JP3460783B2 - Voice switch for talker - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice switch used in a hands-free telephone or the like. In a hands-free telephone that employs a voice switch system, it is necessary to be able to accurately suppress acoustic echo.

[0002]

2. Description of the Related Art In order to realize a hands-free function,
It is necessary to increase the speaker volume and microphone sensitivity. However, in this case, as shown in FIG. 5, an acoustic echo occurs in which the received voice output from the voice output unit such as the speaker wraps around to the voice input unit such as the microphone. This is a phenomenon in which one's voice sounds like a echo to the other party of the call, which is extremely difficult to use. In order to remove this acoustic echo, there are two methods: (1) echo canceller method and (2) voice switch method.

The echo canceller system removes acoustic echoes using an adaptive signal processing technique. For example, in FIG.
As shown in (1), an acoustic echo r in which the received voice that is output wraps around the microphone is artificially generated inside the communication device and subtracted from the signal input to the microphone. The generation of the pseudo echo r'is the transfer function from the speaker to the microphone represented by the FIR filter. Since this transfer function changes depending on the surroundings of the telephone, the filter is adaptively changed so as to minimize the error between the pseudo echo r ′ and the acoustic echo r.

On the other hand, in the voice switch system, as shown in FIG. 7, the power of the speaker output voice is compared with the power of the microphone input voice, and one of them is suppressed to remove the acoustic echo. In other words, the signal input to the microphone during the speaker output has a high probability of being an acoustic echo.
During this period, the microphone input signal is suppressed to prevent the acoustic echo from being transmitted to the other party.

As described above, there are two methods of removing the acoustic echo, which is a problem in realizing the hands-free function, an echo canceller and a voice switch. The advantages and disadvantages of both are compared as shown in FIG. 8, and there is a trade-off between throughput and capacity. If cost is prioritized, the voice switch system will be adopted. The present invention relates to this voice switch.

FIG. 9 shows a detailed conventional structure of a hands-free telephone equipped with this voice switch. In FIG. 9, reference numeral 1 denotes a receiving unit that includes a demodulator or the like that receives a voice signal from the other party, and 2 denotes a power suppressing unit that can suppress and control the power of the received signal by changing the reception gain gain-r.
Reference numeral 3 is a voice output unit that includes an amplifier, a speaker, and the like and emits a received voice (R). Reference numeral 6 is a voice input unit that is composed of a microphone and an amplifier for inputting the transmitted voice (S), 7 is a power suppressing unit that can control the power of the received signal by changing the transmission gain gain-s, and 8 is the transmitted voice signal. Is a transmission unit including a modulator or the like for transmitting the signal to the other party.

Reference numeral 4 indicates whether or not the power suppressing section 2 on the receiving side suppresses the received voice based on the sizes of the received voice and the transmitted voice.
This is a determination unit that determines whether or not the transmitted voice is suppressed by the power suppressing unit 7 on the transmitting side. Reference numeral 41 is a power calculation unit that calculates the power of the reception signal received by the reception unit 1, and 42 is the current received voice state s based on the power calculated by the power calculation unit 41.
-sound detection unit that detects whether s is silent or voiced, 43 is a power calculation unit that calculates the power of the audio signal input to the audio input unit 6, and 44 is the current calculated based on the power calculated by the power calculation unit 43. A voice detecting section for detecting whether the transmitted voice state s-r is silent or voiced, and 45 indicates which side of the power suppressing sections 2 and 7 is in the suppression control state based on the detection results of the voice detecting sections 42 and 44. It is a determination unit that determines whether to set.

Here, the power calculators 41 and 43 calculate the power of the input voice data by the following formula. That is, assuming that the input voice data is x _i , the output power p _i is obtained by p _i = 10 × log [Σ (x _ij × x _ij )]. However, Σ is assumed to be an addition from j = 0 to J.

As shown in FIG. 10, the sound detecting units 42 and 44 are composed of a comparing unit for comparing the input power p _i with a constant threshold th, and the input power p is calculated by the following judgment formula.
_i is compared with a threshold th to determine whether the current voice state S _i is voiced or silent. Where s _i = 0 is silence,
s _i = 1 means voiced. The determination formula is if (p _i <th) s _i = 0 if (p _i > th) s _i = 1. This means that if the input power p _i is smaller than the threshold th, the voice state s _i is “0”, and if it is larger than the threshold th, the voice state s _i is “1”. This prevents background noise equal to or less than the threshold th from being erroneously determined to be voiced.

