close

Вход

Забыли?

вход по аккаунту

?

DESCRIPTION JP2005159678

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 JP2005159678
An object of the present invention is to enable a comfortable loud-speaking call by suppressing
one-sided fall of a voice switch. SOLUTION: A gain G2 of a second amplifier 6 for amplifying a
received signal arriving from a call terminal on the opposite side through a transmission system
is variable, and a background noise power estimated value Pn (R) on the received side is a
predetermined value. A gain control unit 20 is provided which decreases the gain G2 of the
second amplifier 6 if the threshold value Xc is exceeded. If the ambient noise level at the far end
increases and the receiver background noise power estimated value Pn (R) exceeds the threshold
Xc, the gain control unit 20 amplifies the gain G2 of the second amplifier 6 that amplifies the
received signal. In order to reduce the initial value G2 to G2 further to G2, the level of the
reception signal also decreases and the receiving side instantaneous power estimated value Ps
(R) also decreases, so that the erroneous determination of the call state in the insertion loss
amount control unit 10 is prevented. One-sided collapse of the voice switch VS can be
suppressed. [Selected figure] Figure 1
Speakerphone
[0001]
TECHNICAL FIELD The present invention relates to a speakerphone used for an interphone or
the like.
[0002]
Conventionally, it is not necessary to have a handset at the time of a call, and an audio signal
10-04-2019
1
transmitted from the other party's call terminal is amplified by the speaker to the caller away
from the call terminal, and the voice of the caller is A loud speaker communicator is provided
which realizes a loud speaker (hands-free call) by collecting sound with a microphone and
transmitting it to the other party's call terminal.
In such a loudspeaker, part of the voice emitted by the caller is reflected by the acoustic coupling
from the speaker of the other party's call terminal to the microphone, the impedance mismatch
between the call terminal and the transmission path, etc. Due to the cause, the signal may be fed
back again in a superimposed manner to the receiving signal, and if the level of this feedback
component is large, the caller may be heard as an unpleasant echo (acoustic echo or line echo).
Also, if a closed loop is formed in the speech system due to the above acoustic coupling,
reflection, and acoustic coupling at the own terminal, if there is a frequency component where
the closed loop loop gain exceeds one, feedback occurs at that frequency, which is stable It will
be impossible to continue the call. Therefore, in designing a loudspeaker as a call terminal, how
to suppress the above-mentioned unpleasant echo and howling becomes an important issue.
[0003]
In order to solve such problems, conventionally, a voice switch that constantly estimates the call
state (sending state, receiving state, etc.) and inserts loss to the transmitting path and receiving
path with appropriate allocation based on the estimation result A scheme has been widely used
to reduce closed loop loop gain and suppress unpleasant echoes and howlings. FIG. 14 is a block
diagram showing a conventional example of a so-called hands-free intercom, which comprises an
intercom master unit (hereinafter abbreviated as "master") M as a speakerphone and a door
phone handset S as the other party's call terminal. (See Patent Document 1). The master unit M
arrives from the microphone 1, the speaker 2, the two-wire / four-line conversion circuit 30, the
first amplifier 31 for amplifying the transmission signal output from the microphone 1, and the
other party's call terminal via the transmission system. It comprises a second amplifier 32 for
amplifying a received signal and an audio switch VS '. Moreover, although illustration is
abbreviate | omitted, the door phone subunit S is comprised with a microphone, a speaker, a 2
wire-4 wire conversion circuit etc.
[0004]
Further, the voice switch VS ′ is a transmission side loss insertion means 33 for inserting a loss
in the transmission side signal path from the microphone 1 through the first amplifier 31 to the
10-04-2019
2
two-wire to four-wire conversion circuit 30, and a second amplifier 32. Receiver loss insertion
means 34 for inserting losses in the receiver signal path from 32 to the speaker 2 and insertion
loss control means 35 for controlling the insertion loss amount in each of the transmitter and
receiver loss insertion means 33, 34 And The insertion loss amount control means 35 estimates,
for example, the powers of the transmission signal and the reception signal, compares the
magnitude relationship of these estimated values, and compares the magnitude relationship of
the estimated values with the predetermined loss insertion means 33, A process of switching
between the transmitting state and the receiving state is performed by inserting a loss. JP, 2000307745, A
[0005]
However, in the above-described conventional example, when the difference between the
ambient noise level on the far end (door phone handset S) side and the ambient noise level on the
near end (parent device M) side is large, for example, the door phone handset S installed
outdoors. When a loud noise such as wind noise or automobile noise is input to the microphone
of the microphone, the insertion loss amount control means 35 that monitors the transmission
signal and the reception signal to estimate the call state, for example, In a situation where the
ambient noise level is large, it is always determined to be in the receiving state, and in a situation
where the ambient noise level on the near end side is large, it is always determined to be the
transmitting state. The phenomenon of fixing the call state to one of the states (a so-called onesided collapse of the voice switch) occurs.
[0006]
The present invention has been made in view of the above-mentioned circumstances, and an
object thereof is to provide a loudspeaker communication apparatus capable of comfortable
speech communication by suppressing one-sided collapse of a voice switch.
