KR102351819B1 - Method and apparatus for simultaneously supporting remote voice recognition of two microphones and multiple broadcasters or ai service companies - Google Patents

Abstract

Proposed are a sampling method and device of a PDM input when simultaneously supporting a remote voice recognition of multiple broadcasting and communication companies or AI service companies with two microphones. The sampling method of the PDM input when simultaneously supporting the remote voice recognition of the multiple broadcasting communication companies or AI service companies with two microphones according to one embodiment comprises a step of installing a software module so that a CPU receives the microphone data at a correct sampling position when an input of an echo canceled microphone and an input of a non-echo canceled microphone are simultaneously transmitted to the CPU in a DSP. Therefore, the present invention is capable of receiving the microphone data of the CPU at the correct sampling position.

Description

Sampling method and device for PDM input when two microphones simultaneously support remote voice recognition of multiple broadcasters or AI service providers

The following embodiments relate to a sampling method and apparatus of a PDM input when simultaneously supporting Far Field Voice Recognition of a plurality of broadcasters or AI service providers with two microphones.

In recent set-top boxes, you can ask for weather, news, etc. with voice commands, and the set-top box processes the voice commands to notify the results by voice through the built-in speaker (Text To Speech, TTS). Microphone input will install In order to process this voice command, two microphones are basically installed more than a certain distance apart to increase reliability and prevent the TV output sound of the built-in speaker output or the stream output from the set-top box from being input again through the microphone. To cancel it, a technique called Acoustic Echo Cancel (AEC) is used.

Broadcasting and telecommunication service providers who want to receive the input of such echo cancelled microphones use an external digital signal processor (DSP), and use a sound reference that is output to a speaker or TV to the DSP. is output simultaneously, the DSP removes the sound reference component from the microphone input and provides only the voice command input to the microphone to the CPU.

However, broadcasters may request to support Google's Voice Assistant (GVA) or Amazon AI service at the same time. It is provided to the company's server, and the result is provided through its own processing.

In this case, the CPU must be provided with both a microphone input that has undergone DSP processing and a microphone input that has not been processed. It can be difficult to get.

Korean Patent Application Laid-Open No. 10-2010-0100522 describes a technology related to such a multi-channel IPTV set-top box and an IPTV broadcasting system for a multi-channel service.

Korean Patent Publication No. 10-2010-0100522

The embodiments describe a sampling method and apparatus of a PDM input when simultaneously supporting remote voice recognition of a plurality of broadcasting service providers or AI service providers with two microphones, and more specifically, using only two microphones, AI different from broadcasting service providers It provides a technology for sampling PDM input when simultaneously supporting voice recognition of a service provider.

Embodiments provide a method and apparatus for sampling PDM input when simultaneously supporting remote voice recognition of a plurality of broadcasting communication operators or AI service providers with two microphones, which installs a software module to receive the microphone data of the CPU at an accurate sampling position is to provide

In addition, embodiments provide long-distance voice recognition of a plurality of broadcasters or AI service providers with two microphones, in which the left and right positions of the speaker can be identified equally from the DSP and CPU even if the phases of the DSP PDM Clock and the CPU PDM Clock are opposite. It is to provide a sampling method and apparatus for PDM input when simultaneously supporting .

The sampling method of the PDM input when the remote voice recognition of a plurality of broadcasting service providers or AI service providers is supported simultaneously with two microphones according to an embodiment is an input of a microphone with echo cancellation in the DSP, and an echo cancellation (Echo) When the non-canceled microphone input is simultaneously transferred to the CPU, installing a software module may include the step of allowing the CPU to receive the microphone data at the correct sampling position.

Using only two microphones, the input of the microphone whose echo cancelled by the DSP desired by the broadcasting communication service provider and the input of the non-echo-canceled microphone of the other AI service provider can be simultaneously transmitted to the CPU.

The step of allowing the CPU to receive the microphone data at the exact sampling position may include: requesting a speaker to make a specific sound; finding at which position data such as a result of DSP is input while changing a sampling position using the specific sound; and determining a section of the found sampling position and determining the position as an intermediate value of the value to obtain stable microphone input data.

