TWI777265B - Directivity sound source capturing device and method thereof - Google Patents

Abstract

The present invention relates a directivity sound source capturing device which primarily includes a radio module, a processor, a selector, and a speaker. The radio module receives a sound signal in a range which is directed through a beam, and the processor generates a receiving sound source through amplifying a sound feature point audio in a sound source corresponding to a receiving target which is selected by the selector and generates a plurality of adjusted sound sources through reducing or shielding the sound feature point audio in the other sound sources. Then the processor combines the receiving sound source and the plurality of adjusted sound sources to generate an output sound signal. Accordingly, the directivity hearing-aid device of the present invention provides selectively receiving the receiving sound source and reducing or shielding the other impurity sounds to facilitate the user hears audio from a specific object thereby.

Description

Pointing sound source detection device and method thereof

The present invention relates to a technology of sound source detection, in particular to a directed sound source detection device and method thereof for retrieving audio information of a specific target.

In a noisy environment, since the surrounding high-decibel ambient sound or other human voices easily overwhelm the sound of a specific object, it is difficult for the user to hear the sound of the specific object in the environment. For example, students are too noisy in class to hear the teacher's voice in class, or the wind speed in nature is too high, which in turn drives rivers or leaves to make sounds, making users unable to hear birds in the natural environment.

In this case, the prior art usually utilizes a sound-receiving cover to be mounted on a sound-receiving device, so that the sound-receiving device can focus on receiving the sound in the sound-receiving direction through physical confinement.

However, this method cannot effectively discriminate the sound of a specific object, and also cannot effectively discriminate and listen to the sound of a specific object when a plurality of high-decibel sounds are received in the sound collection direction. In addition, through the method of physical constriction, the wind speed of the external environment is easy to be too large, so that the sound collecting cover receives the sound produced when the wind is guided on the sound collecting cover, thereby affecting the judgment of sound collection.

Therefore, there is an urgent need in the current technology for a method that eliminates the physical constriction of the sound range, and can effectively listen to the sound of a specific object, so as to improve the problems existing in the prior art.

The purpose of the present invention is to provide a directed sound source detection device, which mainly uses sound waves to detect the sound on the beam range, and captures the sound of a specific object (detecting the sound source) according to related processing, and then reduces or shields other sounds. The user can effectively listen to the voice of a specific object, thereby effectively improving the problems of the prior art.

In order to achieve the purpose disclosed above, the present invention provides a pointing sound source detection device, which includes: a sound receiving module, which receives a sound signal within the range pointed by a beam; a processing module, which is connected with the The radio module is connected to receive the sound signal and separate a plurality of sound sources in the sound signal. The processing module compares a sound feature point in the sound sources according to the plurality of voiceprint data, and the processing module The system receives a detection target, and uses a deep learning algorithm to extract a sound feature point audio of the audio source corresponding to the detection target among the sound feature points, and amplifies the sound source corresponding to the detection target. sound feature point audio to generate a probe sound source, and reduce or shield the sound feature point audio in the other audio sources to generate a plurality of adjusted audio sources, wherein the processing module is executed for the probe audio source and the adjusted audio sources a synthesizing program for combining the detected sound source and the adjusted sound sources to generate an output sound signal; a selection module connected with the processing module to receive the sound feature points, and selecting at least one of the sound feature points to be the detection target; and a speaker connected with the processing module to receive and output the output sound signal.

Preferably, the directional sound source detection device further comprises: a detection module, which provides a sound wave within a detection range; and a transmission module, which applies a separated sound wave to the detection range within the detection range. On the sound wave, the sound wave is separated into a first sound wave and a second sound wave with two directions; Wherein, the sound pickup module receives the first sound wave and the second sound wave, and forms the beam on the overlapping area of the first sound wave and the second sound wave.