The determination unit 45 follows the determination logic table shown in FIG. 11 as an example, and the reception gain gain-r of the reception power suppression unit 2 and the transmission gain ga of the transmission power suppression unit 7 are shown.
It controls in-s. Here, the receiving gain gain-r
The transmission gain gain-s is in the range of 0.0≤gain≤1.0. In the judgment logic table of FIG. 11, when the transmission voice state s−s = 0 and the reception voice state s−r = 0, the transmission gain gain-s is “0.0” and the reception gain gain-r is Set to "0.0". When the transmission voice state s-s = 1 and the reception voice state s-r = 0, the transmission gain gain-s is set to "1.0" and the reception gain gain-r is set to "0.0". When the transmission voice state s-s = 0 and the reception voice state s-r = 1, the transmission gain gain-s is set to "0.0" and the reception gain gain-r is set to "1.0". When the transmission voice state s-s = 1 and the reception voice state s-r = 1, the reception gain is prioritized and the transmission gain gain-s is set to "0.
0 ", and the receiving gain gain-r is set to" 1.0 ". Control.

According to the judgment result of the judging section 45, the power is increased.
-The suppression units 2 and 7 are input voice data x _iAgainst
Output audio data x_iOutput as. x_i= X_i× gain

As described above, this voice switch system suppresses one of the received voice and the transmitted voice,
If the other is the received voice, it is output to the speaker, and if it is the transmitted voice, it is transmitted. When both of them have a voice, various criteria are conceivable, such as a case where the received voice is prioritized and a case where the voice power is higher is prioritized.

[0013]

As described above, the voice switch method is to remove the acoustic echo by comparing the states of the received voice and the transmitted voice and suppressing one of them and passing the other. is there. This usually eliminates the acoustic echo without any problem, but when frame processing is performed, a time lag occurs between the received voice and the transmitted voice, and the acoustic echo can be completely eliminated. It may not be possible.

The frame processing is used, for example, when performing voice coding / decoding processing, and means collectively processing data for a fixed time. FIG. 12 is a diagram for explaining this frame processing, and shows the timing until the input voice is encoded on the transmitting side, is decoded on the receiving side, and is output. In FIG. 12, on the transmission side, the transmitted voice input by the voice input unit is accumulated for a certain period of time to form a frame. At the timing of the second frame, the encoding process is performed by the encoding unit and the transmission process is performed by the transmission unit. Is sent and sent to the other party. On the other party's side, the frame is received by the receiving section and decoded by the decoding section, and then, at the timing of the third frame, the audio output section outputs the frame as the received voice. As can be seen from FIG. 12, the input voice on the transmitting side is output on the other side with a delay of at least 2 frames.

When this frame processing is performed by the voice switch, as shown in FIG. 13, the reception voice and the transmission voice to be compared at the time of determination by the determination unit are the voices input from the microphone at the same time when they are output from the speaker. It is no longer a voice and a transmitted voice. That is, considering the time point of the determination as a reference, the transmitted voice that is input to the microphone one frame before and the received voice that is output from the speaker one frame later are compared. Due to this deviation, if it is simply compared, it becomes impossible to determine whether or not the transmitted voice input by the voice input unit is an acoustic echo wrapping around from the voice output unit.

This will be described below in detail with reference to FIG. In FIG. 13, time is shown in the horizontal axis in units of frames. Hereinafter, description will be made according to the time in frame units.

First frame: The reception data received by the receiving unit on the receiving side is voiced, and the input data input by the voice input unit on the transmitting side is voiced. Second frame: The judging section compares the voice on the receiving side with the voice on the transmitting side. Since both parties have voice, we prioritize the reception,
It is determined that the voice on the transmitting side is suppressed. Third frame: According to the above determination, the output unit outputs the above-mentioned sound as "sound", suppresses the sound on the transmitting side and makes it "silent" and transmits. At this time, it is assumed that the received voice is interrupted and the reception data of the reception unit is silent because the other party has interrupted the conversation. At this point, speech was observed in the voice input section. However, it is not known whether this voice is a conversation of the local speaker or a voice echo of the voice 'sneaking around from the voice output unit. Fourth frame: The determination unit compares the silence on the receiving side with the voice on the transmitting side, and as a result, determines that the voice on the transmitting side is not suppressed. Fifth frame: According to the above determination, silence is output from the voice output unit as "silence" on the receiving side, and the voice from the voice input unit is transmitted on the transmitting side as "voiceless" without being suppressed.

In the above sequence, the voiced sound is transmitted to the other party at the 5th frame. However, whether the voiced sound is the conversation voice of the local speaker or the voice output from the voice output unit. I don't know if it was an acoustic echo of'sound '. For this reason, in the latter case, the acoustic echo, which originally had to be suppressed, will be transmitted to the other party, and the other party will hear the echo of their own voice when interrupting the talk. This makes it uncomfortable.

The present invention has been made in view of the above problems, and it is an object of the present invention to accurately suppress an acoustic echo regardless of a time shift of a transmitted / received signal in a communication device with frame processing.

