JP2010050512A - Voice mixing device, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a voice mixing device capable of catching voices of all uttering persons when having a voice conference in an environment with much background noise. <P>SOLUTION: The voice mixing device includes: a voice information receiving means for receiving a plurality of pieces of voice information from the outside; an existent voice head part existence/non-existence detecting means for detecting the existence/non-existence of an existent voice head part in outputs in each voice information of the voice information receiving means; a number of voices detecting means for detecting the number of voices which are equal to or more than an existent voice threshold in the outputs of the each voice information of the voice information receiving means; a gain control means for raising only the volume of existent voice information showing the existence of the existent voice head part in the outputs of the existent voice head part existence/non-existence detecting means and not raising the volume of the other voice information when the output of the number of voices detecting means is less than a threshold of the number of valid voices; a voice information combining means 124 for adding the outputs in the each voice information of the gain control means; and voice information transmitting means 101 and 102 for transmitting an output of the voice information combining means to the outside. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a voice mixing device used in a voice call system in which a plurality of voice information is input, and in particular, a handset composed of a microphone and a speaker while a flight attendant or the like stays in each assigned place in an airplane with a lot of background noise. The present invention relates to an audio mixing device in an audio control server when a conference is performed using the same.

  A conventional audio mixing apparatus will be described below.

  Conventionally, an audio mixing device described in Patent Document 1 is known. FIG. 9 is a block diagram showing the configuration of the audio mixing device described in Patent Document 1. In FIG. In FIG. 9, 1-3 are voice information transmission / reception units, 4-6 are signal level detection units, 7 is a priority selection unit, 8 is a control unit, and 9 is a voice information synthesis unit.

The operation of the conventional audio mixing apparatus configured as described above will be described below. First, audio information via the network is received by the audio information transmission / reception units 1 to 3 and transmitted to the signal level detection units 4 to 6 and the priority selection unit 9. When unconditionally synthesizing voices from five or more conference rooms, there is a problem that echoes from each conference room are superimposed, making it difficult to hear the synthesized voice. Therefore, the control unit 8 that has received the presence / absence of voice and the time when the voiced voice part is recognized from the signal level detectors 4 to 6 determines the number of lines in which the voiced voiced part is detected. It is checked whether or not the number is smaller than a preset N (N is a positive integer) line. If the number is smaller, the priority selection unit 9 is controlled, and only the line in which a voiced voice portion is detected is selected as the voice information synthesis unit 7. Connect to. In addition, when there are many, the time when the voiced portion of the voice is recognized on these lines is checked in time series, N lines are selected from the earliest according to the order in which the voices are generated, and the priority selection unit 9 is controlled, The selected N line is connected to the voice information synthesis unit 7.
JP-A-4-84553

  However, in the above conventional configuration, the time when the voiced portion is recognized is checked in time series, and the earliest one is selected according to the order in which the voices are generated. There was a problem of destroying.

  In addition, there is a problem that it is difficult to hear when there is a lot of background noise.

  The present invention solves the above-described conventional problems, and an object thereof is to provide an audio mixing apparatus that can hear the voices of all speakers when using an audio conference in an environment with a lot of background noise. .

  In order to achieve this object, the audio mixing device of the present invention includes audio information receiving means for receiving a plurality of audio information from the outside, and whether or not there is a sound head portion in the output for each audio information of the audio information receiving means. The voiced head part presence / absence detecting means for detecting the voiced voice signal and the voice information indicating that there is a voiced head part in the output of the voiced head part presence / absence detecting means is increased, and the volume of the other voice information is not increased. A gain control unit; a voice information synthesis unit that adds outputs for each voice information of the gain control unit; and a voice information transmission unit that transmits the output of the voice information synthesis unit to the outside. Yes.