Preferably, the directional sound source detection device further comprises: a first detection module, which provides a first sound wave within a first detection range; and a second detection module, which is located in a A second sound wave is provided in the second detection range; wherein, the sound pickup module receives the first sound wave and the second sound wave, and forms the beam on the overlapping area of the first sound wave and the second sound wave.

Preferably, the directional sound source detection device further comprises: a register, which is connected with the processing module to store the voiceprint data, the sound signal, the sound feature points, and the sound feature points audio, the output sound signal, or a combination of both.

Preferably, the range in which the beam is directed is between 10 degrees and 90 degrees.

Preferably, the frequency range of the sound signal is 0 to 20000 Hz.

Preferably, the directional sound source detection device further comprises: an activation module, which is connected with the processing module and the radio module, the activation module receives the sound signal from the radio module, and according to a The activation threshold is compared with the sound source data of the sound signal to determine whether to activate the processing module. When the sound source data exceeds the activation threshold, the activation module activates the processing module to make the processing The module separates the sound sources in the sound signal, and when the sound source data is lower than the activation threshold, the activation module continues to receive another sound signal.

Preferably, the sound source data includes a decibel value or a frequency value.

Preferably, the sound-receiving module receives a plurality of the sound signals during a judgment period, and selects the sound signal received at at least one negative decibel time point during the judgment period and sends it to the processing module, wherein The negative decibel time point includes the time point when the sound decibel of the sound signal is lower than a negative decibel threshold.

Another object of the present invention is to provide a method for detecting a directional sound source, which mainly uses sound waves to detect the sound on the beam range, and extracts the sound of a specific object (detecting the sound source) according to related processing, thereby reducing or shielding other sounds. Afterwards, the user can effectively listen to the voice of a specific object, thereby effectively improving the problems of the prior art.

In order to achieve the purpose disclosed above, the present invention provides a directional sound source detection method applied to the above-mentioned directional sound source detection device.

In order to make the above-mentioned objects, features and advantages of the present invention more clearly understood, the following detailed description is given in conjunction with the specific embodiments listed in the drawings.

10: Radio module

11: Sound signal

111: Sound source information

12: Beam

13: First Sonic

14: Second Sound Wave

20: Processing modules

21: Audio source

211: Sound feature points

212: Sound feature point audio

22: Voiceprint data

23: Explore the source of the sound

24: Adjust the audio source

25: Output sound signal

30: Select the mod

31: Explore the target

40: Speaker

50: Exploration Module

60: Send module

71: The first exploration module

72: Second Exploration Module

80: Activate the mod

81: negative decibel threshold

OA: Overlap Area

S01-S08: Step Flow

FIG. 1 is a schematic diagram of the configuration relationship of components of the present invention; FIG. 2 is a schematic diagram of the configuration of a detection module and a sending module according to an embodiment of the present invention; FIG. 3 is a first detection diagram of another embodiment of the present invention. A schematic diagram of the configuration of the module and the second detection module; FIG. 4 is a schematic diagram of the first acoustic wave and the second wave forming beam of the present invention; FIG. 5 is a schematic diagram of the configuration of the activation module of the present invention; The flow chart of the steps of the present invention.

The advantages, features, and technical means of achieving the present invention will be more easily understood by being described in more detail with reference to the exemplary embodiments and the accompanying drawings, and the present invention may be implemented in different forms, so it should not be construed as the present invention. It is limited to the embodiments set forth herein. On the contrary, to those of ordinary skill in the art, the embodiments are provided so that the present disclosure will be more thorough, complete and complete. reach the scope of the present invention, and the present invention shall be defined only by the appended claims.

Additionally, the terms "comprising" and/or "comprising" refer to the presence of stated features, regions, integers, steps, operations, elements and/or components, but do not exclude one or more other features, regions, integers, steps, operations , elements, components and/or the presence or addition of combinations thereof.