The step of allowing the CPU to receive the microphone data at the exact sampling position may include: when the speaker's left and right positions are to be identified, requesting the speaker to move to the left or right side of the set-top box to produce a specific sound; and using the input data as it is when the acquired intensity of the specific sound shows the same result in both the DSP and the CPU.

The step of allowing the CPU to receive the microphone data at the correct sampling position is the step of using the position of the DSP and reversing the direction on the CPU side when the acquired specific sound intensity shows different results on both sides of the DSP and CPU. may include more.

In the step of allowing the CPU to receive the microphone data at the correct sampling position, the phase of the CPU PDM Clock is corrected when the left and right positions of the speaker are opposite to each other due to the phase difference between the DSP PDM Clock and the CPU PDM Clock. can do.

When remote voice recognition of a plurality of broadcasting communication operators or AI service providers is supported simultaneously with two microphones according to another embodiment, the sampling device of the PDM input is the input of the microphone and the echo cancellation (Echo Cancel) in the DSP. ) may include a software module that allows the CPU to receive microphone data at the correct sampling position when simultaneously transmitting the input of the microphone to the CPU.

According to embodiments, a method for sampling PDM input when a software module is installed to receive the microphone data of the CPU at an accurate sampling location, and the remote voice recognition of a plurality of broadcasters or AI service providers is supported simultaneously with two microphones and devices may be provided.

In addition, according to embodiments, even if the DSP PDM Clock and the CPU PDM Clock have opposite phases, the left and right positions of the speaker can be identified equally from the DSP and CPU. When simultaneously supporting voice recognition, a sampling method and apparatus for PDM input may be provided.

1 is a view for explaining a method of simultaneously supporting remote voice recognition of a plurality of broadcasting communication service providers or AI service providers using four conventional microphones.
2 is a block diagram illustrating a clock input and data output method of a general digital microphone.
3 is a timing diagram for explaining a clock input and data output method of a general digital microphone.
4 is a diagram for explaining the reason why the same voice input cannot be received when clocks of a general DSP and CPU do not exactly match.
5 is a diagram for explaining a method of simultaneously supporting remote voice recognition of a plurality of broadcasting communication operators or AI service providers with two microphones according to an embodiment.
6 is a diagram illustrating an example of a case in which frequencies coincide but phases are different according to an embodiment.
7 is a diagram for describing a method in which a CPU receives an input of a PDM according to an embodiment.
8 is a diagram for explaining the reason why the CPU can normally accept a microphone input even if there is a phase difference in one embodiment.
9 is a flowchart illustrating a method of simultaneously supporting remote voice recognition of a plurality of broadcasting communication operators or AI service providers with two microphones according to an embodiment.
10 is a diagram illustrating a case in which a PDM clock of a CPU and a microphone clock (DSP PDM clock) have opposite phases according to an embodiment.
11 is a flowchart illustrating a sampling method of a PDM input when remote voice recognition is simultaneously supported according to an embodiment.

Hereinafter, embodiments will be described with reference to the accompanying drawings. However, the described embodiments may be modified in various other forms, and the scope of the present invention is not limited by the embodiments described below. In addition, various embodiments are provided in order to more completely explain the present invention to those of ordinary skill in the art. The shapes and sizes of elements in the drawings may be exaggerated for clearer description.

1 is a view for explaining a method of simultaneously supporting remote voice recognition of a plurality of broadcasting communication service providers or AI service providers using four conventional microphones.

Referring to FIG. 1 , conventionally, four microphones 31 and 32 may be used to use Far Field Voice (FFV) of a broadcasting service provider and voice recognition of another AI service provider.

When a broadcasting service provider wants to simultaneously support its own far-field voice (FFV) service and another AI service provider's voice recognition, the broadcasting service provider must ensure that the far-field voice (FFV) service eliminates echoes by the DSP 20 If you want to receive (Echo Canceled) microphone input, and other AI service providers expect to receive pure microphone input that has not gone through DSP (20), use 4 microphones (31, 32) to create 2 microphones (32) is used as a microphone input for the DSP 20, and the other two microphones 31 are used as a microphone input for an AI service provider.

