JPH07303148A - Communication conference equipment - Google Patents

Abstract

PURPOSE:To attain the communication of quality equal to that in a same space conversation by implementing signal processing providing effective sound image localization to a recipient in a short processing time in a voice communication among plural points. CONSTITUTION:A call information detection section 2 detects a call start signal and a call end signal based on a voice signal picked up by a microphone 1. A line decision section 3 decides a line through which a voice signal whose localization is to be processed or a line through which the voice signal is added based on call information. A localization processing adder section 4 synthesizes left and right channels of voice signals to provide a sense of localization to a prescribed sound image position and adds the voice signal sent from other point to the voice signal subject to localization processing. Through the constitution above, n-sets of localization processing sections making sound image localization signal processing whose localization number is (n) are installed in parallel and a sense of localization is provided to the different voice signal among n-points at maximum.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a communication conference apparatus, and more particularly, when a voice is reproduced in voice communication connecting a plurality of points, the voice of a talker who is a communication partner is specified by a listener by a sound image localization technique. The present invention relates to a communication conference device that localizes and listens to a position.

[0002]

2. Description of the Related Art In voice communication between a plurality of points which are separated from each other, a listener only needs to add a voice signal of a communication partner and listen to it to listen to a talker who is a communication partner at many points. Could not be identified at the same time,
It is difficult to understand clearly what happened.

In order to solve this problem, it has been attempted to perform signal processing on the voice signals transmitted from a plurality of points to localize a plurality of sound images at different positions. Human beings perceive the position and direction of the sound source by listening to the sound emitted from the sound source, that is, they can obtain a sense of localization. When sound is reproduced by a sound source such as a speaker or an earphone, the sound can be localized in a spatial direction different from those of the sound source by performing signal processing on the sound signal. Then, when multiple voices are presented to the listener at the same time, the intelligibility of these voices is improved without increasing the volume by localizing each voice in mutually different spatial directions, and speaker identification is performed. It is also known that it can be facilitated.

An outline of the sound image localization technique will be described with reference to FIGS. 1 to 3. FIG. 1 shows a conceptual diagram of a sound image localization technique by convolution of a transfer function. FIG. 1 (a) is a diagram showing a relationship between a sound source and a listener, and FIG. 1 (b) is a sound source and a listener even when using a sound source 1 and a sound source 2 installed at different positions. 1 is reproduced at the listening position by controlling the transfer function between the convolutions with the acoustic signal and reproducing the acoustic signal from the real sound source at the listening position.
The sound image can be localized at the position of the sound source in (a).

FIG. 2 is a diagram for explaining control of the sound image localization position by controlling the level difference. FIG. 2 (a) is a conceptual diagram when a sound source A and a sound source B, which are two sound sources on the left and right in front of the listener, are used and reproduced by giving a level difference to the same signal, and FIG. 2 (b) is a sound source. It is a figure which shows the relationship between the relative signal strength with respect to each of A and the sound source B, and the localization position of the sound image of a listener.

FIG. 3 is a diagram for explaining an example of controlling the sound image localization position by controlling the time difference. FIG. 3A is a diagram showing a conceptual diagram in the case of using the sound source A and the sound source B, which are the two sound sources on the left and right in front of the listener, and reproducing the same signal with a time difference. FIG. 3B is a diagram showing the relationship between the left-right time difference between the sound source A and the sound source B and the localization position of the sound image of the listener. When B precedes, (B-A) is a negative value.

As described above, in order to control the localization position of the sound image, the transfer function between the sound source and the listening position is convoluted into the acoustic signal as shown in FIG. 1 or plural as shown in FIG. 2 or FIG. There is a method of controlling the level difference or time difference between sound sources. However, a large amount of calculation is required to perform such signal processing, and the parameters required for signal processing increase depending on the localization position of the sound image. In the case of giving a localization feeling to different spatial directions, it is considered that a localization process is performed in advance by setting parameters that bring about different localization directions. And
The signal processing time must be shorter than several tens of ms which is a delay time that allows natural communication between a plurality of points separated from each other.

In order to do so, it is necessary to perform the sound image localization signal processing, that is, the localization processing on the audio signals transmitted from the respective points in parallel in one step by using an element dedicated to the signal processing. Then, in voice communication with a plurality of communication partners, in order to realize localization feeling in different spatial directions, it is realistic to perform localization processing using different signal processing elements in each spatial direction. Is. here,
In the voice communication between multiple points, when the voices transmitted from each point are spoken at the same time, the delay time can be reduced by inputting the voice signals transmitted from each point to each different signal processing element. Thus, localization feelings in different directions can be added to the voice transmitted from each point. An audio signal processing circuit element having a short delay time on the order of several tens of ms or less has already been developed.

