KR102288182B1 - Method and apparatus for speech privacy, and mobile terminal using the same - Google Patents

Abstract

The present invention relates to a voice privacy protection method, which comprises the following steps of: transmitting a masking sound to a first speaker; receiving a distorted voice signal including a user's voice signal input to a first microphone and a masking sound signal output from the first speaker; removing the masking sound signal from the distorted voice signal to recover a real voice signal; adjusting a gain of the masking sound signal on the basis of the recovered voice signal; receiving a masking sound signal that gain has been adjusted through a second microphone; filtering the signal received through the second microphone to generate an anti-sound signal for the masking sound signal; and outputting the anti-sound signal through a second speaker. According to the present invention, a masking sound for a user's voice signal is output, thereby achieving voice privacy protection. Feedback effect of the masking sound is removed, thereby transmitting a clear voice signal. In addition, the Lombard effect in which the user's voice becomes louder due to interference caused by the masking sound can be efficiently removed.

Description

Voice privacy protection method, voice privacy protection device, and mobile terminal using same

The present invention relates to a voice privacy protection method, a voice privacy protection device, and a mobile terminal using the same, and more particularly, a voice privacy protection method capable of generating a masking sound for voice privacy protection and controlling interference of the masking sound, It relates to a voice privacy protection device and a mobile terminal using the same.

Today, various portable electronic devices such as smart phones, tablets, personal digital assistants (PDAs), and digital voice recorders are being used. Using a portable electronic device, you can make a call with the other party regardless of time and place. However, there is a problem of privacy when making calls in places where there are many people around. That is, there is a very high possibility that the voice privacy of a user who makes a call is inadvertently violated by people around him or other listening devices. Since everyone does not want others to hear whether it is a private conversation or a business conversation that must be kept confidential, voice privacy in a call is very important to be protected, and a cost-effective solution for this needs to be provided.

One method for this is to remove the user's voice information using an Active Noise Control (ANC) technology. ANC is an electroacoustic technology based on the superposition principle, where an additional secondary source (a loudspeaker) generates noise with the same amplitude and opposite phase to cancel the noise of the original primary source. Although many studies have been conducted on this idea, it is difficult to provide a desired level of voice privacy protection due to physical limitations. In particular, most of the portable electronic devices are not suitable for real-time signal processing, which is the main requirement of ANC. This is because the general-purpose processor of the portable electronic device has a long audio latency that causes ANC performance degradation due to causal constraints. In this situation, listeners around the user can still understand the user's conversation content.

Furthermore, to overcome the above limitation, US Patent No. 6952474B2 proposes a design of an external device used with a portable electronic device to provide voice privacy. A spherically mounted speaker arrangement around a voice input microphone generates an anti-voice signal to cancel out the user's voice signal, and a spherically arranged far-field sensor around the speaker reduces how well the speaker reduces the transmission of voice space to the surrounding environment. sense that you are doing it. Nevertheless, ANC cannot provide complete privacy protection to users for the above reasons, and it is very inconvenient to have external hardware.

Another technology for protecting voice privacy is sound masking technology. It promotes voice privacy by adding digitally generated sound. Conventional systems using this technology are mainly used in open offices and the like to increase work comfort, but it is inefficient to provide voice privacy in most spaces. This is because it uses a fixed loudspeaker (eg the ceiling of a room) to generate the masking sound. The masking sound generated to protect the user's dialogue should be loud enough in the presence of unintended listeners. For this reason, the conventional technology using masking sound uses a large number of loudspeakers installed so that the system covers the entire area, so the installation cost is high, and the configuration becomes very complicated due to the difference in spatial architecture and the position of the listener. I have a problem.

Regarding the conventional attempt to use a sound masking technique in a portable device, US Patent Publication No. 2005-0065778A1 proposes a device for generating a masking sound by processing a voice signal together with an ANC. As disclosed in this patent, the speaker is designed to produce an acoustic field directed in the same direction as a human voice, except that the acoustic field is minimized toward the microphone. Thus, the other listener receives both the speaker's voice and the speaker's output, while the microphone receives only the speaker's voice. However, while the above technique may be correct in theory, in practice the feedback effect due to reverberation makes it impossible to isolate the microphone from picking up the output of the speaker. As a result, the user's voice signal that must be transmitted over the network is inevitably damaged. Also, this signal generated through the speaker interferes with the user's own calls, making it difficult to have a clear conversation.