[0020]

FIG. 1 is a diagram illustrating the principle of the present invention. In order to solve the above-mentioned problems, the voice switch of the telephone set according to the present invention has a delay unit that delays a received voice signal received by a predetermined time, and a received voice signal received and a transmitted voice signal input. Based on the received voice signal judged based on the received voice signal delayed by the delay means and the transmitted voice signal inputted, whether or not the transmitted voice signal is suppressed. And a receiving-side suppressing unit that suppresses the received voice signal according to the determination result of the receiving voice determination unit, and suppresses the transmitting voice signal according to the determination result of the transmitting voice determination unit. And a transmitting side suppressing means. In a telephone set to which this voice switch is applied, voice data is processed in block units by frame processing, and the predetermined time delayed by the delay means is a time lag between a reception voice signal and a transmission voice signal generated by the frame processing. Is the time to compensate for and synchronize the two. This delay means can be constituted by a storage means for temporarily storing. In this voice switch, when determining whether to suppress the received voice signal, the received voice is compared with the input transmitted voice. When determining whether or not to suppress the transmitted voice signal, the transmitted voice that has just been input is compared with the received voice that has already been output from the speaker at the same time as that transmitted voice, and the time lag is calculated. The received voice signal and the transmitted voice signal are synchronized at the time of determination. This is realized by delaying the received voice signal by delay means such as temporary storage means.

The above-mentioned voice switch has a transmission voice estimation means for estimating the state of the transmission voice signal synchronized with the reception voice signal, and the transmission voice signal from the transmission voice estimation means is received voice determination means. It can be configured to be input as a transmission voice signal.

[0022]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings. FIG. 2 shows a hands-free telephone equipped with a voice switch as an embodiment of the present invention. In the figure, the receiving unit 1, the power suppressing units 2 and 7, the voice output unit 3, the voice input unit 6, and the transmitting unit 8 are the same as the circuit elements described in the conventional device of FIG. I will omit the explanation.

Further, the apparatus of this embodiment has a decoding section 9 for decoding the received data and a coding section 10 for coding the transmission data. For example, when making a call by connecting personal computers to each other, it is sufficiently possible that the transmission capacity of the communication path is small, and thus it becomes necessary to perform encoding. Therefore, the apparatus of this embodiment has a hands-free function configuration using a voice switch and a voice codec.

Further, the apparatus of this embodiment has a received voice temporary storage section 48 and delays the received voice from the power suppressing section 2 by a predetermined frame time. This delay time is set to a length that can synchronize the sending voice and the receiving voice by correcting the time difference between them. In this embodiment, one frame of received voice data is delayed by two frames.

Further, instead of the determination unit 4 in the conventional device, in the present embodiment, the reception voice determination unit 46 and the transmission voice determination unit 47.
Has two. The received voice determination unit 46 determines whether or not the received power is suppressed by the power suppressing unit 2 on the receiving side, and the received voice output from the decoding unit 9 is input. The transmitted voice determining unit 47 determines whether or not the transmitted power is suppressed by the power suppressing unit 7 on the transmitting side. As the received voice, the received voice temporary storage unit 48 is used.
The data delayed by the predetermined frame time from is input. Since the received voice determination unit 46 and the transmitted voice determination unit 47 have the same configuration as the determination unit 4 described in the related art, detailed description thereof will be omitted.

The operation sequence of the apparatus of this embodiment will be described with reference to FIG. In FIG. 3, time is shown in the horizontal axis in units of frames. Hereinafter, description will be made according to the time in frame units.

First frame: The reception data received by the receiving section 1 on the receiving side is voiced, and the input data input by the input input section 6 on the transmitting side is voiced.

Second frame: The reception voice judging section 46 compares the voice on the receiving side with the voice on the transmitting side. Since both are voiced, it is determined that the voice priority is given to the voice reception and the voice output unit 3 outputs the voiced voice as the voiced 'without suppressing the voice.

Third frame: According to the above determination, the voice output section 3 outputs the voiced sound as a voiced sound 'without being suppressed by the power suppressing section. At this time, it is assumed that the other party has interrupted the conversation, so that the received voice is interrupted and the reception data of the receiving unit 1 becomes silent. At this point, the voice input unit 6
Sound is observed in. However, it is not yet known whether this voiced sound is a conversation of the local speaker or a voiced sound echo of the received voice circulated from the voice output unit 3.

Fourth frame: In the received voice judging section 46,
The silent side on the receiving side and the voiced side on the transmitting side are compared, and it is determined that the silent side on the receiving side is suppressed and output as "silent" from the voice output unit 3. In addition, the transmission voice determination unit 46 compares the voice on the reception side, which has been delayed by two frames in the reception voice temporary storage unit 48, with the voice on the transmission side, and determines the transmission voice according to the rule of reception priority. It is determined that the power is suppressed by the power suppression unit 7 on the transmitting side and transmitted as "silence".

