictionary for converting into
03-05-2019
3
phonetic characters corresponding to voice patterns, a dictionary for analogizing words from the
relationship between sentences, and a kanji dictionary. In addition, the storage device 33 stores
the set position of the microphone 2 in an writable / readable manner at an address
corresponding to a numeric key button described later of the user selection switch 7.
[0010]
The selection decision switch 5 is composed of a cursor movement switch and a decision / cancel
switch, and enables the operator to select and decide a candidate phrase of the sentence
converted from voice to character displayed on the display 3 . The microphone position
adjustment switch 8 instructs the CPU 31 to move the microphone 2 in the left and right
direction while pressing the microphone position adjustment switch 8, and instructs the
movement in the front and rear direction of the microphone 2 while pressing it. , 42B, a
compound switch having a knob switch 41 for instructing vertical movement of the microphone
2 while pinching with the fingertip and rotating left or right. The user selection switch 7 is a
switch provided with a plurality of numeric key buttons each having a light emitting portion on
which numbers are drawn.
[0011]
Display 3, speaker 4, selection determination switch 5, voice recognition start switch 6a, voice
recognition end switch 6b, user selection switch 7, microphone position adjustment switch 8,
registration switch 9, automatic adjustment switch 11, volume indicator 10, It is arranged
collectively in the mentor panel.
[0012]
The arrangement of the position adjustment motor 21 and the position sensor 23 is shown in
FIG.
A guide rail 25Y is provided so as to extend in the front-rear direction on a ceiling above and in
front of the front seat of the vehicle cabin, and slidably holds the pedestal 28Y. The pedestal 28Y
is driven in the front-rear direction by a loop-shaped wire 27Y stretched between a drive drum
22b rotationally driven by a position adjustment motor 21b capable of rotating in the forward
and reverse directions and the drum 26Y. A guide rail 25X extending in the left-right direction of
the vehicle body is fixed to the pedestal 28Y. The guide rails 25X slidably hold the pedestals
03-05-2019
4
28X. The pedestal 28X is also driven in the left-right direction by a looped wire 27X stretched
between a drive drum 22a, which is rotationally driven by a position adjustment motor 21a that
can rotate in the forward and reverse directions, and the drum 26X.
[0013]
A position adjustment motor 21c is provided on the pedestal 28X, and a screw hole of a drive nut
22c rotationally driven by the position adjustment motor 21c vertically penetrates a screw shaft
29b. The tube 24 is fixed to the case of the position adjustment motor 21c. A pair of opposing
guide grooves (not shown) are provided in the axial direction on the inner surface of the cylinder
24. A pair of guide protrusions 29a provided on the side surface of the microphone 2 is slidably
fitted in the guide grooves. The guide groove and the guide protrusion 29a hold the microphone
2 so as not to rotate. When the drive nut 22c is driven and rotated by the position adjustment
motor 21c, the microphone 2 moves in the vertical direction together with the screw shaft 29b.
The position adjustment motors 21a, 21b, 21c are provided with position sensors 23a, 23b, 23c,
respectively, to detect the rotation angle. FIG. 3 is a view showing a state in which the guide rails
25X and 25Y are installed on the front ceiling of the vehicle interior. The positional relationship
with the map light 43, the front window 44, the rearview mirror 45, and the sun visor 46 is
shown.
[0014]
FIG. 4 is a view showing the action by the upper and lower positions in the tube 24 of the
microphone 2. The open end side of the tube 24 is the lower side. When the microphone 2 is
stored in the cylinder 24 as shown in (a), the directivity is strong. When the microphone 2 is
exposed from the cylinder 24 as shown in (b), the directivity is weak. . If ambient noise is
relatively large, adjusting the knob switch 41 to the state of (a) can improve the speech
recognition rate of the speaker.
[0015]
The functions of the CPU 31 will be described below. When the speech recognition start switch
6a is turned on, speech recognition processing of the speech input from the microphone 2 is
started. The speech recognition process is ended by turning on the speech recognition end switch
6b. The sound collected by the microphone 2 is converted into a digital signal by the A / D
03-05-2019
5
converter 34 a and input to the CPU 31. The CPU 31 converts the speech into speech patterns
for each phrase, and the dictionary stored in the storage device 33 has the highest certainty to
the speech patterns, selects an appropriate character string from the front and back relation, and
displays it on the display 3.