It is clear that the feedback effect from the masking sound speaker goes into the voice mic, which corrupts the voice signal to be used later or transmitted to the other party. In addition, the user suffers from artificial ambient noise caused by the masking sound generation, and in this environment, the user's voice becomes louder during a call. This is called the Lombard Effect. According to the Lombard effect, when the voice of the caller becomes louder, a louder masking sound is generated, thereby creating a vicious cycle in which the voice of the caller becomes louder again. Therefore, it is necessary to find a way to minimize and actively control the interference of masking sound to protect voice privacy.

US Registration Publication No. 6952474B2 (registered on October 4, 2005) US Publication No. 2005-0065778A1 (published on March 24, 2005)

The present invention has been devised in view of the above problems, and an object of the present invention is to remove online feedback that removes the masking sound feedback effect for a clean voice signal, and to reduce the level of the masking sound over a wide frequency range. To provide a voice privacy protection method having a removal algorithm, a voice privacy protection device, and a mobile terminal using the same.

A voice privacy protection method according to the present invention for achieving the above object, transmitting a masking sound to a first speaker; receiving a distorted voice signal including a user's voice signal input to a first microphone and a masking sound signal output from a first speaker; recovering an actual voice signal by removing the masking sound signal from the distorted voice signal; adjusting a gain of the masking sound signal based on the recovered voice signal; receiving a masking sound signal whose gain has been adjusted through a second microphone; generating an anti-sound signal for the masking sound signal by filtering the signal received through the second microphone; and outputting the anti-sound signal through a second speaker.

In addition, the masking sound signal received through the second microphone may be output through the first speaker.

The generating of the anti-sound signal may further include correcting a signal received through the second microphone based on the position of the user's ear and the position of the second microphone.

On the other hand, the voice privacy protection device according to the present invention for achieving the above object, the first microphone for receiving a user's voice signal; a first speaker for outputting a masking sound signal having the same amplitude and opposite phase to the user's voice signal; a second microphone receiving the masking sound signal output from the first speaker; a second speaker for outputting an anti-sound signal for the masking sound signal output from the first speaker; and receiving a distorted voice signal including the user's voice signal and the masking sound signal, removing the masking sound signal from the distorted voice signal to recover an actual voice signal, and then based on the restored voice signal and a processor for adjusting a gain of the masking sound signal.

The processor may generate the anti-sound signal by correcting a signal received through the second microphone based on the position of the user's ear and the position of the second microphone.

On the other hand, the mobile terminal according to the present invention for achieving the above object, a first microphone for receiving a user's voice signal; a first speaker outputting a masking sound; a second speaker outputting a voice signal of the other party; a second microphone receiving the masking sound output from the first speaker; a communication unit for wirelessly transmitting and receiving the user's voice signal and the other party's voice signal; and a processor for generating a masking sound signal for the user's voice signal and outputting the masking sound signal through the first speaker.

In addition, the processor receives a distorted voice signal including the user's voice signal and the masking sound signal from the first microphone, and removes the masking sound signal from the distorted voice signal to recover an actual voice signal. Then, the gain of the masking sound signal may be adjusted based on the recovered voice signal.

In addition, the processor receives the masking sound signal output from the first speaker through the second microphone, filters the received signal to generate an anti-sound signal for the masking sound signal, and sets the second speaker It can be output through

Also, the processor may generate the anti-sound signal by correcting a signal received through the second microphone based on the position of the user's ear and the position of the second microphone.

In addition, the anti-sound signal may attenuate the masking sound signal input through the second microphone.

According to the voice privacy protection method, voice privacy protection device, and mobile terminal using the same according to the present invention, voice privacy protection is achieved by outputting a masking sound for a user voice signal, and a clear voice signal is obtained by removing the feedback effect of the masking sound. It is possible to effectively remove the Lombard Effect, in which the user's voice becomes louder due to interference caused by the masking sound.