[0007]
According to the first aspect of the present invention, in order to achieve the above object, a
microphone and a speaker, a first amplifier for amplifying a transmission signal output from the
microphone, and a reception that arrives from the other party's call terminal via the transmission
system. A second amplifier for amplifying the signal, an audio switch for transmitting a
transmitting signal to the transmission system and attenuating the receiving signal, and a voice
switch for switching between a receiving state for transmitting the receiving signal to the speaker
and attenuating the transmitting signal The voice switch comprises a transmitting side loss
insertion means for inserting a loss in a transmitting side signal path from a microphone to a first
amplifier, and a receiving side for inserting a loss in a receiving side signal path from a second
amplifier to a speaker And an insertion loss amount control means for controlling an insertion
10-04-2019
3
loss amount in each of the transmission and reception side loss insertion means, the insertion
loss amount control means comprising: Speech side instantaneous power estimation unit for
estimating the instantaneous power of the transmission line, transmission side background noise
power estimation unit for estimating background noise power of the transmission signal,
instantaneous power estimation value for the transmission side, and background noise power
estimation value A first voice section detection unit for detecting a voice section of a transmission
signal from the receiver, a receiver-side instantaneous power estimation unit for estimating
instantaneous power of a reception signal, and receiver-side background noise power estimation
for estimating background noise power of a reception signal Section and a second speech zone
detection unit for detecting the speech zone of the reception signal from the reception side
instantaneous power estimation value and the background noise power estimation value, and
comparing the transmission side and reception side instantaneous power estimation values A
transmission mode in which the call state is determined with reference to the result and the
detection results of the first and second voice section detection units, and the insertion loss
amount on the receiver side is relatively increased according to the determination result of the
call state Or send In the speakerphone switched to at least one of the receiving modes in which
the insertion loss amount on the side is relatively increased, the gain of the first amplifier is
reduced if the transmitter background noise power estimated value exceeds a predetermined
threshold And at least one of a first gain control unit for causing the second amplifier to decrease
the gain of the second amplifier when the receiver background noise power estimated value
exceeds a predetermined threshold value. Do.
[0008]
According to the present invention, when the ambient noise level on the far end side (or near end
side) is large, the instantaneous power estimated value on the reception side (or transmission
side) also becomes large, so that the insertion loss amount control means The voice switch may
fall to one side to the receiving side (or transmitting side) by erroneously determining the state
as the receiving state (or transmitting state), and in such a case, background noise from the
receiving side (or transmitting side) The instantaneous power of the receiver (or transmitter) by
reducing the gain of the second (or first) amplifier that amplifies the receiver signal (or the
transmitter signal) if the power estimate exceeds a predetermined threshold. Since the estimated
value also decreases, it is possible to prevent the insertion loss amount control means from
erroneously determining the speech state as the reception state (or transmission state), and to
suppress the one-sided collapse of the voice switch.
[0009]
According to the second aspect of the present invention, in order to achieve the above object, a
microphone and a speaker, a first amplifier for amplifying a transmission signal outputted from
the microphone, and an incoming voice arriving from the other party's call terminal via the
10-04-2019
4
transmission system. A second amplifier for amplifying the signal, an audio switch for
transmitting a transmitting signal to the transmission system and attenuating the receiving
signal, and a voice switch for switching between a receiving state for transmitting the receiving
signal to the speaker and attenuating the transmitting signal The voice switch comprises a
transmitting side loss insertion means for inserting a loss in a transmitting side signal path from
a microphone to a first amplifier, and a receiving side for inserting a loss in a receiving side
signal path from a second amplifier to a speaker And an insertion loss amount control means for
controlling an insertion loss amount in each of the transmission and reception side loss insertion
means, the insertion loss amount control means comprising: Speech side instantaneous power
estimation unit for estimating the instantaneous power of the transmission line, transmission side
background noise power estimation unit for estimating background noise power of the
transmission signal, instantaneous power estimation value for the transmission side, and
background noise power estimation value A first voice section detection unit for detecting a voice
section of a transmission signal from the receiver, a receiver-side instantaneous power estimation
unit for estimating instantaneous power of a reception signal, and receiver-side background noise
power estimation for estimating background noise power of a reception signal Section and a
second speech zone detection unit for detecting the speech zone of the reception signal from the
reception side instantaneous power estimation value and the background noise power estimation
value, and comparing the transmission side and reception side instantaneous power estimation
values A transmission mode in which the call state is determined with reference to the result and
the detection results of the first and second voice section detection units, and the insertion loss
amount on the receiver side is relatively increased according to the determination result of the
call state Or send A first attenuation for attenuating a transmission signal input as a reference
signal to a first voice section detection unit in a speakerphone switched to at least one of
reception modes in which the insertion loss amount on the side is relatively increased And the
first attenuation amount control unit that increases the attenuation amount of the first attenuator
if the estimated value of the transmitter and the transmission side background noise power
exceeds a predetermined threshold, or is input as a reference signal to the second speech section
detection unit At least one of a second attenuator for attenuating the reception signal and a
second attenuation control unit for increasing the attenuation of the second attenuator when the
receiver background noise power estimated value exceeds a predetermined threshold. It is
characterized by
[0010]
According to the present invention, when the ambient noise level on the far end side (or near end
side) is large, the instantaneous power estimated value on the reception side (or transmission
side) also becomes large, so that the insertion loss amount control means The voice switch may
fall to one side to the receiving side (or transmitting side) by erroneously determining the state
as the receiving state (or transmitting state), and in such a case, background noise from the
receiving side (or transmitting side) The receiver side (or the transmitter side) by increasing the
10-04-2019
5
amount of attenuation of the second (or first) attenuator that attenuates the receiver signal (or
the transmitter signal) if the power estimate exceeds a predetermined threshold. Since the
instantaneous power estimated value also decreases, it is possible to prevent the insertion loss
amount control means from erroneously determining the speech state as the reception state (or
transmission state), and to suppress the one-sided collapse of the voice switch.
