U.S. patent number 9,271,089 [Application Number 13/924,071] was granted by the patent office on 2016-02-23 for voice control device and voice control method.
This patent grant is currently assigned to FUJITSU LIMITED. The grantee listed for this patent is FUJITSU LIMITED. Invention is credited to Chisato Ishikawa, Takeshi Otani, Masanao Suzuki, Taro Togawa.
United States Patent |
9,271,089 |
Suzuki , et al. |
February 23, 2016 |
Voice control device and voice control method
Abstract
A voice control device includes a hearing estimate section
configured to estimate hearing of a user based on a
sending/received sound ratio representing a ratio of the volume of
a sending sound to the volume of a received sound; a
compensation-quantity calculating section configured to calculate a
compensation quantity for a received signal of the received sound
responsive to the estimated hearing; and a compensation section
configured to compensate the received signal based on the
calculated compensation quantity.
Inventors: |
Suzuki; Masanao (Yokohama,
JP), Otani; Takeshi (Kawasaki, JP), Togawa;
Taro (Kawasaki, JP), Ishikawa; Chisato (Kawasaki,
JP) |
Applicant: |
Name |
City |
State |
Country |
Type |
FUJITSU LIMITED |
Kawasaki-shi, Kanagawa |
N/A |
JP |
|
|
Assignee: |
FUJITSU LIMITED (Kawasaki,
JP)
|
Family
ID: |
46457338 |
Appl.
No.: |
13/924,071 |
Filed: |
June 21, 2013 |
Prior Publication Data
|
|
|
|
Document
Identifier |
Publication Date |
|
US 20130279709 A1 |
Oct 24, 2013 |
|
Related U.S. Patent Documents
|
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
Issue Date |
|
|
PCT/JP2011/050017 |
Jan 4, 2011 |
|
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
A61B
5/123 (20130101); H04R 25/30 (20130101); G10L
21/0364 (20130101); A61B 5/749 (20130101); H04R
25/70 (20130101); G10L 2021/0575 (20130101) |
Current International
Class: |
H03G
3/20 (20060101); A61B 5/00 (20060101); H04R
25/00 (20060101); G10L 21/0364 (20130101); A61B
5/12 (20060101); G10L 21/057 (20130101) |
Field of
Search: |
;381/57,47
;704/226,246 |
References Cited
[Referenced By]
U.S. Patent Documents
Foreign Patent Documents
|
|
|
|
|
|
|
06-217398 |
|
Aug 1994 |
|
JP |
|
07-066767 |
|
Mar 1995 |
|
JP |
|
08-163121 |
|
Jun 1996 |
|
JP |
|
08-223256 |
|
Aug 1996 |
|
JP |
|
2000-209698 |
|
Jul 2000 |
|
JP |
|
2004-165865 |
|
Jun 2004 |
|
JP |
|
2004-235708 |
|
Aug 2004 |
|
JP |
|
2009-171189 |
|
Jul 2009 |
|
JP |
|
2010-28515 |
|
Feb 2010 |
|
JP |
|
2010-081523 |
|
Apr 2010 |
|
JP |
|
2010/035308 |
|
Apr 2010 |
|
WO |
|
Other References
International Search Report, mailed in connection with
PCT/JP2011/050017 and mailed Feb. 8, 2011. cited by applicant .
Extended European Search Report dated Nov. 18, 2015 for
corresponding European Patent Application No. 11855034.2, 8 pages.
cited by applicant.
|
Primary Examiner: Jamal; Alexander
Attorney, Agent or Firm: Fujitsu Patent Center
Parent Case Text
CROSS-REFERENCE TO RELATED APPLICATIONS
This application is a continuation application of International
Application PCT/JP2011/050017 filed on Jan. 4, 2011 and designated
the U.S., the entire contents of which are incorporated herein by
reference.
Claims
What is claimed is:
1. A voice control device comprising: a hearing estimate section
configured to estimate hearing of a user based on a
sending/received sound ratio representing a ratio of the volume of
sending sound to the volume of received sound, the sending sound
being sent by the user in a form of a sending signal, and the
received sound being received by the user in a form of a received
signal; a compensation-quantity calculating section configured to
calculate a compensation quantity for the received signal of the
received sound responsive to the estimated hearing; and a
compensation section configured to compensate the received signal
based on the calculated compensation quantity.