This means that the digital microphones 31 and 32 used in the far-field voice (FFV) service receive one clock signal from the DSP 20 or the central processing unit (CPU) 10, and are synchronized with the clock signal. This is because one microphone data is output, and the CPU 10 and DSP 20 operate by different input clocks of 24 MHz.

If the CPU 10 and DSP 20 use different 24 MHz clocks, each clock is not an accurate 24 MHz clock, but has a different frequency and phase difference between, for example, 23.9xxx MHz ~ 24.0xxx MHz. Therefore, there is no guarantee that the input from the same microphone is received.

2 is a block diagram illustrating a clock input and data output method of a general digital microphone. 3 is a timing diagram for explaining a clock input and data output method of a general digital microphone.

CPU and DSP operate by different input clocks, and when CPU and DSP use different clocks, they show different frequencies and phase differences. The data output method of the digital microphone used at this time may be represented by a block diagram and a timing diagram as shown in FIGS. 2 and 3 .

4 is a diagram for explaining the reason why the same voice input cannot be received when clocks of a general DSP and CPU do not exactly match.

As shown in Fig. 4, it shows that the DSP and CPU cannot receive the same microphone input in an environment where the frequencies do not match. If the frequency and phase of the CPU and DSP are different, they cannot be shared because the microphone data of each does not match.

As shown in FIG. 1, if four microphones are used in a method that simultaneously supports remote voice recognition of a plurality of broadcasting communication operators or AI service providers, it is not good in terms of space and cost for arranging these four microphones. will show non-existent results. In particular, when multiple microphones are used, the distance between each microphone must be sufficiently far to secure the voice input quality, which causes a problem in miniaturizing the size of the set-top box. It costs a lot of money to install a structure to prevent it.

The following embodiment proposes a method for simultaneously supporting voice recognition of a broadcasting communication service provider and another AI service company using only two microphones. In addition, in the following embodiment, when two microphones simultaneously support remote voice recognition of a plurality of broadcasting communication operators or AI service providers, a sampling method of PDM input is presented.

5 is a diagram for explaining a method of simultaneously supporting remote voice recognition of a plurality of broadcasting communication operators or AI service providers with two microphones according to an embodiment.

Referring to FIG. 5 , a method for realizing far-field voice (FFV) and voice recognition of AI service providers through DSPs of two micro broadcast operators is shown.
In the present specification, the first service method refers to a method in which the DSP cancels echo and delivers voice data to the CPU, and the second service method refers to a method in which non-echo-cancelled voice data is delivered to the CPU without going through the DSP. it means.

A device that simultaneously supports remote voice recognition of a plurality of broadcasting carriers or AI service providers with two microphones according to an embodiment is two microphones 530 for receiving a voice command, and the input of the two microphones 530 . It may be composed of a DSP 520 in charge of receiving an echo cancellation function, and a CPU 510 receiving a non-echo cancellation microphone input.

More specifically, the apparatus for simultaneously supporting remote voice recognition of a plurality of broadcasting communication operators or AI service providers with two microphones according to an embodiment includes a plurality of microphones 530 and microphones input through the plurality of microphones 530 . DSP 520 for echo canceling the voice data acquired through the first service method among the inputs of the microphone, and echo cancellation ( The non-echo-canceled voice data is transferred as it is, and the first service method may include the CPU 510 to which the input of the microphone echo-canceled by the DSP 520 is transferred. On the other hand, a plurality of microphones 530, DSP 520, and CPU 510 may be built into the set-top box, and in particular, the plurality of microphones 530 may use two microphones to reduce the size of the set-top box. . In other words, the set-top box device simultaneously supporting two different voice recognition processing methods may be configured to include a plurality of microphones 530 , DSP 520 , and CPU 510 .

As described above, if the CPU 510 and the DSP 520 use different 24 MHz clock inputs, it cannot be guaranteed that the frequencies of the two clocks exactly match.

In order to solve the above problem, as shown in FIG. 5 , the DSP uses the clock of 24 MHz output by the CPU using the PWM output function of the CPU, thereby solving the problem of different frequencies between the two. . In addition, the microphone outputs the microphone data using the PDM Clock of the DSP, and as shown in FIG. 7, the CPU reads the PDM input, so even if the PDM Clock of the CPU is not provided to the microphone, As shown in Figure 8, the same microphone data as the DSP can be input.