Fifth frame: According to the above determination, the receiving side outputs silence as "silence" from the voice output section, and the transmitting side suppresses the sound from the input section 6 and transmits it as "silence".

In the above sequence, at the 5th frame,
The voice is transmitted as'silence '. In this case, the voiced sound input to the voice input unit 6 is simultaneously output from the voice output unit 3 at the same point.
It is highly possible that this is a sound echo of this voiced 'because the voiced' is output from the voice. However, the conventional device could not make the determination. On the other hand, according to the apparatus of this embodiment, the transmission voice determination unit 47 includes the reception voice temporary storage unit 4
Since the voiced sound delayed by 2 frames in 8 is compared with the voiced sound input to the voice input unit 6, the time difference between the voiced sound on the receiving side and the voiced sound on the transmitting side can be calculated. If they can be matched and if both are voiced, it can be determined that the voice on the transmitting side is likely to be an acoustic echo that the voiced 'output on the receiving side wraps around. It is determined that the voice is suppressed and the silence is transmitted. Thereby, the acoustic echo can be removed.

FIG. 4 shows another embodiment of the present invention.
Similar to the above-described embodiment, the frame delay is dealt with, but in order to eliminate the time lag between the received voice and the transmitted voice, which is also used when the received voice determination unit 46 determines the received voice, it is currently being determined. When the received voice is output, the transmission voice estimation unit 49 estimates the voice state of the transmission voice observed by the microphone of the voice input unit 6 at the same time. For estimation of the transmitted voice, a method of averaging the powers of the transmitted voices of the past several frames can be used. Since the speech changes relatively smoothly in time, the estimation accuracy is considered to be relatively high.

[0034]

As described above, according to the present invention,
It is possible to correct the time lag between the received voice and the transmitted voice at the time of determination, which is a problem of the voice switch using frame processing. This makes it possible to remove acoustic echoes and ensure quick followability to changes in received / transmitted voices, and provide a more natural call.

[Brief description of drawings]

FIG. 1 is a diagram illustrating the principle of the present invention.

FIG. 2 is a diagram showing a hands-free telephone equipped with a voice switch as one embodiment according to the present invention.

FIG. 3 is a diagram for explaining an operation sequence of the embodiment apparatus.

FIG. 4 is a diagram showing another embodiment of the present invention.

FIG. 5 is a diagram illustrating acoustic echo in a hands-free telephone or the like.

FIG. 6 is a diagram illustrating an echo canceller system.

FIG. 7 is a diagram illustrating a voice switch system.

FIG. 8 is a diagram comparing an echo canceller system and a voice switch system.

FIG. 9 is a diagram showing a hands-free telephone equipped with a conventional voice switch.

FIG. 10 is a diagram showing a configuration of a sound detecting unit in a conventional device.

FIG. 11 is a diagram showing an example of a sound / silence determination table.

FIG. 12 is a diagram illustrating delay by frame processing.

FIG. 13 is a diagram showing an operation sequence of a conventional device.

[Explanation of symbols]

1 receiver 2, 7 Power suppression unit 3 Audio output section 4 Judgment section 6 Voice input section 8 transmitter 41, 43 Power calculator 42, 44 voice detector 45 Judgment unit 46 Received voice judgment unit 47 Transmitted voice determination unit 48 Received voice temporary storage 49 Transmitted speech estimation unit

─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Masato Ito 4-1-1 Kamiotanaka, Nakahara-ku, Kawasaki-shi, Kanagawa Fujitsu Limited (56) Reference JP-A-5-75501 (JP, A) 6-78046 (JP, A) JP 7-202767 (JP, A) JP 8-274689 (JP, A) JP 5-75500 (JP, A) (58) Fields investigated (Int .Cl. ⁷ , DB name) H04M 1/60 H04B 3/23

Claims (2)

(57) [Claims] 1. A voice switch for a telephone with frame processing.
In, the delay means for delaying the received voice signal received by a predetermined time, the received voice determination means for determining whether to suppress the received voice signal based on the received voice signal received and the transmitted voice signal input, A transmission voice determination unit that determines whether or not to suppress the transmission voice signal based on the reception voice signal delayed by the delay unit and the input transmission voice signal, and the determination result of the reception voice determination unit wherein the receiving side suppressing means for suppressing the reception voice signal, and a transmitter-side suppression means for suppressing the transmission voice signal according to the determination result of the transmitted voice judging means, the delay time by the delay unit, the following By frame processing
The time lag between the received voice signal and the transmitted voice signal caused by
It is time to compensate and synchronize the two . The voice switch of the telephone. 2. A transmission voice estimation means for estimating a voice state of a transmission voice signal synchronized with the reception voice signal, wherein the transmission voice signal from the transmission voice estimation means is transmitted to the reception voice determination means. The voice switch for a telephone as claimed in claim 1, wherein the voice switch is configured to be inputted as a voice signal.