2. The voice control device as claimed in claim 1, wherein the
hearing estimate section estimates an age of a user from the
sending/received sound ratio for estimating a minimum audible range
based on the estimated age, wherein the compensation-quantity
calculating section obtains the compensation quantity for the
received signal so that the received signal is compensated to be
greater than the estimated minimum audible range.
3. The voice control device as claimed in claim 2, further
comprising: a noise estimate section configured to estimate a noise
quantity from the sending sound; and a hearing compensation section
configured to compensate the minimum audible range based on the
estimated noise quantity, wherein the compensation-quantity
calculating section obtains the compensation quantity for the
received signal so that the received signal is compensated to be
greater than the estimated minimum audible range.
4. The voice control device as claimed in claim 1, wherein the
hearing estimate section determines whether the received sound and
the sending sound include sound, then obtains the sending/received
sound ratio for the received sound and the sending sound if the
received sound and the sending sound include the sound.
5. The voice control device as claimed in claim 3, wherein the
noise estimate section determines whether the received sound
includes no sound, then updates the noise quantity if the received
sound includes no sound.
6. A method for controlling voice in a voice control device,
comprising: estimating hearing of a user based on a
sending/received sound ratio representing a ratio of the volume of
sending sound to the volume of received sound, the sending sound
being sent by the user in a form of a sending signal, and the
received sound being received by the user in a form of a received
signal; calculating a compensation quantity for the received signal
of the received sound responsive to the estimated hearing; and
compensating the received signal based on the calculated
compensation quantity.
7. A non-transitory computer-readable recording medium having a
program stored therein for causing a computer to execute a method
for controlling voice, the method comprising: estimating hearing of
a user based on a sending/received sound ratio representing a ratio
of the volume of sending sound to the volume of received sound, the
sending sound being sent by the user in a form of a sending signal,
and the received sound being received by the user in a form of a
received signal; calculating a compensation quantity for the
received signal of the received sound responsive to the estimated
hearing; and compensating the received signal based on the
calculated compensation quantity.
Description
FIELD
The invention relates to a voice control device, a voice control
method, a voice control program and a portable terminal device that
control received sound.
BACKGROUND
Conventionally, there are portable terminal devices that execute
control to make received voices easy to hear. For example, there is
a technology with which multiple single-tone frequency signals are
reproduced for a user to calculate the minimum hearing level based
on the user's hearing result for processing a voice (Patent
document 1).
Also, there is a technology in which a Lombard effect is utilized
so that if a sending sound volume is loud, it is determined that it
is noisy in the surroundings to increase the received sound volume,
if a sending sound volume is soft, the received sound volume is
decreased automatically (Patent document 2).
Also, there is a technology in which an equalizer is provided for
emphasizing a voice signal in a specific range, and characteristics
of the equalizer are adjusted based on a volume operation by a user
(Patent document 3).
RELATED-ART DOCUMENTS
Patent Document
[Patent document 1] Japanese Laid-open Patent Publication No.
07-66767 [Patent document 2] Japanese Laid-open Patent Publication
No. 2004-165865 [Patent document 3] Japanese Laid-open Patent
Publication No. 2010-81523
However, there is a problem with Patent document 1 in that it is
not easy to use because a user needs to execute a hearing test that
forces a complicated procedure upon the user.
Also, there is a problem with Patent document 2 in that there are
cases where sound quality may be bad depending on a user because
the received sound volume is determined only by the sending sound
volume, hence a hearing characteristic of the user is not taken
into account.
Also, there is a problem with Patent document 3 in that voice
control may not be carried out during a call because it requires a
user to do a volume operation, which is, however, difficult to do
during a call.
SUMMARY
According to one embodiment, a voice control device includes a
hearing estimate section configured to estimate hearing of a user
based on a sending/received sound ratio representing a ratio of the
volume of a sending sound to the volume of a received sound; a
compensation-quantity calculating section configured to calculate a
compensation quantity for a received signal of the received sound
responsive to the estimated hearing; and a compensation section
configured to compensate the received signal based on the
calculated compensation quantity.