[0016]
When the operator presses one of the numeric key buttons of the user selection switch 7, the
CPU 31 lights up the light emitting section to indicate that the numeric key button is in the
selected state, and the illuminated numeric key button is displayed again. When it is pressed, it
goes out and becomes non-selected. When one numeric key button is on and the unlit numeric
key button is pressed, the previously illuminated numeric key button is extinguished, and the
newly pressed numeric key button is on. It will be in the state.
[0017]
When the operator operates the microphone position adjustment switch 8 to adjust the
microphone position for his convenience, when the registration switch 9 is pressed, the operator
corresponds to the numeric key button of the user selection switch 7 selected by the operator.
The position signal from the position sensor 23 is readably stored in the memory address of the
storage device 33. When the microphone position is stored at the address corresponding to the
selected numeral key button, the light emitting unit of the numeral key button blinks for a while,
and it is operated that the registration of the microphone position is performed for the numeral
key button Inform the person and then turn on to indicate the normal selection state.
[0018]
The CPU 31 dynamically displays the volume level collected by the microphone 2 on the volume
indicator 10 with a bar length. The level position of the appropriate target volume is also
displayed on the volume indicator 10. When the position adjustment switch 8 is operated, the
CPU 31 controls the position of the microphone 2 in the front-rear direction, the left-right
direction, and the up-down direction. Also, when the operator selects the numeric key button of
the user selection switch 7, the registered microphone position is read from the corresponding
address of the storage device 33, and a signal for driving the position adjustment motor 21 is
output, and the registered microphone Move the microphone 2 to the position. Furthermore,
03-05-2019
6
when the automatic adjustment switch 11 is turned on, the microphone position is moved
according to the position sequence in a predetermined movement pattern, and the microphone
position is searched for the volume level collected from the microphone 2 to be equal to or
higher than the target level. The position is stored in the address corresponding to the selected
numeric key button of the user selection switch 7.
[0019]
Next, based on FIG. 5, the flow of control for moving the microphone position by operating the
microphone position adjustment switch and registering the position will be described. The voice
recognition apparatus is activated when an ignition switch (not shown) is turned on. The
operator sets the seat to a suitable state for himself, and operates the respective switches of the
microphone position adjustment switch 8 so that the user's voice can be properly picked up by
the microphone 2, while continuing to emit audio at a constant volume. . The flow of the
following processing is processed as a program of the CPU 31.
[0020]
In step 101, it is checked whether the microphone position adjustment switch 8 has been
operated. If it is not operated, step 101 is repeated. If it is operated, the process proceeds to step
102. In step 102, while any one of the composite switches having the left and right direction
switches 42R and 42R, the front and rear direction switches 42F and 42B, and the knob switch
41 is operated, the position corresponding to the moving direction of the microphone position
indicated by the switch A command signal for rotating any of the adjustment motors 21a, 21b,
21c in the instructed direction is output. The position signals from the position sensors 23a, 23b,
and 23c are input to the CPU 31, and when it is detected that the movable range limit has been
reached, any one of the microphone position adjustment switches 8 is operated. , Command
signal is not output.
[0021]
In step 103, position adjustment motors 21a, 21b and 21c corresponding to the operated switch
of the microphone position adjustment switch 8 drive the microphone 2 in the operated direction
based on the command signal. In step 104, the volume indicator 10 displays the target volume
level and the volume level currently picked up by the microphone 2. The target volume level is
03-05-2019
7
set by reading out data preset in the ROM 32a. In step 105, it is checked whether the registration
switch 9 has been operated. If the registration switch 9 is operated, the process proceeds to the
switch 106. If this is not the case, the operator continues to operate the microphone position
adjustment switch 8 so as to reach the standard volume level while watching the volume
indicator 10, returns to step 101, and repeats step 105.
[0022]
The operator looks at the volume indicator 10, finds the proper position of the microphone 2
above the target volume level, and operates the user selection switch 7 to store the microphone
position for himself so as to reproduce it at any time in the future . In step 106, it is checked
whether any numeric key button is on. The number key button turned on lights up.