1 is a perspective view and a schematic view of a voice privacy protection device and a mobile terminal according to the present invention.
2 is a schematic diagram showing the configuration of a voice privacy protection device according to the present invention.
3 is a schematic diagram showing the configuration of a voice privacy protection device according to the present invention.
4 is a schematic diagram for explaining the operation of a feedback remover, which is a component of the voice privacy protection device according to the present invention.
5 is a schematic diagram for explaining the operation of an active noise controller, which is one component of the voice privacy protection device according to the present invention.
6 is a diagram for explaining an error that may occur in a signal processing process of an active noise controller.
7 is a view in which a virtual sensing block for correcting an error that may occur in a signal processing process of an active noise controller is added.
Fig. 8 is a graph showing the cancellation effect of an anti-sound signal generated by an active noise controller.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS [0010] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS [0010] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS [0023] Reference is made to the accompanying drawings, which show by way of illustration specific embodiments in which the present invention may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the present invention. It should be understood that the various embodiments of the present invention are different but need not be mutually exclusive. For example, certain shapes, structures, and characteristics described herein with respect to one embodiment may be embodied in other embodiments without departing from the spirit and scope of the invention. In addition, it should be understood that the location or arrangement of individual components within each disclosed embodiment may be changed without departing from the spirit and scope of the present invention. Accordingly, the detailed description set forth below is not intended to be taken in a limiting sense, and the scope of the invention, if properly described, is limited only by the appended claims, along with all scope equivalents to those claimed. Like reference numerals in the drawings refer to the same or similar functions throughout the various aspects.

Hereinafter, the present invention will be described in detail with reference to the drawings. 1 is a perspective view and a schematic view of a voice privacy protection device and a mobile terminal according to the present invention. Although the voice privacy protection device and the mobile terminal are illustrated in FIG. 1 as a smart phone, the target article is not necessarily limited thereto.

1, the voice privacy protection device and the mobile terminal according to the present invention are provided with two speakers and two microphones.

The first microphone RM receives a user's speech as a reference microphone. The first speaker MS is a main speaker and, as will be described in more detail below, emits a masking sound that prevents the user's speech from spreading far. The first microphone RM and the first speaker MS are disposed adjacent to each other, and when implemented as a smartphone, they are positioned close to the user's mouth to receive a voice signal.

The second speaker ES is an earpiece speaker, and has a function of transmitting the voice of the other party, but in the present invention, the anti-sound is transmitted and output from the first speaker MS to reduce the masking sound for a call. It prevents the masking sound from being input to the user's ear and acting as noise. The second microphone EM is an error microphone, and receives a masking sound output from the first speaker MS on behalf of the user's ear. The signal becomes a basic signal for generating an anti-sound to be transmitted from the second speaker ES. The second microphone EM and the second speaker ES are disposed adjacent to each other, and when implemented as a smartphone, they are positioned close to the user's ear to replace the user's ear or attenuate the masking sound flowing to the user's ear.

2 is a flowchart showing the flow of a voice privacy protection method according to the present invention. In the flowchart of FIG. 2 , the subject receiving or transmitting each signal may be the first and second speakers or the first and second microphones described above. In addition, processing (conversion, filtering, correction, etc.) of the input signal may be performed by various components included in the processor.

In the voice privacy protection method according to the present invention, first, a masking sound signal is transmitted through the first speaker MS (S10. After that, the user's voice signal input to the first microphone RM and the first speaker MS ) receives a distorted voice signal including a masking sound signal output from (S20).

Here, the user's voice signal inputted to the first microphone RM itself should be a clean signal that does not include noise, but a masking sound signal output from the first speaker MS disposed adjacent to the first microphone RM. Since is directly input to the first microphone RM or indirectly reflected by the user's hand or an article placed in the surrounding environment, the signal input to the first microphone RM is the user's voice signal and the masking sound signal It is inevitably a distorted voice signal due to a mixture of

The masking sound signal may be previously recorded, and may be continuously updated by feeding back the user's voice signal.

Thereafter, the actual voice signal is restored by removing the masking sound signal from the distorted voice signal input to the first microphone RM ( S30 ). Then, the gain of the masking sound signal is adjusted based on the recovered voice signal (S40). The gain-adjusted masking sound signal is transmitted again through the first speaker MS to attenuate the user's speech so as not to spread to the surroundings. Steps S10 to S40 correspond to an online feedback cancellation step, and noise is continuously accumulated through a feedback method of recovering a damaged voice signal, generating a masking sound based on it, and adjusting the gain. This will prevent the loss of voice and ensure clear call quality and voice privacy.

When the masking sound signal enters the user's ear, it may act as noise, which makes the user's call voice louder, and it is fed back to cause a problem affecting the masking sound generation. The present invention solves this problem through the active noise cancellation step do.