[0011]
According to the invention of claim 3, in the invention of claim 1 or 2, the first and second gain
control units or the first and second attenuation amount control units are transmitter side or
receiver side obtained at predetermined time intervals. The gain of the first and second amplifiers
or the attenuation of the first and second attenuators is updated at predetermined time intervals
sufficiently longer than the duration of the phoneme based on the background noise power
estimation value of .
[0012]
According to the present invention, since the gains of the first and second amplifiers or the
attenuations of the first and second attenuators are updated at predetermined time intervals
sufficiently longer than the duration of the phoneme, they should be attenuated in principle. It is
possible to prevent non-speech from being attenuated.
[0013]
According to the invention of claim 4, in the invention of claim 1 or 2, the first and second gain
control units or the first and second attenuation amount control units are configured as a
transmitter side or a receiver side background at predetermined time intervals. The gains of the
first and second amplifiers or the attenuations of the first and second attenuators are updated at
predetermined time intervals sufficiently longer than the duration of the phoneme based on the
average value of the noise power estimation value. Do.
[0014]
According to the present invention, since the gains of the first and second amplifiers or the
attenuations of the first and second attenuators are updated at predetermined time intervals
sufficiently longer than the duration of the phoneme, they should be attenuated in principle. It is
possible to prevent non-speech from being attenuated.
[0015]
According to the invention of claim 5, in the invention of claim 1 or 2, the first gain control unit
10-04-2019
6
and the first attenuation amount control unit have a transmitter-side background noise power
estimated value equal to or less than a predetermined first threshold. When the gain of the first
amplifier and the attenuation amount of the first attenuator are constant values, the output of the
first amplifier and the first attenuation when the transmitter background noise power estimate
exceeds the first threshold The gain of the first amplifier and the attenuation amount of the first
attenuator are adjusted so that the power of the output of the amplifier becomes substantially
constant, and the detection result of the first voice section detection unit continues to be not
constant for a predetermined time or more. When it is voice, the first threshold is decreased, and
the second gain control unit and the second attenuation amount control unit determine that the
receiver-side background noise power estimated value is less than or equal to the predetermined
second threshold. Setting the gain of the second amplifier and the attenuation of the second
attenuator to a constant value, and Second amplifier gain and second attenuator attenuation such
that the power of the second amplifier output and second attenuator output is substantially
constant when the noise power estimate exceeds the second threshold. The amount is adjusted,
and the second threshold is decreased when the detection result of the second speech section
detection unit continues to be non-speech for a predetermined time or more.
[0016]
According to the present invention, when the detection results of the first voice section detection
unit and the second voice section detection unit continue to be non-voice continuously for a
predetermined time or more, only background noise is transmitted to the transmission signal and
the reception signal. As it is considered to be included, in such a situation, by reducing the first
threshold and the second threshold, the voice switch can be reduced by suppressing an increase
in background noise power amplified by the first or second amplifier. Can be more reliably
suppressed.
[0017]
According to the invention of claim 6, in the invention of claim 5, the first and second gain
control units and the first and second attenuation amount control units reduce the first or second
threshold from an initial value. The first or second threshold value may be returned to an initial
value if a voice section is detected by the first or second voice section detector when in the state.
[0018]
According to the present invention, if the first or second threshold value is decreased even after
the voice section is detected, the voice may not be sufficiently amplified and it may be difficult to
hear, so if the voice section is detected The first or second threshold can be returned to the initial
value to prevent the voice from being amplified sufficiently to be difficult to hear.
10-04-2019
7
[0019]
According to the invention of claim 7, in the invention of claim 6, the first and second gain
control units and the first and second attenuation amount control units receive detection results
of the first or second voice section detection unit. The first and second thresholds are gradually
reduced each time the duration of the non-voice state exceeds a predetermined threshold.
[0020]
According to the present invention, the first and second threshold values are gradually reduced
as the duration of the non-speech section increases, thereby further suppressing an increase in
background noise power amplified by the first or second amplifier. As a result, the voice switch
can be more reliably suppressed from falling over.