[0023]
If it is on, the process proceeds to step 107, where the set position of the user selection switch 7
is read out. If not, step 106 is repeated. At step 108, the current position signal of each position
sensor 23 is read out and stored in the address of the storage device 33 corresponding to the
numeric key button which the user selection switch 7 is turned on. At step 109, the user turns on
the numeric key button of the user selection switch 7 for a while to inform the operator of the
completion of the registration, and then this series of processing ends.
[0024]
Next, based on FIG. 6, FIG. 7, the flow of control which moves a microphone position
automatically using an automatic adjustment switch, searches for a suitable position to a target
sound volume level, and registers automatically is demonstrated. The flow of the following
processing is processed as a program of the CPU 31. The operator sets the seat to a state suitable
for him. In step 201, it is checked whether the automatic adjustment switch 11 is on. If it is on,
the process proceeds to step 202. If it is not on, step 201 is repeated. In step 202, it is checked
whether any one of the numeric key buttons of the user selection switch 7 is on. If it is on, the
process proceeds to step 203. If it is not on, step 202 is repeated.
[0025]
03-05-2019
8
At step 203, a target volume level to be displayed on the volume indicator 10 is set. This reads
out and sets data preset in the ROM 32a. In step 204, a command signal is output to each
position adjustment motor 21 to set the microphone at each target position of the position
sequence set in the movement pattern of the predetermined microphone position stored in the
ROM 32a.
[0026]
The movement pattern is, for example, the following steps. The initial position of the microphone
2 is the front end of the movable range for the lateral position, the front end of the movable
range for the longitudinal position, and the main body of the microphone 2 in the cylinder 24 for
the vertical position. The position is almost stored. As the movement order, for example, first, the
position adjustment motor 21a is driven to move the microphone position from the assistant
driver's seat side to the driver's seat side along the guide rail 25X. Next, the position adjustment
motor 21b is driven to move the microphone position backward. Finally, the position adjustment
motor 21c is driven to move the microphone position downward. In this series of movement
patterns of the microphone 2, the search for the maximum sound volume position is performed
in accordance with steps 204 to 213.
[0027]
In step 205, the position adjustment motor 21 is driven according to the control of the CPU 31.
The operator waits for the microphone position to be set automatically while emitting sound at a
constant volume level. In step 206, the volume indicator 10 displays the volume level picked up
by the microphone 2. At step 207, the average volume of a predetermined time at each target
position of the pattern is calculated. At step 208, the signal from each position sensor 23 at that
timing is read. In step 209, the automatic adjustment starts to check whether it is the first target
position of the movement pattern. In the first case, the process proceeds to step 211, and in the
case of the second and subsequent target positions, the process proceeds to step 210.
[0028]
In step 210, the volume level of the current target position of the movement pattern is compared
03-05-2019
9
with the stored volume level at the previous target position of the movement pattern, and the
volume level at the current target position of the movement pattern is larger. Check if it is. If the
present is smaller, the process proceeds to step 213, and if the present is larger, the process
proceeds to step 211. At step 211, the current microphone position of the movement pattern and
its volume level are stored in the RAM 32b.
[0029]
In step 212, it is checked whether the target volume level has been exceeded. If it does not
exceed, it proceeds to step 213, and if it exceeds it proceeds to step 215. In step 213, it is
checked whether the predetermined movement pattern position sequence has ended. If not
completed, the process returns to step 204 to continue moving the microphone position. If the
position sequence is finished, the process proceeds to step 214. At step 214, the maximum sound
volume position stored in the RAM 32b is read out, and a command is output to the position
adjustment motor 21 to move the microphone 2 to that position, and the microphone 2 is set.
After step 214, the process proceeds to step 216.
[0030]
If the process proceeds to step 215 after step 212, the command output to each position
adjustment motor 21 is stopped, and the process proceeds to step 216. In step 216, the
microphone position stored in the RAM 32b is stored in the address of the storage device 33
corresponding to the numeric key button of the user selection switch 7 in the on state. Finally, in
step 217, the numeric key button of the user selection switch 7 is blinked for a while to notify
the operator of the registration completion, and then the state returns to the on state indicating
the on state. Thereafter, the registration switch 9 is automatically turned off, and this series of
processing ends.