The object and advantages of the embodiment will be realized and
attained by means of the elements and combinations particularly
pointed out in the claims. It is to be understood that both the
foregoing general description and the following detailed
description are exemplary and explanatory and are not restrictive
of the invention, as claimed.
BRIEF DESCRIPTION OF DRAWINGS
FIG. 1 is a schematic view illustrating a change of hearing level
with aging;
FIG. 2 is a block diagram illustrating an example of functions of a
voice control device according to an embodiment;
FIG. 3 is a block diagram illustrating an example of a hearing
estimate section;
FIG. 4 is a schematic view illustrating an example of a
relationship between sending/received sound ratio and age;
FIG. 5 is a schematic view illustrating an example of a
relationship between age and minimum audible range;
FIG. 6 is a block diagram illustrating an example of a
configuration of a hearing compensation section;
FIG. 7 is a schematic view illustrating an example of a
relationship between noise quantity and compensation quantity;
FIG. 8 is a schematic view illustrating an example of minimum
audible ranges before and after compensation;
FIG. 9 is a schematic view illustrating an example of a
relationship between noise quantity and minimum audible ranges
before and after compensation;
FIG. 10 is a schematic view illustrating spectrum compensation;
FIG. 11 is a flowchart illustrating an example of a voice control
procedure according to the embodiment;
FIG. 12 is a flowchart illustrating an example of a hearing
compensation procedure; and
FIG. 13 is a block diagram illustrating an example of hardware of a
portable terminal device according to an embodiment.
DESCRIPTION OF EMBODIMENTS
First, a relationship between age and hearing will be described.
Hearing is, for example, a minimum audible range. FIG. 1 is a
schematic view illustrating a change of hearing level with aging.
According to an experiment illustrated in FIG. 1, it can be
understood that there is a difference in the average hearing level
between age 40 to 60. The hearing level of 60's notably drops in a
high range (2 kHz to 4 kHz). Also, as for hearing impairment
induced by aging, an investigation result was reported in
Http://triosaka.jp/group/infoele/life/sense/data/katagiri/070622.pdf.
As illustrated in FIG. 1, it is already known that hearing level
drops with aging. Especially, in a high frequency range, hearing
level drops when one gets older.
Here, the Lombard effect will be described. The Lombard effect is
an effect that if it is noisy in the surroundings, or it is
difficult to hear a voice of a counterpart because the counterpart
is a quiet talker, one's speaking voice becomes louder. For
example, it has been investigated that if background noise is 50
dBSPL (simply denoted as dB hereafter), the speaking volume becomes
4 dB greater than in a quiet situation (37 dB). As for this
investigation, refer to FIG. 1 in "Effects of noise on speech
production: acoustic and perceptual analyses", W. Van. Summers et.
al., J. Acoust. Soc. Am., Vol. 84, No. 3, September 1988.
However, the Lombard effect affects not only volumes of the
surrounding noise and the voice of a counterpart, but also the
hearing of a listener. If hearing is reduced, it becomes harder to
hear a voice of a counterpart, hence the speaking voice tends to
become louder. As illustrated in FIG. 1, hearing reduction is
related to age, hence there is a relationship between the Lombard
effect and age.
Therefore, in the following, embodiments will be described in which
The Lombard effect is used to obtain a relationship between a
received sound volume and a sending sound volume, in which age is
estimated to estimate hearing for controlling the received sound to
make the received voice easier to hear. The embodiments will be
described below with reference to the drawings.
Embodiment
<Configuration>
A configuration of a voice control device 1 will be described
according to the embodiment. FIG. 2 is a block diagram illustrating
an example of a configuration of the voice control device 1
according to the embodiment. As illustrated in FIG. 2, the voice
control device 1 includes frequency transform sections 101 and 102,
a hearing estimate section 103, a noise estimate section 104, a
hearing compensation section 105, a spectrum compensation-quantity
calculating section 106, a spectrum compensation section 107, and
an inverse frequency transform section 108.