[0021]
According to the invention of claim 8, according to the invention of any one of claims 3 to 7, the
insertion loss amount control means determines whether the communication state is a
transmission state or a reception state and according to the communication state, the
transmission side and the reception side. The first gain control unit or the first gain control unit
is controlled only by controlling the insertion loss amount in each loss insertion means on the
side and when the determination result of the insertion loss amount control means continues to
be in the transmission state for a predetermined time or more. Operating the second gain control
unit or the second attenuation control unit only when the attenuation control unit is operated
and the determination result of the insertion loss control means continues to be in the receiving
state for a predetermined time or more. It features.
[0022]
According to the present invention, when the determination result of the insertion loss amount
control means is fixed to either the talk state or the talk state continuously for a predetermined
time or longer, the fixed side (transmission side or reception side) It is considered that the
ambient noise level on the side) is relatively large, so that each part can be operated only when
the determination result continues for a predetermined time or more, and it is possible to
efficiently suppress one-sided collapse and prevent attenuation of voice. .
[0023]
According to the present invention, when the ambient noise level on the far end side (or near end
side) is large, the instantaneous power estimated value on the reception side (or transmission
side) also becomes large, so that the insertion loss amount control means The voice switch may
10-04-2019
8
fall to one side to the receiving side (or transmitting side) by erroneously determining the state
as the receiving state (or transmitting state), and in such a case, background noise from the
receiving side (or transmitting side) The instantaneous power of the receiver (or transmitter) by
reducing the gain of the second (or first) amplifier that amplifies the receiver signal (or the
transmitter signal) if the power estimate exceeds a predetermined threshold. Since the estimated
value also decreases, it is possible to prevent the insertion loss amount control means from
erroneously determining the speech state as the reception state (or transmission state), and to
suppress one-sided collapse of the voice switch.
[0024]
(First Embodiment) A loudspeaker communicator according to this embodiment includes a
microphone 1, a speaker 2, a first amplifier 5, a second amplifier 6, and an audio switch VS as
shown in FIG. It is common with the machine (interphone master machine M).
[0025]
The voice switch VS in this embodiment includes a transmitting side loss insertion unit 3 inserted
in a transmitting signal path for transmitting a transmitting signal to a channel, and a receiving
side signal for transmitting a receiving signal to the speaker 2 The receiver loss insertion unit 4
is inserted into the path, and the insertion loss control unit 10 controls the insertion loss amount
of the transmitter loss insertion unit 3 and the receiver loss insertion unit 4 according to the call
state.
In addition, the insertion loss amount control unit 10 includes a first voice interval detection unit
11 that detects a voice interval of a transmission signal, and a second voice interval detection
unit 12 that detects a voice interval of a reception signal.
The first voice activity detection unit 11 estimates the instantaneous power of the transmission
signal (the signal at point B) input to the transmission loss insertion unit 3; A transmitting side
background noise power estimating unit (not shown) for estimating the background noise power
of the speech signal and a ratio (== the instantaneous power estimated value Ps (T) of the
transmitting side to the background noise power estimated value Pn (T) A determination unit
(not shown) that compares Ps (T) / Pn (T)) with a predetermined threshold value δ and
determines that the section is an audio section when the ratio is equal to or greater than the
threshold value δ.
10-04-2019
9
Similarly, the second voice activity detection unit 12 estimates a momentary power of a reception
signal (signal at point C) input to the reception side loss insertion unit 4, and a reception signal
(not shown); Receiver background noise power estimation unit (not shown) for estimating
background noise power and ratio of instantaneous power estimated value Ps (R) to background
noise power estimated value Pn (R) (= Ps (R) A determination unit (not shown) configured to
compare / Pn (R) with a predetermined threshold value δ and determine that the section is a
voice section when the ratio is equal to or greater than the threshold value δ.
When the first and second voice section detectors 11 and 12 detect a voice section, the voice
section detection signals SDF (T) and SDF (R) on the transmitting side and the receiving side are
set to "1", respectively. When the voice section is not detected (in the non-voice section), the
voice section detection signals SDF (T) and SDF (R) are set to "0".
The instantaneous power estimation units on the transmitting side and receiving side are
configured with a filter or the like having a characteristic that the rising is steep and the falling is
gradual, and the background noise power estimating units on the transmitting and receiving
sides are gentle And the like, and is composed of a filter or the like having a steep falling
characteristic.
[0026]
Furthermore, the insertion loss amount control unit 10 returns from the input point B to the
transmission side loss insertion unit 3 to the input point C to the reception side loss insertion
unit 4 after passing through the transmission side loss insertion unit 3 and the line side. Channel
gain multiplication unit 13 having a value determined according to the gain of the system to be
used as a coefficient, receiver loss insertion unit 4 from the input point C to receiver loss
insertion unit 4 and the acoustic side (microphone 1 and speaker 2 Acoustic coupling gain
multiplication unit 14 having as a coefficient a value determined according to the gain of the
path to the input point B to the transmission side loss insertion unit 3 after the wraparound in 2),
and the second voice section detection unit An output signal P2 obtained by inputting the
receiving side instantaneous power estimated value Ps (R) output from 12 to the acoustic
coupling gain multiplication unit 14 and a transmission side instantaneous power estimation
output from the first voice section detection unit 11 A first comparator 15 for comparing the
magnitude relation with the value Ps (T); A second comparator 16 for comparing the magnitude
relationship between an output signal P1 obtained by inputting the talking side instantaneous
power estimated value Ps (T) to the channel feedback gain multiplication unit 13 and the
10-04-2019
10
receiving side instantaneous power estimated value Ps (R) , And the output signals C1 and C2 of
the first comparator 15 and the second comparator 16 and the output signal C3 (= SDF (T)) of
the first voice section detection unit 11 and the second voice section detection unit 12 , C4 (=
SDF (R)) to determine the communication state, and include an insertion loss amount distribution
processing unit 17 that controls the loss and insertion loss amount of the transmission side loss
insertion unit 3 and the reception side loss insertion unit 4 Do.