[0031]
Next, based on FIG. 8, the flow of control for selecting the stored microphone position and setting
the microphone 2 to a suitable position for the operator will be described. The flow of the
following processing is processed as a program of the CPU 31. The operator sets the seat to a
state suitable for him. In step 301, it is checked whether the numeric key button of the user
selection switch 7 is on. If any numeric key button is turned on, the process proceeds to step
03-05-2019
10
302. If not, step 301 is repeated.
[0032]
In step 302, the target position of the microphone position is read out from the address of the
storage device 33 corresponding to the number button of the user selection switch 7 being
turned on. At step 303, the position signal of the current microphone position indicated by each
position sensor 23 is read out. Then, in step 304, the current position and the target position are
compared. In step 305, the movement direction can be determined from the comparison result,
and a command signal to each position adjustment motor 21 is output. At step 306, the position
adjustment motor is driven in accordance with the command signal. In step 307, it is checked
whether the target position and the current position match. If they do not match, the process
returns to step 304 and repeats the steps 304 to 307 until they match. If they match, the process
proceeds to step 308.
[0033]
In step 308, it is checked whether the speech recognition start switch 6a is turned on. If it is not
turned on, step 308 is repeated. If it is turned on, the process proceeds to step 309 to start
speech recognition processing. In step 310, it is checked whether the speech recognition end
switch 6b is turned on. If it is turned on, the process ends. If not, the speech recognition process
is continued.
[0034]
The position adjustment motors 21a, 21b and 21c of the present embodiment are drive means of
the present invention, the position sensors 23a, 23b and 23c are attachment state detection
means, and the user selection switch 7 is user selection means. Constitute a storage means. Steps
101 and 102 of the flowchart are attachment state change output means of the present
invention, steps 105, 107 and 108, or step 216 are storage instruction means, step 201 is
automatic adjustment start means, and steps 302 to 307 are drive instructions. Configure the
means.
[0035]
03-05-2019
11
As described above, according to the present embodiment, once the operator finds the
appropriate microphone position by manual microphone position adjustment using the
microphone position adjustment switch 8 or the microphone position adjustment with the
automatic adjustment switch 11, registration is performed. If this is done, the user can set the
microphone position to an appropriate microphone position by operating the user selection
switch 7, and the speech recognition accuracy can be maintained constant. When one vehicle is
used by a plurality of people, it is convenient because the setting of the microphone position
suitable for each person can be performed by a simple selection operation.
[0036]
Next, a modification of the flow of control of the automatic adjustment of the microphone
position according to the present embodiment will be described based on FIG. In this
modification, since steps 201 to 208 in FIG. 6 are the same except that the predetermined
movement pattern in step 204 is different, the description will be omitted. The movement pattern
in the present modification sets, for example, a reference position at substantially the center of
the movable range in the left and right direction and the front and back direction, and moves a
plurality of ellipses of different diameters surrounding the reference position or a spiral
movement extending out from the reference position. It is a pattern. First, the microphone
position is moved up and down at the reference position to store the rate of change in volume
level due to the change in directivity, and then the microphone 2 is moved in the above
movement pattern in a state of being stored in the tube 24. After step 208, the process proceeds
to step 401. At step 401, the current target position of the movement pattern and the volume
level are stored as a set in the RAM 32b.
[0037]
In step 402, it is checked whether the predetermined movement pattern position sequence has
ended. If not completed, the process returns to step 204 to continue moving the microphone
position, and steps 204 to 208 and steps 401 to 402 are repeated. If the position sequence is
completed, the process proceeds to step 403. At step 403, the position of the position sequence
stored in the RAM 32b and the volume level corresponding to the position are read out to
calculate the sound field.
03-05-2019
12
[0038]
In step 404, a target microphone position to be a target volume is estimated from the calculated
sound field. The sound field is determined for the movement pattern of a reference position and a
plurality of ellipses surrounding it, or the movement pattern of a spiral spreading out from the
reference position. Therefore, it is possible to estimate, for example, by spline interpolation the
volume level with respect to the position coordinates, at which position away from the reference
position the maximum point of the volume level is located. If the estimated volume at that
position does not reach the target volume level, the correction position is calculated by lowering
the microphone 2 toward the open end of the tube 24. The position is stored in the RAM 32b.