  Further, the voice information receiving means further includes a voice number detecting means for detecting the number of voices that are equal to or higher than a sound threshold value in the output for each voice information, and the gain control means is effective when the output of the voice number detecting means is effective. When the number of voices is less than the threshold value, only the volume of voice information indicating that there is a voiced head part in the output of the voiced head part presence / absence detecting means is raised, and the volume of other voice information is not raised. is doing.

  As described above, according to the present invention, when using a voice conference in an environment with a lot of background noise, an excellent effect is obtained that the voices of all the speakers can be heard.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
(Embodiment 1)
FIG. 1 is a block diagram showing a configuration of an audio mixing apparatus according to Embodiment 1 of the present invention. When the number of input voices is small, even if the background noise is large, it is easy to hear if the gain at the beginning of the speech is increased. However, when the number of input voices is large, if the gain of the first sound of the conversation is increased, the voice of the speaker who joins from the middle of the conversation becomes an obstacle to the ongoing conversation and is difficult to hear. Therefore, it has the following configuration. Reference numeral 100 denotes a network that transmits voice information to and from a voice terminal such as a handset, and 101 to 102 denote a first voice information transmission / reception unit that transmits and receives voice information to and from the network 100 to Mth (M is a positive integer). The voice information transmission / reception units 103 to 104 have presence / absence of a voice level (hereinafter referred to as “sound level”) equal to or higher than the voice threshold 120 from the reception output of the first voice information transmission / reception unit 101 to the Mth voice information transmission / reception unit 102. The first signal level detection unit to M-th signal level detection unit 105 to 106 detect voices having a sound threshold value of 120 or more at the outputs of the first signal level detection unit 103 to M-th signal level detection unit 104. A first sounded head part presence / absence detecting unit for detecting the presence / absence of a sounded head part which is a head part having a level to an Mth sounded head part presence / absence detecting unit, 121 is a first signal level detection A voice number detection unit for detecting the number of voices having a voice threshold value of 120 or more (hereinafter referred to as “valid voice number”) among the outputs of the 103rd to Mth signal level detection units 104; Gain change determination units 107 and 108 for determining whether the number is less than the threshold value 122. The gain change determination unit 123 outputs the outputs of the first sound head portion presence / absence detection unit 105 to the Mth sound head portion presence / absence detection unit 106. The first gate unit to the M-th gate unit whose outputs are controlled as gate signals, and 109 to 110 are the reception outputs of the first audio information transmitting / receiving unit 101 to the M-th audio information transmitting / receiving unit 102 as the first gate unit. A first gain control unit to an Mth gain control unit that performs gain control on the basis of the outputs of the 107th to Mth gate units 108, and 124 are outputs of the first gain control unit 109 to the Mth gain control unit 110, respectively. to add A voice information combining unit. The output of the voice information synthesis unit 124 is supplied to the first voice information transmission / reception unit 101 to the Mth voice information transmission / reception unit 102 and sent to the voice terminal via the network 100.

  Next, the operation of the audio mixing apparatus according to Embodiment 1 of the present invention will be described with reference to FIGS. FIG. 2 is a schematic diagram of the output of the m-th (m is an arbitrary integer from 1 to M) signal level detector in the first embodiment of the present invention. In the m-th signal level detection unit, which is a component of the first signal level detection unit 103 to the M-th signal level detection unit 104, first, analog sound that is the signal level of the sound part at time t (n). The level (V (n) (m)) is calculated, and then the analog sound level and the sound threshold 120 are compared, and the binarized sound level (BV (n) (m)) (sound threshold) 1 is calculated in the above case, and 0) is calculated and output if it is less than the sound threshold. Here, in the case of silence and unvoiced sound, the level value is zero.