Here, the output signal C1 of the first comparator 15 is “0” when Ps (T) <P2, and is “1”
when Ps (T) 2P2.
In addition, the output signal C2 of the second comparator 16 is "0" when Ps (R) ≧ P1, and is "1"
when Ps (R) <P1.
[0027]
Thus, the insertion loss amount distribution processing unit 17 determines the call state with
reference to the four binary signals C1 to C4, and the insertion loss amount of the transmission
side loss insertion unit 3 and the reception side loss insertion unit 4 is determined. decide.
Here, if C1 = C2 = 1 and C3 = 1, the transmitting mode, if C1 = C2 = 0 and C4 = 1, the receiving
mode, if C1 ≠ C2 and both C3 and C4 are not 0, fast idle In mode and other states, it is
determined to be slow idle mode, and when the determination result is the transmission mode,
the insertion loss amount of transmitting side loss insertion unit 3 is the minimum value, and the
insertion loss amount of receiving side loss insertion unit 4 is the maximum value When the
judgment result is in the receiving mode, the insertion loss amount of the transmitting side loss
insertion unit 3 is set to the maximum value, and the insertion loss amount of the receiving side
loss insertion unit 4 is set to the minimum value. Sometimes the insertion loss amounts of the
transmission loss insertion unit 3 and the reception loss insertion unit 4 are made equal to each
other in a short transition time, and the transmission loss insertion unit 3 and reception loss
insertion in a long transition time when the determination result is slow idle mode Department
Equal to each other in the insertion loss.
The configuration and operation of the voice switch VS described above are the same as those
disclosed in Patent Document 1, and thus detailed description will be omitted.
10-04-2019
11
[0028]
Next, the gist of the present embodiment will be described.
In this embodiment, the gain G2 of the second amplifier 6 for amplifying the received signal
arriving from the other party's call terminal through the transmission system is variable, and the
background noise power estimated value Pn (R) on the received side is predetermined. The
second embodiment is characterized in that a gain control unit 20 is provided to reduce the gain
G2 of the second amplifier 6 if the threshold Xc of the second amplifier 6 is exceeded.
[0029]
The gain control unit 20 sets the gain G2 of the second amplifier 6 to the initial value G20 when
the receiver-side background noise power estimated value Pn (R) is less than the threshold Xc as
shown in FIG. The gain G2 is decreased to a value G21 smaller than the initial value G20, and the
gain G2 is decreased to a value G22 smaller than G21 when the receiver-side background noise
power estimated value Pn (R) further increases by ΔX or more than the threshold Xc.
[0030]
Thus, when the ambient noise level on the far end side is large, the reception side instantaneous
power estimated value Ps (R) also becomes large, so that the insertion loss amount distribution
processing unit 17 misjudges the call state as the reception state. The voice switch Vs may fall to
the receiving side, but if the ambient noise level at the far end increases and the receiving
background noise power estimated value Pn (R) exceeds the threshold Xc, the gain control unit In
order to decrease the gain G2 of the second amplifier 6 that amplifies the reception signal by 20
from the initial value G20 to G21 to G22, the reception side instantaneous power estimated value
Ps (R) also decreases, and therefore the insertion loss amount distribution processing unit 17 Can
prevent the voice state from being erroneously determined as the receiving state, and can
suppress one-sided collapse of the voice switch VS.
[0031]
Although the gain control unit 20 is provided on the receiving side in the present embodiment,
the gain G1 of the first amplifier 5 is variable, and the background noise power estimated value
Pn (T) on the transmitting side exceeds a predetermined threshold. If a gain control unit for
reducing the gain G1 of the first amplifier 5 is provided, it is possible to prevent misjudgment of
the call state in the insertion loss amount distribution processing unit 17 as in the reception side
10-04-2019
12
and to suppress one fall of the voice switch VS. The gain control unit may be provided not only
on the transmitting side or the receiving side but also on both sides.
[0032]
Second Embodiment FIG. 3 shows a partially omitted block diagram of the present embodiment.
However, since the basic configuration of the present embodiment is the same as that of the first
embodiment, the same reference numerals are given to the common components, the description
is omitted, and only the configuration that is the feature of the present embodiment will be
described.
[0033]
In the present embodiment, the attenuator 21 for attenuating the reception signal input to the
second speech zone detection unit 12 and the attenuation of the attenuator 21 when the
reception side background noise power estimated value Pn (R) exceeds a predetermined
threshold Xc And an attenuation amount control unit 22 that increases the amount Lr.