6 is a diagram illustrating an example of a case in which frequencies coincide but phases are different according to an embodiment.

If the CPU and DSP use different 24MHz clock inputs, the DSP cannot guarantee that the frequencies of the two clocks exactly match, so as shown in FIG. The 24MHz Clock created using the system will be provided and used. In this way, the problem of frequency mismatch is solved.

7 is a diagram for describing a method in which a CPU receives an input of a PDM according to an embodiment.

As shown in Fig. 7, even if the CPU and DSP use the clock input of the same frequency (24MHz), the FCLOCK (0~5MHz) used by the microphone is set to the respective Phase Locked Loop (PLL) module. However, there is no guarantee that these two clocks are in phase with each other. In other words, there is a problem that two clocks that are in phase with each other cannot be provided to the microphone.

The CPU may use an internal clock of a much higher frequency than the PDM clock (eg, 133 MHz) to adjust the position at which the PDM data is to be received.

8 is a diagram for explaining the reason why the CPU can normally accept a microphone input even if there is a phase difference in one embodiment.

Referring to FIG. 8 , it is possible to explain why the CPU can normally accept a microphone input even if there is a phase difference between the PDM Clock of the CPU and the microphone clock (DSP PDM Clock) according to an embodiment.

Accordingly, in this embodiment, the FCLOCK provided by the micro is used as the PDM (Pulse Duration Modulation) CLOCK provided by the DSP, and the PDM CLOCK provided by the CPU is not used. FIG. 6 shows that in the system configured in this way, data input from the microphone by the PDM clock of the DSP may not coincide with the PDM clock of the CPU that is not used in terms of phase.

To solve this problem, you need to look at the CPU's PDM Timing Spec. 7 shows that the CPU can select an input position (sampling position) by using a more divided high-frequency internal clock, and using this function, microphone data can be received from the same position as the DSP as shown in FIG. have.

9 is a flowchart illustrating a method of simultaneously supporting remote voice recognition of a plurality of broadcasting communication operators or AI service providers with two microphones according to an embodiment.

Referring to FIG. 9 , in the method for simultaneously supporting remote voice recognition of a plurality of broadcasting communication operators or AI service providers with two microphones according to an embodiment, the step of transmitting the input of the microphone of the first service method to the DSP (S110) ), the step of Echo Canceling the input of the microphone received from the DSP and transmitting it to the CPU (S120), and the second service method of transferring the input of the microphone that has not been echo canceled to the CPU. It may be made including step (S130).

Hereinafter, each step of the method of simultaneously supporting remote voice recognition of a plurality of broadcasting communication operators or AI service providers with two microphones according to an embodiment will be described in more detail.

A method of simultaneously supporting remote voice recognition of a plurality of broadcasting communication operators or AI service providers with two microphones according to an embodiment is two microphones according to an embodiment described with reference to FIG. A device that simultaneously supports voice recognition may be described as an example. As described above, a device that simultaneously supports remote voice recognition of a plurality of broadcasting communication operators or AI service providers with two microphones according to an embodiment includes a plurality of microphones 530 , DSP 520 , and CPU 510 . can be done by

In step S110 , the input of the microphone of the first service method may be transmitted to the DSP 520 . In this case, the plurality of microphones 530 may transmit the microphone input received from the first service method to the DSP 520 .

In step S120 , after echo cancellation for the microphone input received from the DSP 520 , it may be transmitted to the CPU 510 .

For example, the first service method may be a broadcasting communication service provider. In the case of the microphone input of the broadcasting communication service provider, the input of the microphone echo cancelled by the DSP 520 desired by the broadcasting communication service provider may be transmitted to the CPU 510 .

In step S130 , an input of a microphone that is not echo canceled of the second service method may be transmitted to the CPU 510 . In this case, the plurality of microphones 530 may transfer the microphone input that is not echo canceled received from the second service method to the CPU 510 .

For example, the second service method may be an AI service provider. In the case of the input of the microphone of the AI service company, the input of the microphone that has not been echo canceled may be transferred to the CPU 510 .