In order to facilitate your reviewers to understand the content of the present invention and the effects that can be achieved, the specific embodiments listed in the drawings are described in detail as follows: Please refer to FIG. 1 , which is a schematic diagram of the component configuration relationship of the present invention . As shown in the figure, the present invention is mainly composed of a radio module 10 , a processing module 20 , a selection module 30 and a speaker 40 . The radio module 10 can specifically be a microphone or other related devices that can receive external sounds. In order to effectively receive the sound of a specific object, in the present invention, the radio range of the radio module 10 is limited to a Within the range pointed by the beam, a sound signal 11 is received from the range of the beam, and when the sound within the range of the beam is too noisy to effectively capture the sound of a specific object, the radio module 10 Further, a plurality of the sound signals are received during a judgment time, and the sound signal 11 received at at least one negative decibel time point is selected to be sent to the processing module 20 during the judgment time, wherein the negative decibel time point It includes the time point when the sound decibel of the sound signal 11 is lower than a negative decibel threshold (eg, 60 decibels or other decibel values that will not overwhelm the sound of a specific object). The purpose of setting the negative decibel time point is that, for example, in a class environment, although the voice of the classmates is noisy, the volume of people's speech cannot be kept at a high decibel volume at any time. Therefore, when the radio module 10 captures When acquiring the sound signal 11 at the negative decibel time point (ie, the time point when the students' speech is low), the radio module 10 can use the sound signal to perform subsequent processing and judgment.

Specifically, the processing module 20 can be a central processing unit or other device capable of data processing. When the processing module 20 is connected to the radio module 10 , it can receive the radio module 10 . The received sound signal 11, at this time, the processing module 20 will first separate a plurality of sound sources 21 in the sound signal 11 (such as the sound of a specific object and other people's voices), when the sound sources 21 are separated out Afterwards, in order to effectively distinguish the differences between the sound sources 21 and select a specific object based on the differences, the processing module will also use a plurality of voiceprint data 22 to compare a sound feature in the sound sources 21 At point 211, the voiceprint data 22 can be obtained through a period of voice learning or data stored in the original register. At this time, the selection module 30 can receive the sound feature points 211 by connecting with the processing module 20. In a specific embodiment, the selection module 30 can display a selection interface to The user selects, or uses an automatic selection method (for example, when the object to be captured has been set) to receive a selection signal, and according to the selection signal, at least one of the sound feature points 211 is selected as a detection target 31, and then sent to the Processing module 20 .

When the processing module 20 receives the detection target 31, a deep learning algorithm is used to capture a sound feature point audio 212 of the sound source 21 corresponding to the detection target 31 among the sound feature points 211, And amplifying the audio feature point audio 212 in the audio source 21 corresponding to the detection target 31 to generate a detection audio source 23 .

In addition to the above-mentioned action of highlighting the detected sound source 23, the processing module 20 further reduces or shields the sound feature point audio 212 in the other sound sources 21 to generate a plurality of adjusted sound sources 24, and for the detected sound source 23 and the adjusted audio sources 24 execute a synthesis procedure, so that the probed audio source 23 and the adjusted audio sources 24 are combined to generate an output audio signal 25. In this way, the output audio signal 25 will include the highlighted probed audio signal 25. The sound source 23 and the adjusted sound sources 24 that are attenuated or muted.

After that, the speaker 40 can receive the output sound signal 25 by connecting with the processing module 20, and output the output sound signal 25 to the user for listening. In this way, the sound source 23 can be selectively captured, and other noises can be reduced or shielded, so that the user can listen to the sound of a specific object. information and other functions.

In addition, since the conversion or processing between digital signals or analog signals is a conventional technique, the above-mentioned operations of receiving or outputting signals are not repeated in the operations known in the prior art.