[0034]
The attenuation amount control unit 22 sets the attenuation amount Lr of the attenuator 21 to
the initial value Lr0 (= 0 dB) when the receiver-side background noise power estimated value Pn
(R) is less than the threshold value Xc as shown in FIG. When Xc is exceeded, the attenuation Lr
is increased to a value Lr1 larger than the initial value Lr0, and when the receiver-side
background noise power estimated value Pn (R) further increases by ΔX or more from the
threshold Xc, the attenuation is further increased to a value Lr2 larger than Lr1. Increase Lr.
[0035]
Thus, when the ambient noise level on the far end side is large, the reception side instantaneous
power estimated value Ps (R) also becomes large, so that the insertion loss amount distribution
processing unit 17 misjudges the call state as the reception state. Voice switch Vs may fall to the
receiving side, but if the ambient noise level at the far end increases and the receiving
background noise power estimated value Pn (R) exceeds the threshold value Xc, attenuation
control Since the attenuation amount Lr of the attenuator 21 for attenuating the reception signal
by the unit 22 is increased from the initial value Lr0 to Lr1 to Lr2, the reception side
instantaneous power estimated value Ps (R) also decreases, and the insertion loss amount
10-04-2019
13
distribution processing unit 17 Can prevent the voice state from being erroneously determined
as the receiving state, and can suppress one-sided collapse of the voice switch VS.
[0036]
In the present embodiment, although the attenuator 21 and the attenuation amount control unit
22 are provided on the receiving side, an attenuator for attenuating the transmission signal input
to the first voice section detection unit 11, and transmission side background noise power If the
attenuation amount control unit is provided to increase the attenuation amount of the attenuator
if the estimated value Pn (T) exceeds the predetermined threshold value Xc, erroneous
determination of the call state in the insertion loss amount distribution processing unit 17 as in
the receiving side It is possible to suppress the one-sided collapse of the voice switch VS, and an
attenuator and an attenuation amount control unit may be provided not only on either the
transmitter side or the receiver side but also on both sides.
[0037]
(Third Embodiment) The basic configuration of the present embodiment is the same as that of
the first embodiment, and therefore common components are denoted by the same reference
numerals, and illustration and description thereof will be omitted as appropriate, and the gain as
a feature of the present embodiment Only the control unit 20 will be described.
[0038]
Each unit constituting the voice switch VS in the present embodiment is realized by combining
general-purpose hardware (processor) such as a DSP or a CPU and dedicated software.
Then, as shown in FIG. 5, the gain control unit 20 in the present embodiment is a second one for
converting the reception side background noise power estimated value Pn (R) into a value before
being amplified by the second amplifier 6. The multiplier 20a for multiplying the reciprocal of
the gain G2 (n) of the amplifier 6 and the converted value Pn (R) 'of the receiver-side background
noise power estimated value output from the multiplier 20a are compared with the threshold
value Xc. Gain determination unit 20b that determines gain G2 (n + 1) of amplifier 6, switch unit
20c that switches input / output of receiver background noise power estimated value Pn (R) to
multiplier 20a, and switch unit 20c And a timer unit 20d for controlling on / off of the
The timer unit 20d turns on the switch unit 20c every predetermined time (a time sufficiently
10-04-2019
14
longer than the duration time of the phoneme) to input the receiver-side background noise power
estimated value Pn (R) to the multiplier 20a.
[0039]
Although the voice is attenuated if the gain G2 (n) of the second amplifier 6 is decreased despite
the fact that the received signal includes voice, the gain control unit 20 determines that the voice
control is performed in this embodiment. In order to update the gain G2 (n) of the second
amplifier 6 every fixed time sufficiently longer than the duration of the phoneme based on the
receiver background noise power estimated value Pn (R) obtained every time, it is basically
attenuated Can be prevented from being attenuated.
[0040]
In addition, a gain control unit for controlling the gain G1 of the first amplifier 5 based on the
background noise power estimated value Pn (T) on the transmission side is provided, and the
transmission side background obtained by the gain control unit every predetermined time If the
gain G1 (n) of the first amplifier 5 is updated at predetermined time intervals sufficiently longer
than the duration of the phoneme based on the noise power estimated value Pn (T), the voice
contained in the transmission signal is generated. It can be prevented from being attenuated.
Also, based on the receiver-side background noise power estimated value Pn (R) obtained by the
attenuation amount control unit 22 in the second embodiment at each predetermined time, the
attenuation amount Lr of the attenuator 21 is a fixed time sufficiently longer than the duration of
the phoneme. The attenuation amount of the attenuator based on the transmission side
background noise power estimated value Pn (T) which is updated for each predetermined time or
the attenuation amount control unit is obtained for each predetermined time, and the attenuation
amount of the attenuator is sufficiently longer than the duration of the phoneme It does not
matter as a configuration to be updated.
[0041]
(Fourth Embodiment) The basic configuration of the present embodiment is the same as that of
the first embodiment, and therefore common components are denoted by the same reference
numerals, and illustration and description thereof are omitted as appropriate, and the gain is a
feature of the present embodiment. Only the control unit 20 will be described.