The time-frequency transform section 101 applies a time-frequency
transform to a received signal r(t) of a received sound to obtain a
spectrum R(f) according to the following formula (1). The
time-frequency transform is, for example, a fast Fourier transform
(FFT). R(f)=Re{R(f)}+jIm{R(f)} FORMULA 1
f: frequency (f=0, 1, 2, . . . , K-1)
K: Nyquist frequency
Re { }: real part
Im { }: imaginary part
The frequency transform section 101 outputs the obtained spectrum
R(f) to the hearing estimate section 103, the spectrum
compensation-quantity calculating section 106, and the spectrum
compensation section 107.
The time-frequency transform section 102 applies a time-frequency
transform to a sending signal s(t) of a sending sound to obtain a
spectrum S(f) according to the following formula (2). The
time-frequency transform is, for example, a fast Fourier transform
(FFT). S(f)=Re{S(f)}+jIm{S(f)} FORMULA 2
f: frequency (f=0, 1, 2, . . . , K-1)
K: Nyquist frequency
Re { }: real part
Im { }: imaginary part
The frequency transform section 102 outputs the obtained spectrum
S(f) to the hearing estimate section 103 and the noise estimate
section 104.
The hearing estimate section 103 estimates a user's hearing based
on the received sound volume and the sending sound volume. FIG. 3
is a block diagram illustrating an example of the hearing estimate
section 103. In an example illustrated in FIG. 3, the hearing
estimate section 103 includes a sending/received-sound-ratio
calculating section 131, an age estimate section 132, and a minimum
audible range estimate section 133.
The sending/received-sound-ratio calculating section 131 calculates
the average electric power of the spectrum R(f) of the received
sound and the spectrum S(f) of the sending sound by the following
formula.
.times..times..times..function..times..times..times..times..function..tim-
es..function..times..times..times..times..times..function..times..times..t-
imes..times..function..times..function..times..times.
##EQU00001##
R_ave: average electric power of the spectrum of the received
sound
S_ave: average electric power of the spectrum of the sending
sound
The sending/received-sound-ratio calculating section 131 obtains a
sending/received sound ratio sp_ratio, for example, from the
average power R_ave of the received sound and the average power
S_ave of the sending sound by the following formula.
sp_ratio=S_ave/R_ave FORMULA (5)
sp_ratio: sending/received sound ratio
The sending/received-sound-ratio calculating section 131 sets the
sending/received sound ratio to the ratio of the volume of the
received sound to the volume of the sending sound. The
sending/received-sound-ratio calculating section 131 outputs the
obtained sending/received sound ratio to the age estimate section
132.
Having obtained the sending/received sound ratio from the
sending/received-sound-ratio calculating section 131, the age
estimate section 132 estimates the age of a user with referring to
information that indicates a relationship between the
sending/received sound ratio and age, which is stored
beforehand.
FIG. 4 is a schematic view illustrating an example of a
relationship between a sending/received sound ratio and age. An
experiment has been executed based on a notion that when one gets
older, hearing is impaired, which makes it difficult to hear a
voice of a counterpart, hence a speaking voice becomes louder. By
the experiment similar to that for obtaining the Lombard effect, a
relationship between age and the sending/received sound ratio has
been estimated beforehand.
For example, a relationship between age and the sending/received
sound ratio can be estimated with the following steps:
(1) For each age (or by generation such as teenagers, twenties),
estimate sending sound volume with respect to received sound volume
(for example, 60 dB) with an examinee.
(2) Obtain the average sending sound volume of all examinees of the
ages measured in (1).
(3) Obtain a ratio of the average sending sound volume in (2) to
the received sound volume (sending/received sound ratio).
(4) Execute the steps (1) to (3) for other values of the received
sound volume (for example, 30 to 80 dB).
Thus, the information that indicates a relationship between age and
the sending/received sound ratio can be obtained for the values of
the received sound volume. The age estimate section 132 holds the
information that indicates a relationship between age and
sending/received sound ratio.