The second microphone EM positioned adjacent to the user's ear receives the gain-adjusted masking sound signal (S50). Thereafter, processing (filtering) is performed on the signal received through the second microphone EM to generate an anti-sound signal for the masking sound signal (S60). The anti-sound signal is output through the second speaker ES (S70).

Since the anti-sound signal has an opposite phase to the masking sound signal, when the two signals are combined, the amplitude of the signal is attenuated, making it almost inaudible to the user's ear. In this way, the masking sound protects the user's voice privacy, but does not act as a noise to the user's ear, so that a high-quality call can be realized without raising the voice.

Meanwhile, the generation of the anti-sound signal may further perform a correction step. A position where the masking sound, which is a basic signal for generating an anti-sound signal, is directly received, is an area where the second microphone EM exists, but in reality, the basic signal must be a masking sound input to the user's ear. That is, the basic signal for generating the anti-sound signal may be different from the actual signal. Accordingly, the signal received through the second microphone EM should be corrected based on the position of the user's ear and the position of the second microphone EM.

Accordingly, the present invention may further include correcting the signal received through the second microphone EM based on the position of the user's ear and the position of the second microphone in generating the anti-sound signal.

On the other hand, the masking sound signal is a signal having the same amplitude and opposite phase as the user's voice signal input through the first microphone (RM), and the anti-sound signal is output from the first speaker (MS) to the second microphone (EM) It may be a signal having the same amplitude and opposite phase as that of the masking sound signal input to . However, the amplitude and phase difference of each signal may be set differently depending on the surrounding environment and system.

3 is a schematic diagram showing the configuration of a voice privacy protection device according to the present invention. In FIG. 3, a signal flow is shown by an arrow along with each configuration. The schematic diagram shown in FIG. 3 may be directly applied to the mobile terminal according to the present invention.

The voice privacy protection apparatus according to the present invention includes a first microphone RM, a first speaker MS, a second microphone EM, a second speaker ES, and a processor 100 . The processor 100 may include a sound masking unit 110 , a gain controller 120 , a feedback removing unit 130 , and an active noise controller 140 .

The first microphone RM is configured to receive a speech of a near-end user. However, the signal input to the first microphone RM may include a masking sound signal output from the first speaker MS disposed adjacent to the first microphone RM as well as the user's voice signal.

The sound masking unit 110 may include a memory for storing a masking sound signal and/or a signal processing unit for generating a masking sound signal. The masking sound signal transmitted from the sound masking 110 is transmitted through the first speaker MS to interfere with the user's (near-end user) voice signal, making it difficult for an outsider to recognize the user's voice. The masking sound signal may have an opposite phase to the user's voice signal, but is not limited thereto.

Although the main function of the first microphone RM is to receive the user's voice signal, as described above, the masking sound signal transmitted from the sound masking unit 110 and transmitted through the first speaker MS can also be input as noise. can That is, since the masking sound signal is directly input to the sound masking first microphone RM or is reflected in the user's hand or an article placed in the surrounding environment, the signal input to the first microphone RM is input to the user. Inevitably, it is a distorted voice signal by mixing the voice signal and the masking sound signal.

If the first microphone RM receives a distorted voice signal and the sound masking unit 110 generates a masking sound signal based thereon again, there is a problem in that the masking sound signal is gradually distorted.

The feedback remover 130 solves this problem. The feedback remover 130 receives a corrupted speech input through the first microphone RM, and removes the inputted masking sound signal from the received distorted signal. Through the above process of the feedback remover 130, the clean speech signal may be restored again.

Meanwhile, the feedback removing unit 130 generates an interference signal of the masking sound by filtering the masking sound signal, and a signal processing process of updating a feedback filter that adaptively filters a change of the interference signal over time. can be performed.

The gain controller 120 adjusts the gain of the masking sound signal based on the restored voice signal, and the gain-adjusted masking sound signal is again transmitted through the first speaker MS. At this time, the gain controller 120 adjusts the gain of the masking sound according to the amplitude of the restored voice signal in order to provide desired voice privacy. The gain-adjusted masking sound protects voice privacy by making it difficult for outsiders to recognize the user's voice.

By the feedback remover 130 , the damaged voice signal is recovered and a masking sound generated based thereon is generated, and then the gain is adjusted by adopting a feedback method to achieve clean call quality and protection of voice privacy.

