An audio signal processing device, method, and storage medium

Abstract

The present application discloses an audio signal processing device, which includes a pickup, a speaker, and an audio signal processing chip. The audio signal processing chip includes a voiceprint recognition module, a comparison module, and a storage module. The pickup is used to obtain an audio signal and convert the audio signal into an analog signal. The voiceprint recognition module is used to detect whether the analog signal contains human voices. The comparison module is used to compare whether the target analog signal is the same as the sample signal stored in the storage module. If they are the same, the target analog signal is deleted. If they are different, the target analog signal is transmitted to the storage module as a sample signal for storage. The speaker is used to obtain the latest stored sample signal from the storage module, convert the latest stored sample signal into a corresponding audio signal, and then output it. The present application can filter echoes and accurately obtain the user's voice.

Description

Technical Field

[0001] This application relates to the technical field of audio signal processing, and in particular to an audio signal processing device, method, and storage medium. Background Art

[0002] Existing intelligent devices usually have a voice interaction function for interacting with users. During the interaction between an intelligent device and a user, when collecting the user's voice signal as the target signal, the intelligent device collects the user's voice signal through a microphone. After the user's voice signal is converted into an analog signal by the microphone, it is transmitted to a speaker. The speaker can convert the analog signal into an audio signal and play the sound, but the sound played by the speaker will generate an echo after reflection, and the echo will be collected by the microphone again, which will interfere with the microphone's collection of the user's voice. Although existing intelligent devices with voice interaction functions are all configured with an audio signal processing chip that can be used to process analog signals, and the audio signal processing chip includes a power generation module, a clock signal generation module, an analog signal detection module, an analog signal processing module, a hardware acceleration module, a storage module, and an interface module, etc., it is still not sufficient to help the intelligent device accurately collect the user's voice. Summary of the Invention

[0003] In view of the above technical problems, this application provides an audio signal processing device, method, and storage medium, which can filter echoes and accurately obtain the user's voice.

[0004] An embodiment of this application provides an audio signal processing device, including a pickup, a speaker, and an audio signal processing chip. The audio signal processing chip includes a voiceprint recognition module, a comparison module, and a storage module;

[0005] The pickup is configured to obtain an audio signal and convert the audio signal into an analog signal and then transmit it to the voiceprint recognition module;

[0006] The voiceprint recognition module is configured to detect whether the analog signal contains human voices. If so, it transmits the analog signal as a target analog signal to the comparison module; otherwise, it deletes the analog signal;

[0007] The storage module is configured to store a sample signal, and the sample signal is a user voice signal;

[0008] The comparison module is configured to compare whether the target analog signal is the same as the sample signal stored in the storage module. If they are the same, it deletes the target analog signal; if they are different, it transmits the target analog signal as a sample signal to the storage module for storage; and when the storage module does not store the sample signal, it transmits the target analog signal as a sample signal to the storage module for storage;

[0009] The comparison module includes a plurality of comparison units, and each comparison unit compares the target analog signal with the sample signal according to a preset comparison method; if each comparison unit determines that the analog signal is different from the sample signal, the comparison module determines that the target analog signal is different from the sample signal; otherwise, the comparison module determines that the target analog signal is the same as the sample signal.

[0010] The loudspeaker is configured to obtain the latest stored sample signal from the storage module, and convert the latest stored sample signal into a corresponding audio signal and output it.

[0011] Optionally, at least one of the plurality of comparison units includes a time comparison unit.

[0012] The comparison unit compares the target analog signal with the sample signal according to a preset comparison method, including:

[0013] The time comparison unit compares the starting point of the target analog signal with the starting point of the sample signal.

[0014] If the starting point of the target analog signal is different from the starting point of the sample signal, the time comparison unit determines that the target analog signal is different from the sample signal; otherwise,

[0015] The time comparison unit compares the ending point of the analog signal with the ending point of the sample signal.

[0016] If the ending point of the target analog signal is different from the ending point of the sample signal, the time comparison unit determines that the target analog signal is different from the sample signal; otherwise, it determines that the target analog signal is the same as the sample signal.

[0017] Optionally, at least one of the plurality of comparison units includes a waveform comparison unit.

[0018] The comparison unit compares the target analog signal with the sample signal according to a preset comparison method, including:

[0019] The waveform comparison unit compares the waveform of the target analog signal with the waveform of the sample signal.

[0020] If the similarity between the waveform of the target analog signal and the waveform of the sample signal reaches a preset threshold, the waveform comparison unit determines that the target analog signal is the same as the sample signal; otherwise, it determines that the target analog signal is different from the sample signal.