Based on the sending/received sound ratio obtained from the
sending/received-sound-ratio calculating section 131, the age
estimate section 132 estimates an age from the relationship
illustrated in FIG. 4. For example, the age estimate section 132
identified the information indicating the relationship illustrated
in FIG. 4 that corresponds to the received sound volume whose
sending/received sound ratio has been obtained. The age estimate
section 132 estimates an age from the information indicating the
identified relationship and the calculated sending/received sound
ratio. The age estimate section 132 outputs the estimated age to
the minimum audible range estimate section 133.
Based on the age obtained from the age estimate section 132, the
minimum audible range estimate section 133 estimates a minimum
audible range. The minimum audible range estimate section 133 holds
a minimum audible range for each generation, based on the
relationship illustrated in FIG. 1.
FIG. 5 is a schematic view illustrating an example of a
relationship between age and a minimum audible range. A1
illustrated in FIG. 5 represents a minimum audible range of 60's,
A2 represents a minimum audible range of 40's, and A3 represents a
minimum audible range of 20's. Here, although the relationship
between generation and a minimum audible range are provided with
intervals of 20 years in the example illustrated in FIG. 5, it may
be provided with intervals of 10 years.
As illustrated in FIG. 5, when the generation goes up higher, the
minimum audible range becomes higher, which means it is harder to
hear. The minimum audible range estimate section 133 holds data
that indicates the relationship between generation and a minimum
audible range as illustrated in FIG. 5 to obtain a minimum audible
range corresponding to the age obtained from the estimate section
132. The minimum audible range estimate section 133 outputs the
obtained minimum audible range to the hearing compensation section
105.
Other than the relationship between generation and a minimum
audible range, a hearing reduction quantity for each generation may
be used. Also, a minimum audible range or a hearing reduction
quantity based on gender may be used. For a difference of hearing
characteristics by gender, see p. 72-73 of "Building Environment
for Aged People", edited by Architectural Institute of Japan, 1994
Jan. 10, Shokokusha Publishing Co., Ltd.
Returning to FIG. 2, the noise estimate section 104 estimates noise
of the surroundings from the sending sound. For example, the noise
estimate section 104 estimates noise from the average power S_ave
of the sending sound of a current frame.
The noise estimate section 104 compares the average power S_ave of
the sending sound and a threshold value TH. If S_ave.gtoreq.TH, the
noise estimate section 104 does not update the noise quantity. If
S_ave<TH, the noise estimate section 104 updates the noise
quantity by the following formula.
noise_level(f)=.alpha..times.S(f)+(1-.alpha.).times.noise_level(f)
FORMULA (6)
noise_level(f): noise quantity
.alpha.: constant
Here, an initial value of noise_level(f) is arbitrary. For example,
the initial value may be set to 0. Also, .alpha. is a constant
between 0 and 1. .alpha. is set to, for example, 0.1.
The threshold value TH may be set from 40 to 50 dB. The threshold
value TH is set smaller than a volume of a human voice because the
volume of voices in people's conversation is 70 to 80 dB. The noise
estimate section 104 outputs the estimated noise quantity to the
hearing compensation section 105.
The hearing compensation section 105 compensates hearing (for
example, a minimum audible range) based on the minimum audible
range obtained from the hearing estimate section 103 and the noise
quantity obtained from the noise estimate section 104. FIG. 6 is a
block diagram illustrating an example of a configuration of the
hearing compensation section 105. The hearing compensation section
105 includes a compensation-quantity calculating section 151 and a
minimum audible range compensating section 152.
The compensation-quantity calculating section 151 calculates a
compensation quantity in response to the noise quantity obtained
from the noise estimate section 104. The compensation-quantity
calculating section 151 outputs the noise quantity to the minimum
audible range compensating section 152.
The minimum audible range compensating section 152 compensates the
minimum audible range based on the minimum audible range obtained
from the hearing estimate section 103 and the compensation quantity
obtained from the compensation-quantity calculating section 151.
The minimum audible range compensating section 152, for example,
adds the obtained compensation quantity to the obtained minimum
audible range.
In the following, a concrete example of compensation of the minimum
audible range will be described.