10-04-2019
15
[0042]
As in the third embodiment, each unit constituting the voice switch VS in the present
embodiment is realized by combining general-purpose hardware (processor) such as a DSP or a
CPU and dedicated software.
Then, as shown in FIG. 6, the gain control unit 20 in the present embodiment calculates an
average value Pn (R) ′ of the receiver-side background noise power estimated value Pn (R) in a
predetermined time frame (= sampling number). In order to convert the average value calculation
unit 20 e and the average value Pn (R) ′ of the receiver-side background noise power estimated
value into a value before being amplified by the second amplifier 6, the gain G 2 (n The gain of
the second amplifier 6 by comparing the multiplier 20f for multiplying the inverse number of)
and the converted value Pn (R) of the average value of the receiver background noise power
estimated value output from the multiplier 20f with the threshold value Xc. And G2 (n + 1) are
determined.
[0043]
Thus, also in the present embodiment, the gain control unit 20 determines the gain G2 (n) of the
second amplifier 6 based on the average value Pn (R) 'of the receiver-side background noise
power estimated value Pn (R) at predetermined time intervals. Since the) is updated every fixed
time which is sufficiently longer than the duration of the phoneme, it is possible to prevent the
speech which should not be attenuated from being attenuated.
[0044]
A gain control unit for controlling the gain G1 of the first amplifier 5 based on the background
noise power estimated value Pn (T) on the transmission side is provided, and this gain control
unit controls the transmission side of each predetermined time frame. If the gain G1 (n) of the
first amplifier 5 is updated at predetermined time intervals sufficiently longer than the duration
of the phoneme based on the average value of the background noise power estimated value, the
voice contained in the transmission signal is generated. It can be prevented from being
attenuated.
In addition, the attenuation amount control unit 22 in the second embodiment determines the
attenuation amount Lr of the attenuator 21 based on the average value of the background noise
10-04-2019
16
power estimation values on the reception side for each predetermined time frame, for a fixed
time sufficiently longer than the duration of the phoneme. The attenuation amount of the
attenuator is set to a constant time longer than the duration of the phoneme based on the
configuration to be updated every time, or the attenuation amount control unit based on the
average value of the background noise power estimated value on the transmitting side every
predetermined time frame. It does not matter as a configuration to be updated.
[0045]
Fifth Embodiment FIG. 7 is a partially omitted block diagram of the present embodiment, and FIG.
8 is a block diagram of a gain control unit 20. As shown in FIG.
However, since the basic configuration of this embodiment is the same as that of the first
embodiment, the same reference numerals are attached to the common components and the
description is omitted, and the configuration of the gain control unit 20 which is the feature of
this embodiment. Only explain.
[0046]
The gain control unit 20 calculates an average value Pn (R) ′ of the receiver-side background
noise power estimated value Pn (R) in a predetermined time frame (= sampling number) as
shown in FIG. , Multiplying the inverse of the gain G2 (n) of the second amplifier 6 to convert the
average value Pn (R) 'of the receiver background noise power estimated value to the value before
being amplified by the second amplifier 6 The gain G2 (n + 1) of the second amplifier 6 by
comparing the multiplier 20f and the converted value Pn (R) of the average value of the receiver
background noise power estimated value output from the multiplier 20f with the threshold value
Xc A counter 20h for monitoring the voice segment detection signal SDF (R) output from the
second voice segment detector 12 and measuring the duration of the non-voice segment, and a
counter 20h Gain is determined when the duration of the non-voice section measured by ; And a
threshold adjusting unit 20i to reduce the threshold Xc in 20g.
[0047]
The counter unit 20h increments the count value when the voice section detection signal SDF (R)
output from the second voice section detecting unit 12 is non-voice (SDF (R) = 0) every
predetermined time frame, When the voice section detection signal SDF (R) is voice (SDF (R) = 1),
10-04-2019
17
the count value is reset, and when the count value reaches a value corresponding to a
predetermined time, a count up signal is output to the threshold adjustment unit 20i Do.
Further, when the count up signal is input, the threshold adjustment unit 20i instructs the gain
determination unit 20g to change the threshold Xc to a smaller threshold Xc '.
Then, when receiving the instruction from the threshold adjustment unit 20i, the gain
determination unit 20g changes the threshold (initial value) Xc to the threshold Xc '(<Xc), and
the receiver-side background noise power estimated value Pn (R ) Decreases the gain G2 (n) of
the second amplifier 6 if the threshold value Xc 'is exceeded.
[0048]
Thus, when the detection result of the second voice section detection unit 12 is non-voice
continuously for a predetermined time or more, it is considered that only the background noise is
included in the reception signal, such a situation By reducing the threshold Xc to the threshold
Xc ′, it is possible to suppress an increase in background noise power amplified by the second
amplifier 6 and to more reliably suppress one-sided collapse of the voice switch VS.
[0049]
By the way, when the threshold value Xc 'which is reduced even after the voice section is
detected by the second voice section detection unit 12, the voice included in the reception signal
may not be sufficiently amplified and may be difficult to hear.