[0021] Optionally, the audio signal processing chip further includes an adjustment module;

[0022] The adjustment module is configured to compare the amplitude of the target analog signal with the amplitude of the sample signal when the time comparison unit determines that the target analog signal is the same as the sample signal;

[0023] If the amplitude of the target analog signal is greater than the amplitude of the sample signal, reduce the amplitude of the target analog signal; if the amplitude of the target analog signal is less than the amplitude of the sample signal, increase the amplitude of the target analog signal;

[0024] Transmit the adjusted target analog signal to the waveform comparison unit.

[0025] Optionally, the storage module includes a plurality of storage units, each storage unit stores a corresponding sample signal, each storage unit is numbered in sequence, and each number is recorded in a mapping table, wherein the mapping table is stored in one of the storage units of the storage module;

[0026] The storage module transmits the sample signal to the comparison module, including:

[0027] The storage module looks up the corresponding storage unit according to the number recorded in the mapping table;

[0028] The storage module obtains the corresponding sample signal from the found storage unit and transmits the corresponding sample signal to the comparison module.

[0029] Optionally, the storage module is a volatile memory, and each time the audio signal processing chip is powered off and restarted, the data in the storage module will be cleared.

[0030] An embodiment of the present application provides an audio signal processing method for an audio signal processing device, the audio signal processing device includes a pickup, a speaker, and an audio signal processing chip, the audio signal processing chip includes a voiceprint recognition module, a comparison module, and a storage module, and the audio signal processing method includes the following steps:

[0031] The pickup obtains an audio signal and converts the audio signal into an analog signal and then transmits it to the voiceprint recognition module;

[0032] The voiceprint recognition module detects whether the analog signal contains human voice. If so, it transmits the analog signal as a target analog signal to the comparison module, otherwise, deletes the analog signal;

[0033] The comparison module compares whether the target analog signal is the same as the sample signal stored in the storage module. If they are the same, the target analog signal is deleted; if they are different, the target analog signal is transmitted to the storage module as a sample signal for storage; and it is used to transmit the target analog signal to the storage module as a sample signal for storage when the storage module does not store the sample signal.

[0034] Wherein, the storage module is used to store sample signals, and the sample signals are user voice signals.

[0035] The comparison module includes a plurality of comparison units, and each comparison unit compares the target analog signal with the sample signal according to a preset comparison method. If each comparison unit determines that the analog signal is different from the sample signal, the comparison module determines that the target analog signal is different from the sample signal; otherwise, the comparison module determines that the target analog signal is the same as the sample signal.

[0036] The speaker obtains the latest stored sample signal from the storage module, and converts the latest stored sample signal into a corresponding audio signal and outputs it.

[0037] Optionally, at least one time comparison unit is included in the plurality of comparison units.

[0038] The comparison unit compares the target analog signal with the sample signal according to a preset comparison method, including:

[0039] The time comparison unit compares the starting point of the target analog signal with the starting point of the sample signal.

[0040] If the starting point of the target analog signal is different from the starting point of the sample signal, the time comparison unit determines that the target analog signal is different from the sample signal; otherwise,

[0041] The time comparison unit compares the ending point of the analog signal with the ending point of the sample signal.

[0042] If the ending point of the target analog signal is different from the ending point of the sample signal, the time comparison unit determines that the target analog signal is different from the sample signal; otherwise, it determines that the target analog signal is the same as the sample signal.

[0043] Optionally, at least one waveform comparison unit is included in the plurality of comparison units.

[0044] The comparison unit compares the target analog signal with the sample signal according to a preset comparison method, including:

[0045] The waveform comparison unit compares the waveform of the target analog signal with the waveform of the sample signal;

[0046] If the similarity between the waveform of the target analog signal and the waveform of the sample signal reaches a preset threshold, the waveform comparison unit determines that the target analog signal is the same as the sample signal; otherwise, it determines that the target analog signal is different from the sample signal.

[0047] An embodiment of the present application provides a storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the audio signal processing method described in any one of the above are implemented.

[0048] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0049] The audio signal processing device provided by the embodiment of the present application screens the picked-up audio signal through a pick-up to obtain a target analog signal containing human voice. The sample signal stored in the storage module is the user voice signal input last time. The sample signal stored in the storage module and the target analog signal are compared through a comparison module. When the target analog signal is the same as the sample signal, it indicates that the target analog signal is an echo, and the target analog signal is deleted; when the target analog signal is different from the sample signal, it indicates that the target analog signal is not an echo, and the target analog signal is stored as the sample signal in the storage module to ensure that the sample signal obtained by the storage module is a real user voice signal rather than an echo. In addition, multiple comparison units can be used to achieve multiple comparisons, further improving the judgment accuracy of the comparison module. The speaker can obtain the latest stored sample signal from the storage module, convert it into a corresponding audio signal and output it, avoiding playing the echo through the speaker again, so as to filter the echo and accurately obtain the user's voice. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 is the first structural schematic diagram of the audio processing device provided by the embodiment of the present application;