Example 1
The compensation-quantity calculating section 151 holds a
compensation quantity suited to a noise quantity. FIG. 7 is a
schematic view illustrating an example of a relationship between
noise quantity and compensation quantity. In an example illustrated
in FIG. 7, noise is classified into three levels, loud, normal, and
soft, and corresponding compensation quantities to the levels are
illustrated. In an example illustrated in FIG. 7, B1 represents a
compensation quantity of a spectrum whose noise quantity is loud,
B2 represents a compensation quantity of a spectrum whose noise
quantity is normal, and B3 represents a compensation quantity of
the spectrum whose noise quantity is soft.
The compensation-quantity calculating section 151 determines which
noise level corresponds to the obtained noise quantity with a
determination using a threshold value or the like, to obtain a
compensation quantity in response to the determination result from
the relationship illustrated in FIG. 7. The compensation-quantity
calculating section 151 outputs the obtained compensation quantity
to the minimum audible range compensating section 152.
The minimum audible range compensating section 152 adds the
compensation quantity obtained from the compensation-quantity
calculating section 151 to the minimum audible range obtained from
the hearing estimate section 103. FIG. 8 is a schematic view
illustrating an example of minimum audible ranges before and after
compensation. In an example illustrated in FIG. 8, C1 represents a
minimum audible range after compensation and C2 represents a
minimum audible range before compensation.
The minimum audible range compensating section 152 obtains a
minimum audible range after compensation (C1 illustrated in FIG. 8)
by adding to either of the compensation quantity B1 to B3 in FIG. 7
to the minimum audible range of C2 illustrated in FIG. 8. The
minimum audible range compensating section 152 outputs the
compensated minimum audible range H'(f) to the spectrum
compensation-quantity calculating section 106. The compensated
minimum audible range H'(f) is, for example, C1 illustrated in FIG.
8.
Example 2
The compensation-quantity calculating section 151 calculates a
compensation quantity by multiplying the noise quantity, or
noise_level(f) which has been obtained from the noise estimate
section 104, by constant .beta.. .beta. is a constant set to, for
example, 0.1. The compensation-quantity calculating section 151
outputs the calculated compensation quantity to the minimum audible
range compensating section 152.
The minimum audible range compensating section 152 obtains a
compensated minimum audible range by the following formula.
H'(f)=H(f)+.beta..times.noise_level(f) FORMULA (7)
H'(f): minimum audible range after compensation
H(f): minimum audible range before compensation
.beta.: constant
noise_level(f): noise quantity
FIG. 9 is a schematic view illustrating an example of a
relationship between noise quantity and minimum audible ranges
before and after compensation. In an example illustrated in FIG. 9,
D1 represents a minimum audible range after compensation, D2
represents a minimum audible range before compensation, and D3
represents a noise quantity.
The minimum audible range compensating section 152 obtains the
compensated minimum audible range (D1 illustrated in FIG. 9) by
adding a compensation quantity multiplied by the constant .beta.,
to the minimum audible range of D2 illustrated in FIG. 9. The
minimum audible range compensating section 152 outputs the
compensated minimum audible range H'(f) to the spectrum
compensation-quantity calculating section 106.
Thus, based on the estimated noise, it is possible to compensate a
minimum audible range estimated from an age of a user.
Returning to FIG. 2, the spectrum compensation-quantity calculating
section 106 compares the spectrum R(f) of the received sound and
the compensated minimum audible range H'(f) to obtain the spectrum
compensation quantity G(f). For example, the spectrum
compensation-quantity calculating section 106 obtains a spectrum
compensation quantity based on the following condition.
G(f)=H'(f)-R(f) if R(f)<H'(f), G(f)=0 if R(f).gtoreq.H'(f)
The spectrum compensation-quantity calculating section 106 outputs
the obtained spectrum compensation quantity G(f) to the spectrum
compensation section 107.
The spectrum compensation section 107 obtains compensated received
sound spectrum R'(f) from, for example, the spectrum R(f) of the
received sound and the spectrum compensation quantity G(f) by the
following formula. R'(f)=R(f)+G(f) FORMULA (8)
FIG. 10 is a schematic view illustrating spectrum compensation. E1
illustrated in FIG. 10 represents a compensated minimum audible
range H'(f), E2 represents a received sound spectrum R(f) before
compensation, and E3 represents a compensated received sound
spectrum R'(f). The spectrum compensation section 107 compensates
the received sound spectrum R(f) so that the compensated received
sound spectrum R'(f) becomes greater than the minimum audible range
H'(f). The spectrum compensation section 107 outputs the
compensated spectrum R'(f) of the received sound to the inverse
frequency transform section 108.