On the other hand, when a plurality of speakers are talking at the same time without using a communication device in the same space, the listener can identify only up to about 4 to 5 speakers by using only the voice as a clue. It is known. And
In a normal dialogue, it is rare that a large number of speakers are speaking at the same time as compared with a state where only one speaker is speaking, and in reality, speaking is rarely started at the same time. It has also been investigated that the sound immediately after the start of speech greatly contributes to sound image localization and speaker identification.

[0010]

By the way, when the number of communicating parties during simultaneous utterance is larger than the number of localization processing units equipped in the receiving circuit device, a voice signal transmitted from a certain point where the last utterance is started. By prioritizing the sound image localization process and adding the sound signal transmitted from other points to the sound signal subjected to the sound image localization process, the sound can be reproduced without impairing the localization feeling and speaker identification. It is considered. However, if the voice signals are simply added and played back for a long time, the listener forgets the features of the voice of each speaker over time, and it becomes difficult to identify the speaker.

The present invention is based on the above-mentioned human perception,
Utilizing the features of dialogue, in voice communication between multiple points,
Signal processing that provides effective sound image localization to the listener with a short processing time facilitates speaker identification by the listener, and the naturalness is less likely to be impaired by delays caused by acoustic signal processing. It is intended to provide a communication conference apparatus for performing communication with a quality equivalent to that of dialogue in a space.

[0012]

Means for Solving the Problems A transmitter circuit device (1) for recording a voice signal and transmitting it to another point, a receiver circuit device (3, 4, 5) for reproducing a voice signal transmitted from another point, In a communication conferencing apparatus comprising a transmission line (6) having a transmission channel for physically or logically connecting different m points, a localization processing unit (41) for performing sound image localization processing on a transmitted audio signal, and A transmission for transmitting a voice signal of a communication partner, which is provided with n sets of localization processing addition units (4) in parallel, each of which includes an addition unit (42) for adding a voice signal of a communication partner who is uttering a voice signal subjected to sound image localization processing. A communication conferencing apparatus is provided with a line determining unit (3) for switching the connection between the line (6) and the n sets of localization processing units (41) and adding units (42).

The line determining unit (3) constitutes a communication conference apparatus which is controlled based on the utterance information indicating the utterance start and utterance end of the communication partner. Further, the transmitting circuit device is provided with the utterance information detecting section (2), and the communication conferencing device configured to transmit the detected utterance information to the receiving circuit device via the transmission line (6). Furthermore, a communication conferencing apparatus that prioritizes the sound image localization process on the last voice signal of the voice signals being uttered and adds the other voice signals to the voice signal subjected to the sound image localization process is provided. Configured.

Further, a communication conferencing apparatus for resuming the sound image localization processing by prioritizing the voice signal transmitted from the point where the voice signal whose addition has been started first is uttered is also configured.

[0015]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT An embodiment of the present invention will be specifically described with reference to FIGS. First, referring to FIG. 5, utterance information is detected in the transmitting circuit device at each point, the detected utterance information is transmitted to the receiving circuit device at another point, and based on the utterance information transmitted in the receiving circuit device. A line determining unit for determining a line transmitting a voice signal to be subjected to localization processing and a line transmitting a voice signal to be added is provided, and a set of localization processing unit and addition unit is provided to transmit from another point. 1 is an example of a communication conference apparatus of the present invention that reproduces a voice signal by a binaural earphone. This communication conferencing device comprises a transmitter circuit device, a receiver circuit device and a transmission line 6. Reference numeral 1 is a microphone, 2 is an utterance information detection unit, and the microphone 1 and the utterance information detection unit 2 constitute a transmission circuit device. The utterance information detection unit 2 is provided with the microphone 1
The utterance start signal and utterance end signal, which are utterance information, are detected based on the voice signal recorded by. Reference numeral 3 is a line determination unit, and 4 is a localization processing addition unit.
And the localization processing addition unit 4 constitutes a reception circuit device. The line determining unit 3 is a unit that determines a line that has transmitted a voice signal to be localized or a line to be added, based on the utterance information. The localization processing addition unit 4
This is a unit for synthesizing audio signals that give a feeling of localization to a predetermined sound image position to the audio signals for the left and right two channels, and adding the audio signals transmitted from other points to the audio signals subjected to localization processing.