Therefore, as shown in FIG. 10, the threshold value is decreased from Xc to Xc 'when the nonspeech duration (the time when the voice section detection signal SDF (R) = 0) reaches the
predetermined time Tc. When the voice section is detected by the voice section detection unit 12
and the voice section detection signal SDF (R) becomes 1, a reset signal is output from the
threshold adjustment unit 20i, and the threshold in the gain determination unit 20g is set to an
initial value from Xc '. It is desirable to return to Xc to prevent sufficient amplification of the
voice contained in the received signal to make it difficult to hear.
[0050]
Further, as shown in FIG. 11, every time the duration of the state where the detection result of
10-04-2019
18
the second voice section detection unit 12 is non-voice exceeds the predetermined time Tc, the
threshold value Xc is changed from Xc to Xc ′, Xc ′ ′ (<Xc ′ In this case, the increase of the
background noise power amplified by the second amplifier 6 can be further suppressed, and the
one-sided collapse of the voice switch VS can be suppressed more reliably.
[0051]
Although the gain control unit 20 provided on the receiving side has been described in the
present embodiment, the detection result of the first voice segment detection unit 11 continues
to be non-continuous for a predetermined time or more also in the gain control unit provided on
the transmitting side. In the case of voice, the threshold in the gain determination unit is
decreased, or when the first voice interval detection unit 11 detects a voice interval, the
threshold in the gain determination unit is returned to the initial value, or the first voice interval
detection is performed. The same effect can be obtained by decreasing the threshold value
stepwise each time the duration of the state where the detection result of the unit 11 is non-voice
exceeds the predetermined time Tc.
In addition, such a configuration is applied to the second embodiment, and the attenuation
amount control unit on the transmission side or the reception side causes the detection result of
the first or second voice section detection unit 12 to continue for a predetermined time or more
and is non-speech The threshold is compared with the transmitter or receiver instantaneous
power estimated value Ps (R) when the attenuation amount is increased if there is a voice interval
detected by the first or second voice interval detection unit 12 When it is detected, the threshold
value is returned to the initial value, or the threshold value is gradually decreased each time the
duration of the state where the detection result of the first or second voice segment detection
unit 12 is non-voice exceeds a predetermined time. You may
[0052]
Sixth Embodiment FIG. 12 shows a partially omitted block diagram of the present embodiment.
However, since the basic configuration of the present embodiment is common to the first
embodiment and the third embodiment, the same reference numerals are given to the common
components and the description is omitted, and only the configuration that is the feature of the
present embodiment explain.
10-04-2019
19
[0053]
This embodiment is characterized in that the gain control unit 20 is operated only when the
determination result of the insertion loss amount distribution processing unit 17 continues to be
in the receiving state for a predetermined time or more, and the insertion loss amount
distribution processing unit 17 The process shown in the flowchart of FIG. 13 is performed each
time the determination of the call state is performed.
[0054]
That is, the insertion loss amount distribution processing unit 17 determines the call state with
reference to the four output signals C1 to C4 (step 1), and then decrements the count value of
the counter if the determination result is the reception mode (step 2) Further, it is determined
whether the count value is 0 (step 3).
Then, if the count value is 0, the count value is set to 1 (step 4) and then the operation of the
gain control unit 20 is started (step 5). If the count value is not 0, the process is ended without
doing anything. Do.
On the other hand, when the determination result of the call state is other than the receiving
mode, the count value is initialized to an initial value of 2 or more (step 6), and the process is
ended.
[0055]
If it is determined that the reception mode is continued for a predetermined time or more until
the count value is decremented from the initial value to 0, the ambient noise level on the
reception side is considered to be relatively large. Only when the determination result continues
for a predetermined time or more, the gain control unit 20 can be operated to efficiently
suppress the one-sided collapse.
[0056]
Even if the configuration of the present embodiment is applied to the fourth or fifth embodiment,
the same effect can be obtained.
10-04-2019
20
[0057]
FIG. 1 is a block diagram showing a first embodiment.
FIG.
FIG. 7 is a block diagram showing a second embodiment.
FIG.
FIG. 16 is a block diagram showing a gain control unit in a third embodiment.
FIG. 16 is a block diagram showing a gain control unit in a fourth embodiment.
FIG. 16 is a block diagram showing Embodiment 5;
It is a block diagram which shows the gain control part in the same as the above.
FIG.
It is a time chart for operation explanation same as the above.
It is a time chart for operation explanation same as the above.
FIG. 18 is a block diagram showing a sixth embodiment.
It is a flowchart for operation | movement description same as the above.
10-04-2019
21
It is a block diagram which shows a prior art example.
Explanation of sign
[0058]
Reference Signs List 1 microphone 2 speaker 3 transmitting side loss insertion unit 4 receiving
side loss insertion unit 6 second amplifier 10 insertion loss amount control unit 12 second voice
interval detection unit 20 gain control unit
10-04-2019
22
Документ
Категория
Без категории
Просмотров
0
Размер файла
37 Кб
Теги
description, jp2005159678
1/--страниц
Пожаловаться на содержимое документа