The spectrum compensation-quantity calculating section 106 may
compensate the received sound spectrum only within a predetermined
frequency band. The predetermined frequency band is, for example, a
low frequency band and/or a high frequency band where hearing tends
to be reduced. This is because bands where hearing tends to be
reduced are known.
Returning to FIG. 2, the inverse frequency transform section 108
applies an inverse frequency transform (inverse FFT) to the
compensated received sound spectrum R'(f) obtained from the
spectrum compensation section 107, to obtain compensated received
signal r'(t). The compensated received signal r'(t) is output from
a speaker to make an output sound.
Thus, the voice control device 1 estimates hearing of a user based
on a ratio of volume of sending sound and volume of received sound,
and controls the voice in response to the hearing of a user to
provide a voice easy to hear for the user automatically.
Also, the voice control device 1 compensates a minimum audible
range estimated from the age of the user based on estimated noise
to provide a voice even easier to hear for the user.
Here, the noise estimate section 104 and the hearing compensation
section 105 are not necessarily required to be configured. In this
case, the spectrum compensation-quantity calculating section 106
may calculate the spectrum compensation quantity using the hearing
(minimum audible range) estimated by the hearing estimate section
103.
<Operation>
Next, operations of the voice control device 1 will be described
according to the embodiment. FIG. 11 is a flowchart illustrating an
example of a voice control procedure according to the
embodiment.
At Step S101 illustrated in FIG. 11, the
sending/received-sound-ratio calculating section 131 calculates a
sending/received sound ratio from the volume of a received sound
and the volume of a sending sound.
At Step S102, based on the calculated sending/received sound ratio,
the age estimate section 132 estimates an age from the information
that indicates a relationship between the sending/received sound
ratio and age.
At Step S103, based on the estimated age, the minimum audible range
estimate section 133 estimates a minimum audible range from the
information that indicates a relationship between age (or
generation) and minimum audible range.
At Step S104, the hearing compensation section 105 compensates the
estimated minimum audible range based on noise included in the
sending sound. This compensation procedure will be described using
FIG. 12.
At Step S105, the spectrum compensation-quantity calculating
section 106 calculates the compensation quantity of the received
sound spectrum so that the received sound becomes greater than the
compensated minimum audible range.
At Step S106, the spectrum compensation section 107 compensates the
received signal by adding the calculated compensation quantity, or
the like.
Thus, it is possible to provide a voice easy to hear for a user
during a call in response to the hearing of the user.
FIG. 12 is a flowchart illustrating an example of a hearing
compensation procedure. At Step S201 illustrated in FIG. 12, the
noise estimate section 104 determines whether the average electric
power of sending sound (also called, "sending power" hereafter) is
less than the threshold value TH. If the sending power is less than
the threshold value TH (Step S201--YES), Step S202 is taken, or if
the sending power is greater than or equal to the threshold value
TH (Step S202--NO), Step S203 is taken. If the sending power is
less than TH, the noise estimate section 104 determines that the
sending sound includes no sound.
At Step S202, the noise estimate section 104 updates the noise
quantity using the sending sound spectrum of a current frame by the
formula (6).
At Step S203, the hearing compensation section 105 compensates the
minimum audible range based on the estimated noise quantity (see
FIGS. 8-9).
Thus, it is possible to make a voice easier to hear in response to
the surrounding noise by compensating the minimum audible range if
the noise quantity of the surroundings is loud. Here, according to
the embodiment, compensation of the minimum audible range by a
noise quantity is not necessarily required to obtain a sufficient
effect.
As above, it is possible to execute voice control in response to a
user's hearing without forcing a burden on the user according to
the embodiment. Also, it is possible to execute voice control
suited to a user because the user is not required to carry out a
voice control operation, hence voice control can be done
automatically during a call.
Also, the processing by the hearing estimate section 103 may be
done at a predetermined timing (once a week, once a month, etc.) so
that only hearing compensation by a noise quantity is executed
usually. This is because it is not required to execute a hearing
estimation every time if the user remains unchanged.