  FIG. 3 is a schematic diagram of the output of the m-th sound head portion presence / absence detection unit according to Embodiment 1 of the present invention. From the output of the m-th signal level detection unit, the m-th sound head part presence / absence detection unit 105, which is a component of the first sound head part presence / absence detection unit 105 to the M-th sound head part presence / absence detection unit 106, The presence / absence detection of the head part of the sound is detected. As a method of detecting the presence / absence of a voiced head part, for example, the voiced head part is only when the value BV (n) (m) at the current time−the value BV (n−1) (m) at the previous time is 1. Can be assumed to be detected, and in the case of 0 or −1, it can be assumed that the head portion of sound is not detected. If the head part of sound is detected, a gain is increased, and if not, a signal that does not change the gain is output.

  FIG. 4 is a schematic diagram of the output of the voice number detection unit 121 according to Embodiment 1 of the present invention. In the sound number detection unit 121, a value obtained by adding the values of the binarized sound level BV (n) (m), which are the outputs of the first signal level detection unit 103 to the Mth signal level detection unit 104, is valid. Detect as voice number. The number of effective voices at time t (n) is represented by VP (n).

  FIG. 5 is a schematic diagram of the output of the gain change determination unit in Embodiment 1 of the present invention. The gain change determination unit 123 determines whether or not the number of effective sounds is less than the effective sound number threshold 122 by size comparison. When the number of valid voices is equal to or greater than the valid voice number threshold, the gain change is OFF, and when the number is less than the valid voice number threshold, the gain change is output as ON.

  FIG. 6 is a schematic diagram of the output of the mth gate portion in the first embodiment of the present invention. In the m-th gate unit, which is one component of the first gate unit 107 to the M-th gate unit 108, the output of the m-th sound leading portion presence / absence detection unit is controlled by the output of the gain change determination unit 123. For example, when the m-th sound head portion presence / absence detection unit indicates gain up, the gain change determination unit 123 outputs gain up when the gain change is ON, but gain invariant when the gain change is OFF. Is output. When the m-th sound head portion presence / absence detection unit indicates that the gain is unchanged, the gain unchanged is output when the output of the gain change determination unit 123 is either gain change ON or gain change OFF. The outputs of the first gate unit 107 to the Mth gate unit 108 control the gains of the first gain control unit 109 to the Mth gain control unit 110, respectively. Increase the volume for gain up. If the gain remains unchanged, the volume is not changed.

  FIG. 7 is a flowchart of gain control of the audio mixing apparatus according to Embodiment 1 of the present invention. In S701, the voice information via the network 100 is received by the first voice information transmitting / receiving unit 101 to the Mth voice information transmitting / receiving unit 102, and the first voice information to the Mth voice information at time t (n) are input. Is done. In S702, the gains of all of the first gain control unit 109 to the Mth gain control unit 110 are reset. In step S <b> 703, the first signal level detection unit 103 to the M-th signal level detection unit 104 detect the sound levels of the outputs from the first audio information transmission / reception unit 101 to the M-th audio information transmission / reception unit 102. To do. In S <b> 704, the outputs of the first signal level detection unit 103 to the M-th signal level detection unit 104 respectively correspond to the first sounded head part presence / absence detection unit 105 to the M-th sounding head part presence / absence detection unit 106. The presence / absence of the head part of the sound is detected. In step S <b> 705, the number-of-voices detection unit 121 detects the number of valid voices. In step S706, the gain change determination unit 123 determines whether the number of effective sounds is less than the effective sound number threshold. If it is greater than or equal to the valid voice number threshold, the process proceeds to S707, and if it is less than the valid voice number threshold, the process proceeds to S708. In S707, the gains of the first gain control unit 109 to the Mth gain control unit 110 are not changed. In S708, the gain is increased only for the corresponding gain control unit that has detected the presence of the leading part of the sound, and the gain is unchanged for the other gain control units. In S709, the outputs of the first gain control unit 109 to the Mth gain control unit 110 are input to the voice information synthesis unit 124 to synthesize voice information. The output of the voice information synthesis unit 124 is input to the first voice information transmission / reception unit 101 to the Mth voice information transmission / reception unit 102 and is sent to each speaker via the network 100. In S710, the process at time t (n) is ended and the process waits for the process at time t (n + 1).