Please refer to FIG. 2 to FIG. 4 again, which are schematic diagrams of the configuration of a detection module and a transmission module according to an embodiment of the present invention, and schematic diagrams of a configuration of a first detection module and a second detection module according to another embodiment of the present invention. , and a schematic diagram of the first acoustic wave and the second acoustic wave forming a beam. As shown in the figure, in order to effectively provide the sound-receiving module 10 to obtain the sound signal 11 in the range pointed by the beam 12, the present invention provides two implementations as an example. One embodiment is to detect the directional sound source. The device further includes a detection module 50 and a transmission module 60, the detection module 50 can provide a sound wave within a detection range, and the detection range can be, for example, directed to the sound source detection device. Facing the wide-angle range, the transmitting module 60 provides a separation sound wave to the sound wave in the detection range, so that the sound wave is separated into two directional first sound waves 13 and 13 in the detection range. A second acoustic wave 14, and an overlapping area OA is formed between the first acoustic wave 13 and the second acoustic wave 14. At this time, since the radio module 10 can be equipped with the function of forming the beam 12, when the When the sound-receiving module 10 receives the first sound wave 13 and the second sound wave 14 , the beam 12 can be formed on the overlapping area OA, and the sound signal 11 can be captured in the range of the beam 12 .

In another embodiment, the directional sound source detection device may include a first detection module 71 and a second detection module 72, so as to directly use the first detection module 71 in a first detection module 71 A first sound wave 13 is provided within the detection range, and the second detection module 72 is provided with a second sound wave 14 within the second detection range. It is worth noting that, in order to provide the radio module 10 to form the beam , so the first detection range and the second detection range should have the overlapping area OA, so that when the radio module 10 receives the first sound wave 13 and the second sound wave 14, it can be in the overlapping area OA form the beam 12 on the The audio signal 11 is captured over the range of the beam 12 .

In this way, the present invention can at least utilize the methods provided by the above-mentioned embodiments to achieve the effect of directional acquisition of the sound signal.

Please refer to FIG. 5 again, which is a schematic diagram of the configuration of the activation module of the present invention. As shown in the figure, in order to effectively save power, the directional sound source detection device of the present invention may further include an activation module 80, which is connected to the processing module 20 and the radio module 10. The activation module 80 The sound signal 11 is received from the sound pickup module 10, and the negative decibel threshold value 81 is compared with the sound source data 111 (for example, a decibel value) of the sound signal 11 to determine whether to activate the processing module 20. For example, when the sound source data 111 is lower than the negative decibel threshold 81 (for example, the above-mentioned 60 decibels), the activation module 80 activates the processing module 10, so that the processing module 10 separates the sound For the sound sources 21 in the signal 11 , when the sound source data 111 is higher than the negative decibel threshold 81 , the activation module 80 continues to receive another sound signal 11 to determine the relevant activation procedure. Thereby, the power saving function of the device for detecting the directed sound source of the present invention is effectively provided.

Please refer to FIG. 6 again, which is a flow chart of the steps of the present invention. As shown in the figure, the present invention can mainly be based on the following steps to achieve the above-mentioned function of effectively listening to the sound source, which includes: S01: the radio module receives sound signals within the range pointed by a beam; : the processing module separates a plurality of sound sources in the sound signal; S03: the processing module compares the sound feature points in these sound sources according to the voiceprint data; S04: the selection module receives these sound feature points, and selects at least A sound feature point is the detection target; S05: The processing module receives the detection target, and uses a deep learning algorithm to extract a sound feature point audio of the sound feature points corresponding to the sound source of the detection target; S06: The processing module amplifies or amplifies and frequency shifts the corresponding detection Obtaining the sound feature point audio in the target sound source to generate a probe sound source, and reducing or shielding the sound feature point audio in the other sound sources to generate a plurality of adjusted sound sources; S07: The processing module is directed to the probe sound source and the sound source. A synthesis procedure is performed on the adjusted sound sources, so that the detected sound source and the adjusted sound sources are combined to generate an output sound signal; S08: the speaker receives and outputs the output sound signal.

In this way, the present invention can effectively provide a device capable of selectively capturing and probing sound sources, and reducing or shielding other noises, so that the user can listen to the audio of a specific object.