The masking sound cancels out the user's voice and prevents it from spreading to the surroundings, thereby protecting privacy. This makes the user's call voice louder, which causes a problem in that it is fed back and affects the masking sound generation. The active noise controller 140, which is one component of the present invention, solves the above problem through anti-sound generation.

The active noise controller 140 transmits anti-sound to the masking sound signal transmitted from the first speaker MS through the second speaker ES, so that the masking sound is input to the user's ear during a call and is converted into noise. prevent it from working

First, the second microphone EM receives the masking sound signal output from the first speaker MS on behalf of the user's ear. The signal is transmitted to the active noise controller 140 , and the active noise controller 140 generates an anti-sound signal by filtering the corresponding signal. A filtering process of generating an anti-sound signal having an opposite phase and/or the same amplitude based on the masking sound signal input to the second microphone EM may use various conventional methods. That is, the present invention is not limited to a specific signal processing method, and various signal processing methods used in the conventionally disclosed sound masking technology may be selectively applied. Since the essence of the invention may be obscured here, a detailed description of signal processing will be omitted.

The anti-sound signal generated by the active noise controller 140 is output through the second speaker ES, which cancels the masking sound signal coming into the user's ear so that it is not transmitted to the user's ear.

4 is a schematic diagram for explaining the operation of a feedback remover, which is a component of the voice privacy protection device according to the present invention.

As shown in FIG. 4 , the masking sound signal transmitted from the sound masking unit 110 or transmitted through the gain controller 120 is transmitted to a direct or indirect path through the first speaker MS. 1 is input to the microphone RM, which distorts the user's voice signal input to the first microphone RM.

When the direct and/or indirect path is considered as one path, the adaptive algorithm 131 models this feedback path as an adaptive filter. This is referred to as a modeling path 132 . Filtering the masking sound signal through this modeling path and removing the masking sound signal transmitted through the direct and/or indirect path from the distorted voice signal can finally remove the feedback effect of the masking sound, A clear voice signal (a pure voice signal inputted from the user) can be used. The cleared voice signal may be transmitted to the other party of the call, or stored in a memory (not shown) and used later.

5 is a schematic diagram for explaining the operation of an active noise controller, which is one component of the voice privacy protection device according to the present invention. As described above, the main function of the active noise controller 140 is to block noise caused by the masking sound at the user's ear position.

As shown in FIG. 5 , the masking sound signal transmitted from the sound masking unit 110 or transmitted through the gain controller 120 is transmitted to a direct or indirect path through the first speaker MS. 2 Input to the microphone (EM). The masking sound signal input to the second microphone EM is input to the user's ear and acts as noise.

Although illustrated as a user's ear in FIG. 5 , it may actually correspond to the second microphone EM. The masking sound signal transmitted through the first speaker MS is directly or indirectly input to the user's ear, more specifically, the second microphone EM through the first path. At this time, the adaptive algorithm 141 of the active noise controller 140 uses the masking sound signal as a reference signal, and generates an anti-noise signal through the adaptive filtering 142 . do.

The anti-sound signal generated by the active noise controller 140 is input to the user's ear as a second path through the second speaker ES. Both signals received through the first path and the second path cancel each other out or lower the strength of the masking sound signal.

6 is a view for explaining an error that may occur in the signal processing process of the active noise controller, and FIG. 7 is a virtual sensing block for correcting an error that may occur in the signal processing process of the active noise controller. Added drawing.

As shown in FIG. 6 , when the first path and the second path are set based on the user's ear, there will be no error in generating the anti-sound, but in reality, there is a gap between the user's ear position and the second microphone EM. There is a slight displacement. Accordingly, in order to estimate the error signal on the ear side from the collected error signal, a signal processing technique such as a virtual sensing technique may be implemented. The error signal at the user's ear transmits the error signal to the adaptive algorithm to track the change of the adaptive filter, and improves the reliability of the anti-sound signal by correcting the error using virtual sensing technology.

FIG. 7 shows that a virtual sensing block is further added to the configuration shown in FIG. 5 . x(n) is a masking sound signal transmitted from the first speaker MS, d _p (n) is an interference signal from an actual microphone (the second microphone EM), and d _v (n) is the interference signal from the virtual sensor VS (located in the user's ear), e _p (n) is the error signal from the real microphone, and e _v (n) is the error signal from the virtual sensor VS signal, W(z) is an adaptive filter, LMS is an adaptive algorithm, and y(n) is a control signal that has passed through the adaptive filter and the adaptive algorithm.