  FIG. 8 is a sound state diagram for explaining the operation of the sound mixing apparatus according to the first embodiment of the present invention. It is assumed that the speakers A to E participate in the conversation and the conversation is made at the timing shown in the figure. It is assumed that the effective voice number threshold = 3. When the voiced head portion of speaker B is detected at time t (a), the number of effective voices is calculated. Since speaker A is speaking, the number of effective voices including B is two. Since the effective voice number threshold value and the effective voice number are compared and become less than the threshold value, the voice packet of the speaker B is gained up, and voice information is synthesized. When the voiced head portion of the speaker E is detected at time t (b), A and B are speaking, so the number of effective voices is 3. The voice packet of the speaker E is synthesized with the voice information synthesized with the gain unchanged. The same is true at time t (c), and the number of effective voices is three. The voice information of the speaker A is synthesized with the gain unchanged and the voice packet of the speaker A becomes equal to or greater than the effective voice threshold. When the voiced head portion of speaker D is detected at time t (d), the number of effective voices is 2, so that the voice packet of speaker D is increased in gain and voice information is synthesized.

  As described above, according to the first embodiment, when a voice conference is used with a small number of people in an environment with a lot of background noise, the gain of the first tone of the conversation that is difficult to hear is increased, so there is little recollection. It is possible to advance the conversation efficiently. On the other hand, when using a voice conference with a large number of people, the gain of the leading sound of the interrupted voice conversation is stopped, so that the voice of the speaker who joins from the middle of the conversation does not become an obstacle to the ongoing conversation. Thus, the user can use the audio conference without performing an operation that hinders the conference such as volume adjustment.

  In the above description, M voice information transmitting / receiving units, a signal level detection unit, a voiced head portion presence / absence detection unit, a gate unit, and a gain control unit are assumed. However, when performing time-sharing processing, it is sufficient to provide one for each.

  It is also possible to perform software processing using a program for causing the above-described means to function.

Further, in the above description, the inputs from the first signal level detection unit 103 to the Mth signal level detection unit 104 are the reception inputs from the first audio information transmission / reception unit 101 to the Mth audio information transmission / reception unit 102. The inputs to the first signal level detection unit 103 to the Mth signal level detection unit 104 participate in the reception inputs from the first audio information transmission / reception unit 101 to the Mth audio information transmission / reception unit 102 via the AGC. The volume level of the person may be adjusted. In this case, a more comfortable call environment can be obtained by removing the difference in the volume level of the participant caused by the distance between the microphone and the speaker being different by AGC.
(Embodiment 2)
In the first embodiment, an audio mixing apparatus that can be applied regardless of the number of participants has been described. However, when it is determined that the number of participants is less than the number of valid voices, it is not necessary to consider the size of the number of valid voices, so the following can be performed.

  FIG. 9 is a block diagram showing a configuration of an audio mixing apparatus according to Embodiment 2 of the present invention. The means and signals that can be omitted compared to FIG. 1 are the voice number detection unit 121, the gain change determination unit 123, the first gate unit 107 to the Mth gate unit 108, and the effective voice number threshold value 122. In this case, the first sounded head portion presence / absence detecting unit 105 to the Mth sounded head portion presence / absence detecting unit 106 detects that the sounded head portion is detected, and the gain is increased. Are directly supplied to the first gain control unit 109 to the Mth gain control unit 110.

  FIG. 10 is a flowchart of gain control of the audio mixing apparatus according to the second embodiment of the present invention. Steps that can be omitted in comparison with FIG. 7 include detecting the number of effective sounds in the sound number detecting unit 121 (S705), and determining whether the number of effective sounds is less than the effective sound number threshold in the gain change determining unit 123. (S706) and making the gain invariant for all of the first gain control unit 109 to the Mth gain control unit 110 (S707).