What is disclosed in this case is a preferred embodiment, and any partial changes or modifications that originate from the technical ideas of this case and are easily inferred by those who are familiar with the art are within the scope of the patent right of this case.

To sum up, in terms of purpose, means and efficacy, this case is showing its technical characteristics that are completely different from those of the prior art, and its first invention is suitable for practical use, and it also meets the requirements of a patent for invention. Granting a patent as soon as possible will benefit the society, and it will be a real sense of virtue.

10: Radio module

11: Sound signal

20: Processing modules

21: Audio source

211: Sound feature points

212: Sound feature point audio

22: Voiceprint data

23: Explore the source of the sound

24: Adjust the audio source

25: Output sound signal

30: Select the mod

31: Explore the target

40: Speaker

Claims (9)

A pointing sound source detection device, comprising: a sound-receiving module, which receives a sound signal within the range pointed by a beam; a processing module, which is connected with the sound-receiving module to receive the sound signal, and separates the For the plurality of sound sources in the sound signal, the processing module compares a sound feature point in the sound sources according to the plurality of voiceprint data, and the processing module receives a detection target and applies a deep learning The algorithm extracts a sound feature point audio of the sound feature point corresponding to the sound source that should be detected, and the processing module amplifies the sound feature point audio in the sound source corresponding to the detected target to generate a sound feature point audio Take the sound source, and reduce or shield the sound feature point audio in the other sound sources to generate a plurality of adjusted sound sources, wherein the processing module executes a synthesis program for the detected sound source and the adjusted sound sources, so that the detected sound source The sound source and the adjusted sound sources are combined to generate an output sound signal; a selection module is connected with the processing module to receive the sound feature points, and the selection module receives a selection signal to select at least one sound signal according to the selection signal The sound feature point is the detection target; and a speaker, which is connected with the processing module to receive and output the output sound signal; wherein, the sound receiving module receives a plurality of the sound signals within a judgment period, and the The sound signal received at at least one negative decibel time point is selected within the judgment period and sent to the processing module, wherein the negative decibel time point includes the time when the sound decibel of the sound signal is lower than a negative decibel threshold value point. The directional sound source detection device according to claim 1, further comprising: a detection module, which provides a sound wave within a detection range; and a transmitting module, which applies a separation sound wave to the sound wave within the detection range, so that the sound wave is separated into two directivity, a first sound wave and a second sound wave; wherein, the sound receiving module receives the first sound wave The acoustic wave and the second acoustic wave form the beam on the overlapping area of the first acoustic wave and the second acoustic wave. The directional sound source detection device according to claim 1, further comprising: a first detection module, which provides a first sound wave within a first detection range; and a second detection module, which A second sound wave is provided in a second detection range; wherein, the sound pickup module receives the first sound wave and the second sound wave, and forms the beam on the overlapping area of the first sound wave and the second sound wave. The device for detecting a sound source according to claim 1, further comprising: a register connected with the processing module to store the voiceprint data, the sound signal, the sound feature points, the sounds Feature point audio, the output audio signal, or a combination of two or more. The directed sound source detection device as claimed in claim 1, wherein the range of the beam directed is between 10 degrees and 90 degrees. The directional sound source detection device as claimed in claim 1, wherein the frequency range of the sound signal is 0 to 20000 Hz. The directed sound source detection device according to claim 1, further comprising: an activation module connected to the processing module and the radio module, the activation module receiving the sound signal from the radio module, and Comparing an activation threshold with the sound source data of the sound signal to determine whether to activate the processing module, When the sound source data exceeds the activation threshold, the activation module activates the processing module, so that the processing module separates the sound sources in the sound signal, and when the sound source data is lower than the activation threshold, The activation module continues to receive another sound signal. The device for detecting a sound source according to claim 7, wherein the sound source data includes a decibel value or a frequency value. A directional sound source detection method applied to the directional sound source detection device according to any one of claim items 1 to 8.

Images