The details of the localization processing addition section 4 are shown in FIG. n sets (2n left and right in the case of listening to earphones) of localization processing units 41 _L and 41 _R , n sets of addition units 42 _L and 42 _R correspondingly connected to the subsequent stages of the localization processing units, and audio transmitted from each point Transfers audio signals to terminals T ₁ and T _{2 for} transferring signals, and localization processing unit and addition unit 42 (n + 1)
It is configured by a line determining unit 41 including individual terminals t ₁ , t ₂ , and t ₃ . The voice signal is transferred between the respective functions according to the route shown by the solid line with an arrow and the utterance information according to the broken line with an arrow. Reference numeral 5 denotes a left and right two-channel earphone. 5 and 6 show the case where n = 1.

To switch between the localization process and the addition process in the communication conference apparatus of FIG. 5, utterance information indicating the utterance start and utterance end of the speaker at each point is required.
This is detected and transmitted by the utterance information detector 2. In addition,
In a normal conversation, the amount of utterance information is on the order of several tens of b / s or less, which is significantly smaller than the amount of voice information on the order of several tens of kb / s · channel. The increase in is almost negligible. Then, the line determining unit 3 is controlled by the utterance information detected and transmitted by the utterance information detecting unit 2, and the connection between the voice signal line connected to each point and the localization processing adding unit 4 is exchanged. That is, since the utterance information is independent for each point, the utterance start and utterance end are detected by the utterance circuit device at each point and the utterance information detected is transmitted to the receiver circuit device at another point, The receiving circuit device is provided with a line determining unit 3 which receives the speech information transmitted from each transmitting circuit device and determines a line for transmitting a voice signal to be localized and a line for transmitting a voice signal to be added. . Further, the connection between the line transmitting the audio signal and the localization processing unit and the addition unit may be not only a physical connection but also a logical connection.

Here, FIG. 4 shows a situation in which the sounds at the points A and B are heard at the point C in the case of a dialogue between the points A, B, and C. Note that the number n of localization processing functions is one. FIG. 4 assumes, as an example, a situation in which utterance starts first at point A, then utterance starts and ends at point B, and finally utters at point A. First, when speech is started at point A, localization processing is performed on the voice recorded at point A until speech is started at point B. Next, from the start to the end of utterance at the point B, the sound signal recorded at the point B is subjected to localization processing, and the sound image recorded at the point A is interrupted by the sound image localization signal processing. It is added to the audio signal recorded at the point and subjected to localization processing. When the utterance ends at the point B, the localization processing is performed again on the audio signal recorded at the point A. The communication conference apparatus of the present invention shown in FIG. 5 carries out the above-described processing.

FIG. 7 is a flow chart for explaining the operation of the line deciding unit 3 for deciding the line for transmitting the voice signal to be localized and the line for transmitting the voice signal to be added.
After making a call, the location of the other party connected for communication is confirmed. Next, position information and parameters necessary for controlling the localization processing unit 41 or the addition unit 42 are set. Communication is started after parameter setting is completed. Immediately after the utterance start signal is received, the line is connected to the localization processing unit 41 or the localization processing unit 41 is switched to the addition unit 42 when the utterance starts. Immediately after receiving the utterance end signal, the line connection to the localization processing unit 41 is disconnected or the addition unit 42 is switched to the localization processing unit 41 when the utterance ends. During communication, the utterance start and utterance end signals are placed on standby until the call ends. As an example, localization priority p is introduced as a control parameter for determining a line for transmitting a voice signal to be localized or added, and prio is introduced for a line from each point, and p is obtained by performing the following operation. The value of is set to indicate the number of lines currently connected to the adder, and the prio is set to indicate the time sequence in which the lines connected to the adder are started to be connected.

FIG. 8 shows a connection or switching procedure of a line at the start of speech at another point. First, the localization processing unit 41 sets a line for transmitting a voice signal from a point corresponding to the utterance start signal.
Connect to. When the line connecting the other points is not already connected to the localization processing unit 41, the process proceeds to the next operation as it is, and waits for reception of the utterance start or end signal. When the other party's line is already connected to the localization processing unit 41,
1 is added to the value of p, the line already connected to the localization processing addition unit 4 is connected to the addition unit 42, and the value of p is added to the value of prio corresponding to the line newly connected to the addition unit 42. give. After that, the operation proceeds to the next operation.