[0039]
In step 405, the target microphone position and the current microphone position from the
position sensor 23 are compared. In step 406, a command signal is output to each position
adjustment motor 21. At step 407, the position adjustment motor 21 is driven in accordance with
the command signal. In step 408, it is checked whether the position signal from the position
sensor 23 matches the target position. If they do not match, the process returns to step 405, and
steps 405 to 408 are repeated. If they match, the process proceeds to step 409. In step 409, the
microphone position stored in the RAM 32b is stored in the address of the storage device 33
corresponding to the numeric key button of the user selection switch 7 in the on state. Finally, in
step 410, the numeric key button of the user selection switch 7 is blinked for a while to notify
the operator of the registration completion, and then the state returns to the on state indicating
the on state. Thereafter, the registration switch 9 is automatically turned off, and this series of
processing ends. Step 409 of this modification constitutes the storage instruction means of the
present invention.
[0040]
According to this modification, the movement pattern in the case of automatic adjustment is not
a sequential search for the maximum sound volume, but is a method in which the sound field is
determined from the center movement of the reference or the movement pattern of the spiral
and then estimated The time required for a person to utter a constant volume can be short.
[0041]
03-05-2019
13
In the embodiment or the modification, the microphone 2 is installed on the front ceiling of the
passenger compartment, and the microphone 2 can be moved in the front-rear direction, the leftright direction, and the up-down direction. Absent.
For example, as shown in FIG. 10, it is installed on the instrument panel in front of the driver's
seat, and the microphone position is positioned forward and backward by the position
adjustment motor 21b, and in the cylinder 24 by the position adjustment motor 21c. The change
may be made by the adjustment motor 21a '. In the drawing, the position sensor 23 provided for
each position adjustment motor is not shown. Furthermore, although the configuration has been
described in which the microphone position is moved by the wire drive by the motor and the
drive by the drive nut and screw shaft by the motor, it is not limited thereto. In the embodiment
described above, the microphone is provided in the vehicle compartment. However, for example,
the microphone is used by a customer support center in a company or a microphone used by an
operator of a telephone exchange, or a podium used in a lecture or the like. The same effect can
be obtained even when applied to a microphone or the like.
[0042]
It is a block diagram showing composition of an embodiment to which the present invention is
applied. It is a figure explaining arrangement of a position adjustment motor. It is a figure
explaining the position which attaches a microphone to a vehicle interior. It is a figure explaining
the change of the directivity by the position of a microphone. It is a figure explaining the flow of
control which adjusts a microphone position manually and registers. It is a figure explaining the
flow of control which adjusts a microphone position automatically and registers. It is a figure
explaining the flow of control which adjusts a microphone position automatically and registers. It
is a figure explaining the flow of control which adjusts a microphone position to the position
which selected the user selection switch by operating. It is a figure explaining the flow of control
which adjusts and registers a microphone position automatically in a modification. It is a figure
explaining the case where a microphone is installed on an instrument panel.
Explanation of sign
[0043]
1 voice recognition control unit 2 microphone 3 display 4 speaker 5 selection decision switch 6a
voice recognition start switch 6b voice recognition end switch 7 user selection switch 8
microphone position adjustment switch 9 registration switch 10 volume indicator 11 automatic
03-05-2019
14
adjustment switch 21a, 21a ' , 21b, 21c Position adjustment motor 22a, 22b, Drive drum 22c
Drive nut 23a, 23b, 23c Position sensor 24 Tube 25X, 25Y Guide rail 26X, 26Y Drum 27X, 27Y
Wire 28X, 28Y Base 29a Guide projection 29b Screw shaft 31 center Arithmetic processing unit
32a read only memory 32b random access memory 33 storage unit 34a A / D converter 34b D /
A converter 41 knob switch 42L left direction switch 42R right direction switch 42F front
direction switch 42B rear direction switch 43 map light 44 front window 45 rearview mirror 46
sun visor
03-05-2019
15
1/--страниц
Пожаловаться на содержимое документа