Since the point where noise reduction is desired is located in the virtual sensor VS, not the point where the second microphone EM of FIG. 6 is located, the secondary path is corrected by the virtual sensing block indicated by the dotted line of FIG. A more precise control is possible only when the and plant calibration is performed. Of course, in an embodiment, this correction process may be omitted.

Here, the second path refers to a path between the second speaker ES and the control point (the location of the virtual sensor VS), and the plant refers to a path between the masking sound signal, which is a noise source, and the control point. means path.

In the virtual sensing block, signal processing for correcting an error is performed. First, a modeling signal is generated by performing secondary path modeling on a control signal input to an actual microphone, and after adding the modeling signal and an error signal from the actual microphone (EM), the observation plant model (observation) It is filtered through the plant model). The filtered signal is fed back to the modeling signal obtained by performing secondary path modeling on the control signal input to the virtual sensor VS, and then goes through the adaptive algorithm step again.

)를 통하여 제어 신호를 제거함으로써, 제2 마이크(EM)에 도달하는 마스킹 사운드 신호를 복원하고(

), 측정한 가상 센서(VS)의 제2 경로(

) 및 '제2 마이크(EM)와 가상 센서(VS) 간의 음향 경로(H(z))를 통하여 사용자의 귀에 위치한다고 상정된 가상 센서(VS)에서의 에러 신호(error signal)(e_v(n))를 얻게 된다. 이를 통하여, 기존의 제2 마이크(EM)에서 측정되는 신호로 가상 센서(VS)에서의 에러 신호(e_v(n))를 구할 수 있고, 가상 센서(VS)의 위치에서의 소음 저감이 가능해진다.More specifically, the second path (

By removing the control signal through ), the masking sound signal reaching the second microphone EM is restored (

), the second path of the measured virtual sensor (VS) (

) and 'error signal (e _v () n)) is obtained. _{Through this, an error signal e v} (n) from the virtual sensor VS can be obtained with a signal measured by the existing second microphone EM, and noise reduction at the location of the virtual sensor VS is possible becomes

That is, noise reduction at the position of the virtual sensor VS (sensor assumed at the user's ear position) is possible with the signal measured by the second microphone EM through the second path and plant correction.

The adaptive algorithm, modeling and filtering functions described above may be defined based on the position of the real microphone (the second microphone EM) and the position of the virtual sensor VS. That is, based on the position of the real microphone (the second microphone EM) and the position of the virtual sensor VS, the signal (masking sound signal) received through the second microphone is corrected and filtered to generate an anti-sound signal will do

Fig. 8 is a graph showing the cancellation effect of an anti-sound signal generated by an active noise controller. In each graph of FIG. 8 , the horizontal axis indicates frequency (Hz), and the vertical axis indicates magnitude (dB).

In each graph, the signal shown above represents the signal without using the active noise controller (w/o ANC), and the signal shown below in the graph represents the signal using the active noise controller (w/ ANC). And, each signal was measured at time intervals of 1 sec, 3 sec, 6 sec, and 10 sec.

As shown in FIG. 8 , it can be seen that when the active noise controller is used, the magnitude of the signal is significantly reduced compared to the case where the active noise controller is not used. This means that the noise entering the user's (near-end user) ear, more specifically, the noise level due to the masking sound output from the first speaker MS is reduced, which is the anti-sound generated by the active noise controller. This is because the signal is transmitted from the second speaker ES.

Meanwhile, the mobile terminal according to the present invention includes a first microphone (RM), a first speaker (MS), a second microphone (EM), a second speaker (ES), a communication unit (not shown), and a processor.

Matters related to the first microphone RM, the first speaker MS, the second microphone EM, the second speaker ES, and the processor are the same as described above, and thus will be omitted.

The communication unit (not shown) has a function of transmitting and receiving signals through wireless communication. In addition to communication methods such as the Internet, WiBro, and Wi-Fi, short-range communication methods such as Bluetooth, Near Field Communication (NFC), Zigbee, UWB, and RFID may be used, but the present invention is not limited thereto. In some cases, a wired communication method may be used.