  As described above, according to the second embodiment, compared with the first embodiment, it is possible to omit means necessary when using the audio conference with a large number of people, and thus the user can adjust the volume such as volume control. Audio conferencing can be used economically without performing any operation that hinders the user.

  In the above description, it is assumed that there are M audio information transmitting / receiving units, a signal level detection unit, a voice head portion presence / absence detection unit, and a gain control unit. However, when performing time-sharing processing, it is sufficient to provide one for each.

  It is also possible to perform software processing using a program for causing the above-described means to function.

  The voice mixing device of the present invention has an excellent effect of being able to hear the voices of all the speakers at an appropriate volume, and thus is useful in an intercommunication system or the like in a plane with high background noise such as an airplane. It is.

The block diagram which shows the structure of the audio | voice mixing apparatus in Embodiment 1 of this invention. Schematic diagram of the output of the m-th signal level detection unit in Embodiment 1 of the present invention Schematic diagram of output of m-th sound head portion presence / absence detection unit in Embodiment 1 of the present invention Schematic diagram of the output of the voice number detection unit in Embodiment 1 of the present invention Schematic diagram of output of gain change determination unit in Embodiment 1 of the present invention Schematic diagram of the output of the m-th gate part in the first embodiment of the present invention Flowchart of gain control of audio mixing apparatus in embodiment 1 of the present invention Audio state diagram for explaining the operation of the audio mixing apparatus according to Embodiment 1 of the present invention. The block diagram which shows the structure of the audio | voice mixing apparatus in Embodiment 2 of this invention. Flowchart of gain control of audio mixing apparatus in embodiment 2 of the present invention Block diagram showing the configuration of a conventional audio mixing device

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Network 101-102 1st-Mth audio | voice information transmission / reception part 103-104 1st-Mth signal level detection part 105-106 1st-Mth sound head part presence-and-absence detection part 107-108 1st-1st M-th gate unit 109 to 110 First to M-th gain control unit 120 Voice threshold 121 Voice number detection unit 122 Effective voice number threshold 123 Gain change determination unit 124 Voice information synthesis unit

Claims (4)

Voice information receiving means for receiving a plurality of voice information from outside;
A voice head part presence / absence detecting means for detecting the presence or absence of a voice head part in the output of each voice information of the voice information receiving means;
Gain control means for raising only the volume of voice information indicating that there is a voiced head part in the output of the voiced head part presence / absence detecting means, and not raising the volume of other voice information;
Voice information synthesis means for adding outputs for each voice information of the gain control means;
Voice information transmitting means for transmitting the output of the voice information synthesizing means to the outside;
An audio mixing device having A voice number detecting means for detecting the number of voices equal to or higher than a voice threshold in the output of each voice information of the voice information receiving means;
The gain control means is only for the volume of the voice information indicating that there is a voice head part in the output of the voice head part presence / absence detection means when the output of the voice number detection means is less than an effective voice number threshold. The sound mixing apparatus according to claim 1, wherein the sound volume of other sound information is not raised. Computer
Voice information receiving means for receiving a plurality of voice information from outside;
A voice head part presence / absence detecting means for detecting the presence or absence of a voice head part in the output of each voice information of the voice information receiving means;
Gain control means for raising only the volume of voice information indicating that there is a voiced head part in the output of the voiced head part presence / absence detecting means, and not raising the volume of other voice information;
Voice information synthesis means for adding outputs for each voice information of the gain control means;
A program for functioning as voice information transmitting means for transmitting the output of the voice information synthesizing means to the outside. A voice number detecting means for detecting the number of voices equal to or higher than a voice threshold in the output of each voice information of the voice information receiving means;
The gain control means is only for the volume of the voice information indicating that there is a voice head part in the output of the voice head part presence / absence detection means when the output of the voice number detection means is less than an effective voice number threshold. The program according to claim 3, wherein the volume of other audio information is not increased.

Images