FIG. 9 shows a procedure for disconnecting or switching the line when the utterance ends. First, the connection between the line corresponding to the point where the utterance end signal is received and the localization processing unit 41 or the addition unit 42 is disconnected. When the disconnected line is connected to the localization processing unit 41 and the value of the corresponding prio is 0, the process directly proceeds to the next operation. If the value of p is greater than 0, p
The line with rio = 1 is connected to the localization processing unit. Pri corresponding to all the lines still connected to the adding unit 42
Subtract 1 from the value of o. Further, 1 is subtracted from the current value of p. If the disconnected line is connected to the addition unit 42, the prio value larger than the corresponding prio value from the disconnected partner (however, the one corresponding to the line still connected to the addition unit) Subtract 1 from. Furthermore,
Subtract 1 from the value of p.

FIG. 10 shows the lines corresponding to the respective parties and the connection statuses to the localization processing section 41 and the addition section 42, the prio values corresponding to the respective parties, and p in the utterance and signal processing situations shown in FIG. Indicates a value. FIG. 10A shows a state in which the line for transmitting the audio signal at the point A is connected to the localization processing unit 41, and the localization processing 41 is performed on the audio signal transmitted from the point A. FIG. 10B shows a state in which the line transmitting the audio signal at the point A is connected to the adding unit 42 and the line transmitting from the point B is connected to the localization processing unit 41, and the audio signal transmitted from the point B is shown. Is localized and A
The audio signal transmitted from the point is added to the audio signal transmitted from the point B, which has been subjected to the localization processing.

[0023]

As described above, according to the present invention, in the voice communication between a plurality of m points, the sound image localization signal processing of the localization direction number n is performed with a short delay time as shown in FIGS. 5 and 6, for example. It has a line determination unit that installs a set of localization processing units in parallel and exchanges a line with a voice line, and provides a maximum of n localization feelings in different directions for a plurality of voices transmitted from distant points.
It can be given in communication between points.

When the number k of people who are uttering at the same time exceeds the number n of localization directions, localization processing is performed to give a feeling of localization preferentially to the speech uttered at the nearest m point, and other The number of localization processing units can be reduced by simply adding the uttered voices to the (k−n) opponents. Also, when the utterance that has undergone localization processing ends, by prioritizing the localization processing for the voice signal that was first started to be added, the listener easily captures the characteristics of the voice again by sound image localization. And the speaker identification can be restored.

[Brief description of drawings]

FIG. 1 is a diagram showing a concept of a sound image localization technique by convolution of a transfer function.

FIG. 2 is a diagram showing a concept of a sound image localization technique based on a level difference.

FIG. 3 is a diagram showing a concept of a sound image localization technique based on a time difference.

FIG. 4 is a diagram showing processing to be performed on a transmitted audio signal.

FIG. 5 is a diagram illustrating an example.

FIG. 6 is a diagram illustrating a localization processing addition unit.

FIG. 7 is a flowchart illustrating the operation of the line determining unit.

FIG. 8 is a diagram for explaining an operation accompanying the start of speech in FIG. 7.

FIG. 9 is a diagram for explaining an operation accompanying the end of utterance in FIG. 7.

FIG. 10 is a diagram showing line connection statuses and p and prio values in a localization processing unit and an addition unit.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Microphone 2 Speech information detection unit 3 Line determination unit 4 Localization processing addition unit 41 Localization processing unit 42 Addition unit 5 Earphone

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Office reference number FI technical display location H04S 1/00 K

Claims (5)

[Claims] 1. A transmitter circuit device for recording a voice signal and transmitting it to another point, a receiver circuit device for reproducing a voice signal transmitted from another point, and a plurality of different m points physically or logically. In a communication conference apparatus including a transmission line having a transmission channel to be connected, a localization processing unit that performs a sound image localization process on a transmitted audio signal and an addition that adds a voice signal by a communication partner who is speaking to the sound image localization processed voice signal. And a line determining unit for switching the connection between the transmission line for transmitting the voice signal of the communication partner and the n sets of the localization processing unit and the adding unit. Characteristic teleconference equipment. 2. The communication conferencing apparatus according to claim 1, wherein the line determining unit is controlled based on utterance information indicating utterance start and utterance end of the communication partner. . 3. The communication conference apparatus according to claim 2, wherein the transmission circuit device includes an utterance information detection unit, and transmits the detected utterance information to the reception circuit device via a transmission line. Teleconference equipment to be. 4. The communication conference apparatus according to any one of claims 1 to 3, wherein the sound image localization processing is prioritized to the last voice signal of the voice signals being uttered which is started to be uttered. The communication conferencing apparatus is characterized in that the other audio signals are added to the audio signals subjected to the sound image localization processing. 5. The communication conferencing apparatus according to claim 4, wherein the sound image localization process is restarted by giving priority to the audio signal transmitted from the point where the audio signal first started to be added is uttered. Teleconference equipment to do.