Also, the sending/received-sound-ratio calculating section 131 may
calculate a sending/received sound ratio if a sending sound and a
received sound include sound (voice). A determination of sound
inclusion may be done with a known technology.
For example, in Japanese Patent No. 3849116, voice or non-voice is
determined for each frame of an input signal, based on a first
voice characterizing quantity calculated using electric power, a
zero crossing rate, a peak frequency of a power spectrum, pitch
cycle, etc., and a second voice characterizing quantity calculated
based on a difference of peak frequency of power spectrum only in
high order components. This makes it possible to estimate the
hearing of a user based on the volume of a sending sound and the
volume of a received sound.
Modified Example
FIG. 13 is a block diagram illustrating an example of hardware of a
portable terminal device 200 according to a modified example. The
portable terminal device 200 includes an antenna 201, a radio
section 202, a baseband processing section 203, a control section
204, a microphone 205, a speaker 206, a main memory section 207, an
auxiliary storage section 208, and a terminal interface section
209.
The antenna 201 sends a radio signal amplified by the sending
amplifier, and receives a radio single sent from a base station.
The radio section 202 applies a D/A conversion to the sending
signal spread by the baseband processing section 203, converts it
to a high frequency signal by a quadrature modulation, and
amplifies the signal by a power amplifier. The radio section 202
amplifies the received radio signal and applies an A/D conversion
to the amplified signal to send it to the baseband processing
section 203.
The baseband section 203 executes various baseband processing such
as an addition of error correcting codes to sending data, data
modulation, spread modulation, despreading of a received signal,
determination of receiving environment, determination of a
threshold value for each channel signal, error correction decoding,
etc.
The control section 204 executes radio control such as
sending/receiving of a control signal. Also, the control section
204 executes the voice control program stored in the auxiliary
storage section 208 to execute voice control according to the
embodiment.
The main memory section 207 is a ROM (Read-Only Memory), a RAM
(Random Access Memory) or the like, which is a storage device to
store or to temporarily store an OS, or the basic software executed
by the control section 204, programs such as application software
or the like, and data.
The auxiliary storage section 208 is an HDD (Hard Disk Drive) or
the like, which is a storage device to store data related to the
application software and the like. For example, the information
illustrated in FIGS. 4, 5 and 7 is stored in the auxiliary storage
section 208.
The terminal interface section 209 executes adapter processing for
data, and interface processing with a handset and an external data
terminal.
This makes it possible to provide a voice in response to hearing of
a user automatically during a call on the portable terminal device
200. Also, it is possible to implement the voice control device 1
according to the embodiments as one or multiple semiconductor
integrated circuits in the portable terminal device 200.
Also, the disclosed technology can be implemented not only on the
portable terminal device 200, but also on other devices. In the
modified example, although the example is explained in which the
voice control device is implemented on a portable terminal device,
the voice control device described above or the voice control
procedures described above are applicable, for example, to a TV
telephone conference device and an information processing device
having a telephone function, a fixed telephone, and the like.
Also, it is possible to have a computer execute voice control
described in the above embodiment by recording a program
implementing voice control processing according to the above
embodiment in a recording medium.
Also, it is possible to implement the above voice control
processing by recording the program on a recording medium and
having a computer or a portable terminal device read the recording
medium on which the program is recorded. Here, various types of
recording media can be used including a recording medium that
records information optically, electrically, or magnetically such
as a CD-ROM, a flexible disk, an optical magnetic disk and the
like, and a semiconductor memory and the like that records
information electrically such as a ROM, a flash memory, and the
like.
All examples and conditional language recited herein are intended
for pedagogical purposes to aid the reader in understanding the
invention and the concepts contributed by the inventor to
furthering the art, and are to be construed as being without
limitation to such specifically recited examples and conditions,
nor does the organization of such examples in the specification
relate to a showing of the superiority and inferiority of the
invention. Although the embodiments of the present invention have
been described in detail, it should be understood that the various
changes, substitutions, and alterations could be made hereto
without departing from the spirit and scope of the invention.
* * * * *