As such, according to embodiments, the input of the microphone that is echo-canceled in the DSP 520 desired by the broadcasting operator using only two microphones and the microphone that is not echo-canceled by other AI service providers The input may be simultaneously transferred to the CPU 510 .

Here, the clock generated by PWM of the CPU 510 may be supplied to the DSP 520 so that the CPU 510 and the DSP 520 use a clock having the same frequency. In addition, the clock of the DSP 520 may be selected from two different PDM clocks output from the CPU 510 and the DSP 520 and provided as the input clock of the microphone.

In addition, the problem of the phase difference caused by selecting the clock of the DSP 520 from among the two different PDM clocks output from the CPU 510 and the DSP 520 and providing it as the input clock of the microphone is solved by the PDM of the CPU 510 . This can be overcome by using the method of changing the sampling position of the data.

According to embodiments, a far-field voice (FFV) service can be provided by receiving an echo-canceled microphone input using the DSP 520, and at the same time, when it is desired to support voice recognition of other AI service providers, the conventional 4 By allowing only two microphones to be used instead of using a dog microphone, space and cost can be significantly reduced.

When the CPU receives microphone data in the same way as above, one of the sampling positions should be selected. For example, as shown in FIG. 8 , one of sampling positions 0 to 9 may be selected. Considering that the position at which the microphone data can be received most stably is the middle part of the microphone data, in the case of FIG. 8 , the sampling position 3 or 4 may be used. This is because the microphone data will be read correctly from 8 positions 0 to 7 out of 10 sampling positions. Meanwhile, although 0 to 9 is described as an example of the sampling position, various settings such as 0 to 15 as well as 0 to 9 can be made.

Here, we present a method of finding the correct data input at sampling positions 0 to 7. In addition, even if this method is applied, the position of the speaker can be reversed by the phase difference between the DSP PDM Clock and the CPU PDM Clock. A specific set-top box may need to distinguish which of the two microphones has the strongest voice data.

On the other hand, as shown in FIGS. 2 and 3, the clock input and data output method of the digital microphone can be shown. In this case, when two microphones are installed on the left and right, the SEL input of the microphone is set to VDD for the left microphone and VDD for the right microphone. The microphone can be set to Ground. Then, it can be distinguished that the data input when the clock is high is from the left microphone, and the data input when the clock is low is the data from the right microphone.

10 is a diagram illustrating a case in which a PDM clock of a CPU and a microphone clock (DSP PDM clock) have opposite phases according to an embodiment.

As shown in FIG. 10 , if the phases of the DSP PDM Clock and the CPU PDM Clock are opposite to each other, the DSP and the CPU may recognize that the direction of the speaker is opposite to each other. In this case, a method for correct recognition of the speaker's direction is also presented.

If the phases of the DSP PDM Clock and the PDM Clock inside the CPU are opposite, the data input at the point where the DSP clock is high is input at the point where the clock is low in the CPU, so the left and right positions can be determined differently. .

For example, a special software module is required to know that the correct data is read from positions 0 to 7. When this module is activated, the set-top box will make a specific sound to the speaker (eg "ahhhhh"). At this time, the software module can change the sampling position from 0 to 9 and find out where the same data as the DSP result is input. Stable microphone input data can be obtained if the section of the sampling position found in this way is identified and the position is determined as the intermediate value of the values.

In addition, as described above, when the speaker's left and right positions are to be identified, the speaker moves to the left or right side of the set-top box to make a sound. At this time, if the strength of the acquired data shows the same result in both DSP and CPU, the input data can be used as it is. Do it.

11 is a flowchart illustrating a sampling method of a PDM input when remote voice recognition is simultaneously supported according to an embodiment.

Referring to FIG. 11 , in the case of simultaneously supporting remote voice recognition of a plurality of broadcasting communication operators or AI service providers with two microphones according to an embodiment, the sampling method of PDM input is to install a software module so that the CPU operates the microphone at the correct sampling position. It may include a step (S210) of receiving data. In addition, the DSP may further include a step (S220) of simultaneously transferring an input of an echo cancelled microphone and a non-echo canceled microphone input to the CPU (S220).