A voice privacy protection device and a mobile terminal are portable electronic devices, such as a smart watch, various wearable devices, a desktop, a tablet PC, a laptop, and a personal digital (PDA). assistants), voice recorders, cameras with communication functions, camcorders, electronic dictionaries, smart TVs, and the like.

According to the voice privacy protection method, voice privacy protection device, and mobile terminal using the same according to the present invention, voice privacy protection is achieved by outputting a masking sound for a user voice signal, and a clear voice signal is obtained by removing the feedback effect of the masking sound. It is possible to effectively remove the Lombard Effect, in which the user's voice becomes louder due to interference caused by the masking sound.

Features, structures, effects, etc. described in the above embodiments are included in one embodiment of the present invention, and are not necessarily limited to one embodiment. Furthermore, features, structures, effects, etc. illustrated in each embodiment can be combined or modified for other embodiments by those of ordinary skill in the art to which the embodiments belong. Accordingly, the contents related to such combinations and modifications should be interpreted as being included in the scope of the present invention.

In addition, although the embodiment has been described above, it is merely an example and does not limit the present invention, and those of ordinary skill in the art to which the present invention pertains are exemplified above in a range that does not depart from the essential characteristics of the present embodiment. It can be seen that various modifications and applications that have not been made are possible. For example, each component specifically shown in the embodiment can be implemented by modification. And differences related to such modifications and applications should be construed as being included in the scope of the present invention defined in the appended claims.

MS: first speaker
RM: first microphone
ES: second speaker
EM: second microphone
100: processor
110: sound masking unit
120: gain controller
130: feedback removal unit
140: active noise controller

Claims (10)

transmitting a masking sound to a first speaker;
receiving a distorted voice signal including a user's voice signal input to a first microphone and a masking sound signal output from the first speaker;
recovering an actual voice signal by removing the masking sound signal from the distorted voice signal;
adjusting a gain of the masking sound signal based on the recovered voice signal;
receiving a masking sound signal whose gain has been adjusted through a second microphone;
generating an anti-sound signal for the masking sound signal by filtering the signal received through the second microphone; and
outputting the anti-sound signal through a second speaker;
The masking sound signal received through the second microphone is outputted through the first speaker. delete According to claim 1,
The generating of the anti-sound signal comprises:
Compensating the signal received through the second microphone based on the position of the user's ear and the position of the second microphone; Voice privacy protection method further comprising a. a first microphone for receiving a user's voice signal;
a first speaker for outputting a masking sound signal for the user's voice signal;
a second microphone receiving the masking sound signal output from the first speaker;
a second speaker for outputting an anti-sound signal for the masking sound signal output from the first speaker; and
Receive a distorted voice signal including the user's voice signal and the masking sound signal, remove the masking sound signal from the distorted voice signal to recover an actual voice signal, and then, based on the restored voice signal, a processor for adjusting a gain of the masking sound signal;
The second microphone receives a masking sound signal whose gain is adjusted by the processor,
The second speaker is a voice privacy protection device for outputting an anti-sound signal in which the gain-adjusted masking sound signal received through the second microphone is filtered. 5. The method of claim 4,
The processor is configured to correct the signal received through the second microphone based on the position of the user's ear and the position of the second microphone to generate the anti-sound signal. a first microphone for receiving a user's voice signal;
a first speaker outputting a masking sound;
a second speaker outputting a voice signal of the other party;
a second microphone receiving the masking sound output from the first speaker;
a communication unit for wirelessly transmitting and receiving the user's voice signal and the other party's voice signal; and
a processor for generating a masking sound signal for the user's voice signal and outputting it through the first speaker;
The processor receives a distorted voice signal including the user's voice signal and the masking sound signal from the first microphone, removes the masking sound signal from the distorted voice signal, and restores an actual voice signal. A mobile terminal for adjusting a gain of the masking sound signal based on the recovered voice signal. delete 7. The method of claim 6,
The processor receives the masking sound signal output from the first speaker through the second microphone, filters the received signal to generate an anti-sound signal for the masking sound signal, and outputs it through the second speaker A mobile terminal that makes 9. The method of claim 8,
The processor is configured to correct the signal received through the second microphone based on the position of the user's ear and the position of the second microphone to generate the anti-sound signal. 9. The method of claim 8,
The anti-sound signal is a signal for attenuating the masking sound signal input through the second microphone.

Images