Here, by using only two microphones, the input of the microphone that has been echo canceled by the DSP desired by the broadcasting service provider and the input of the microphone that is not echo canceled by other AI service providers can be simultaneously transferred to the CPU.

Hereinafter, a sampling method of a PDM input when two microphones simultaneously support remote voice recognition of a plurality of broadcasting communication operators or AI service providers according to an embodiment will be described in more detail.

The sampling method of the PDM input when simultaneously supporting remote voice recognition of a plurality of broadcasting communication operators or AI service providers with two microphones according to an embodiment is two microphones according to an embodiment When simultaneously supporting voice recognition, a sampling device for PDM input can be described as an example.

When two microphones according to an embodiment support different remote voice recognition processing methods of a plurality of broadcasting communication operators or AI service providers at the same time, the sampling device of the PDM input is When audio data that is not directly transmitted without echo cancellation is transmitted to the CPU at the same time, a software module that allows the CPU to receive microphone data at the correct sampling position may be included. In addition, according to an embodiment, as described with reference to FIG. 5 , a plurality of microphones, a DSP, and a CPU may be further included. In other words, the set-top box device that simultaneously supports two different voice recognition processing methods includes a plurality of microphones 530 , DSP 520 and CPU 510 together with a software module (not shown) that selects the sampling position of the PDM input. ) may be included. The software module may be implemented as a component included in the set-top box device, like the microphone 530 , the DSP 520 , and the CPU 510 .

In step S210, the software module transmits the microphone input that has been echo canceled by the DSP and the microphone input that is not echo canceled to the CPU at the same time, and the CPU receives the microphone data at the correct sampling position. You can get input.

First, the software module can find the exact sampling location.

To this end, the software module may ask the speaker to make a specific sound. The software module can find where data such as the DSP result is input while changing the sampling position from 0 to 9 using a specific sound. In this case, the sampling positions of 0 to 9 are only an example, and are not limited thereto and may be set in various ways. In addition, the software module may obtain stable microphone input data by determining the interval of the found sampling position and setting the position as an intermediate value of the value.

In addition, the software module may determine the left and right positions of the speaker.

To this end, the software module may request the speaker to move to the left or right side of the set-top box to make a specific sound when the speaker's left and right positions are to be identified. The software module can use the input data as it is if the acquired specific sound intensity shows the same result in both DSP and CPU. In addition, the software module can use the position of the DSP and reverse the direction on the CPU side when the acquired specific sound intensity shows different results on both sides of the DSP and CPU. That is, the software module can correct the phase of the CPU PDM Clock when the left and right positions of the speaker are opposite to each other due to the phase difference between the DSP PDM Clock and the CPU PDM Clock.

In step S220 , the input of the microphone that has been echo canceled by the DSP and the input of the microphone that are not echo canceled may be simultaneously transmitted to the CPU. This has been described in detail in FIG. 9 .

As described above, according to the embodiments, using only two microphones, a microphone input that is echo-canceled from a DSP desired by a broadcasting service provider and a microphone input that is not echo-canceled from another AI service provider can be simultaneously processed by the CPU. In case of using the method of transmitting to the CPU, it is possible to install a software module to receive the microphone data of the CPU at the correct sampling position. Also, even if the DSP PDM Clock and the CPU PDM Clock are out of phase, the left and right positions of the speaker can be identified equally from the DSP and CPU.

Therefore, since the PDM Clock of the DSP is used and the PDM Clock of the CPU is not used in the method presented in this embodiment, the possibility that the microphone data on the CPU side may show a different result from the actual result is prevented, and the DSP PDM Clock and the CPU PDM Clock are not used. It is possible to correct the case where the left and right positions of the speaker are detected opposite to each other due to the phase difference of the clock.

The device described above may be implemented as a hardware component, a software component, and/or a combination of the hardware component and the software component. For example, devices and components described in the embodiments may include, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable array (FPA), It may be implemented using one or more general purpose or special purpose computers, such as a programmable logic unit (PLU), microprocessor, or any other device capable of executing and responding to instructions. The processing device may execute an operating system (OS) and one or more software applications running on the operating system. A processing device may also access, store, manipulate, process, and generate data in response to execution of the software. For convenience of understanding, although one processing device is sometimes described as being used, one of ordinary skill in the art will recognize that the processing device includes a plurality of processing elements and/or a plurality of types of processing elements. It can be seen that can include For example, the processing device may include a plurality of processors or one processor and one controller. Other processing configurations are also possible, such as parallel processors.

Software may comprise a computer program, code, instructions, or a combination of one or more thereof, which configures a processing device to operate as desired or is independently or collectively processed You can command the device. The software and/or data may be any kind of machine, component, physical device, virtual equipment, computer storage medium or apparatus, to be interpreted by or to provide instructions or data to the processing device. may be embodied in The software may be distributed over networked computer systems and stored or executed in a distributed manner. Software and data may be stored in one or more computer-readable recording media.

The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the medium may be specially designed and configured for the embodiment, or may be known and available to those skilled in the art of computer software. Examples of the computer-readable recording medium include magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic such as floppy disks. - includes magneto-optical media, and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like.

As described above, although the embodiments have been described with reference to the limited embodiments and drawings, various modifications and variations are possible from the above description by those skilled in the art. For example, the described techniques are performed in an order different from the described method, and/or the described components of the system, structure, apparatus, circuit, etc. are combined or combined in a different form than the described method, or other components Or substituted or substituted by equivalents may achieve an appropriate result.

Therefore, other implementations, other embodiments, and equivalents to the claims are also within the scope of the following claims.

Claims (7)

delete delete In a set-top box device,
a plurality of microphones;
DSP (digital signal processor) responsible for echo cancellation (Echo Cancel) function;
For voice data input through the microphone, the first service method, which is a voice recognition processing method of the broadcasting communication service provider, and the second service method, which is a voice recognition processing method of the broadcasting communication service provider and another AI (artificial intelligence) service company, are supported simultaneously a central processing unit (CPU); and
A software module for selecting a sampling position of the voice data received by the CPU so that the CPU receives the same data as the DSP from the microphone
including,
Voice data input through the microphone and echo-cancelled through the DSP is transmitted to the CPU as the first service method,
Voice data input through the microphone and not echo-cancelled is delivered to the CPU as it is as the second service method,
The PWM (Pulse Wave Modulation) clock provided by the CPU is supplied to the DSP so that the DSP uses a clock of the same frequency as the CPU,
The microphone does not use the PDM (Pulse Density Modulation) Clock directly provided by the CPU, but uses the PDM Clock provided by the DSP,
The software module is
Ask the speaker to make a specific sound,
Using the specific sound, while changing the sampling position of the voice data received by the CPU, find out where the data such as the result of the DSP is input,
To obtain stable microphone input data by identifying the section of the found sampling position and setting the position as the intermediate value of the value
characterized in that, a set-top box device. In a set-top box device,
a plurality of microphones;
DSP (digital signal processor) responsible for echo cancellation (Echo Cancel) function;
For voice data input through the microphone, the first service method, which is a voice recognition processing method of the broadcasting communication service provider, and the second service method, which is a voice recognition processing method of the broadcasting communication service provider and another AI (artificial intelligence) service company, are supported simultaneously a central processing unit (CPU); and
A software module for selecting a sampling position of the voice data received by the CPU so that the CPU receives the same data as the DSP from the microphone
including,
Voice data input through the microphone and echo-cancelled through the DSP is transmitted to the CPU as the first service method,
Voice data input through the microphone and not echo-cancelled is delivered to the CPU as it is as the second service method,
The PWM (Pulse Wave Modulation) clock provided by the CPU is supplied to the DSP so that the DSP uses a clock of the same frequency as the CPU,
The microphone does not use the PDM (Pulse Density Modulation) Clock directly provided by the CPU, but uses the PDM Clock provided by the DSP,
The software module is
Ask the speaker to move to the left or right side of the set-top box device to make a specific sound,
When the acquired specific sound intensity shows the same result in the DSP and the CPU, using the input data as it is
characterized in that, a set-top box device. delete 5. The method of claim 4,
The software module is
Correcting the phase of the PDM Clock of the CPU when the acquired specific sound intensity shows a different result between the DSP and the CPU
characterized in that, a set-top box device. delete

Images