[0051] Figure 2 is the second structural schematic diagram of the audio processing device provided by the embodiment of the present application;

[0052] Figure 3 is the first structural schematic diagram of the comparison module provided by the embodiment of the present application;

[0053] Figure 4 is the second structural schematic diagram of the comparison module provided by the embodiment of the present application;

[0054] Figure 5 is the third structural schematic diagram of the audio processing device provided by the embodiment of the present application;

[0055] Figure 6 is a schematic structural diagram of a storage module provided by an embodiment of the present application;

[0056] Figure 7 is a schematic flowchart of an audio signal processing method provided by an embodiment of the present application;

[0057] Among them, 1 is a pickup; 2 is an audio signal processing chip; 3 is a speaker; 21 is a voiceprint recognition module; 22 is a comparison module; 23 is a storage module; 24 is a central processing module; 25 is an adjustment module; 220 is a comparison unit; 221 is a time comparison unit; 222 is a waveform comparison unit; 230 is a storage unit. Detailed implementation manners

[0058] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0059] Please refer to Figure 1 , Figure 1 is a first schematic structural diagram of an audio processing device provided by an embodiment of the present application.

[0060] The present application provides an audio signal processing device including a pickup 1, a speaker 3, and an audio signal processing chip 2. Among them, the pickup 1 is used to obtain an audio signal and convert the audio signal into an analog signal and then transmit it to the voiceprint recognition module 21; the speaker 3 is used to obtain the latest stored from the storage module 23 and convert the latest stored sample signal into a corresponding audio signal and then output it.

[0061] In an application scenario, the pickup 1 is a microphone, and the speaker 3 is a speaker. In order to avoid the audio signal output by the speaker 3 being obtained by the pickup 1 as much as possible, the pickup 1 is arranged on the back of the speaker 3, and the direction in which the pickup 1 obtains the audio signal is opposite to the direction in which the speaker 3 outputs the audio signal. In another application scenario, in order to avoid the audio signal output by the speaker 3 being picked up by the pickup 1 as much as possible, the pickup 1 is arranged proximally, and the speaker 3 is arranged distally, that is, the position where the pickup 1 is located is far from the position where the speaker 3 is located.

[0062] The audio signal processing chip 2 includes a voiceprint recognition module 21, a comparison module 22, a storage module 23, and a central processing module 24. Among them, the central processing module 24 is electrically connected to the pickup 1, the voiceprint recognition module 21, the comparison module 22, the storage module 23, and the speaker 3, and is used to transmit instructions to the pickup 1, the voiceprint recognition module 21, the comparison module 22, the storage module 23, and the speaker 3, so that the pickup 1, the voiceprint recognition module 21, the comparison module 22, the storage module 23, and the speaker 3 process corresponding tasks.

[0063] In this embodiment, when the pickup 1 transmits an analog signal to the voiceprint recognition module 21, the central processing module 24 issues an instruction to the voiceprint recognition analog module 21 so that the voiceprint recognition module 21 receives the analog signal and recognizes the analog signal; when the voiceprint recognition module 21 determines that the analog signal contains human voice, the central processing module 24 sends an instruction to the voiceprint recognition module 21 so that the voiceprint recognition module 21 sends the target analog signal to the comparison module 22, and the central processing module 24 sends an instruction to the storage module 23 so that the storage module 23 transmits the sample signal to the comparison module 22, and the central processing module 24 sends an instruction to the comparison module 22 so that the comparison module 22 receives the target analog signal and compares the target analog signal with the sample signal; when the comparison module 22 determines that the target analog signal is different from the sample signal, the central processing module 24 sends an instruction to the comparison module 22 so that the comparison module 22 sends the target analog signal to the storage module 23 for storage as a sample signal. After the storage is completed, the central processing module 24 calls the latest stored sample in the storage module 23 to the speaker 3; the speaker 3 converts the latest stored analog signal in the storage module 23 into a corresponding audio signal and outputs it.

[0064] The voiceprint recognition module 21 is used to determine whether the analog signal contains human voice. If so, the analog signal is transmitted to the comparison module 22 as the target analog signal. Otherwise, the analog signal is deleted.

[0065] In this embodiment, the voiceprint recognition module 21 uses the voiceprint recognition model trained in the model library to recognize whether the analog signal obtained from the pickup 1 contains human voice. If it does not contain human voice, the analog signal is deleted, so as to automatically exclude audio signals other than human voice, which is conducive to avoiding noise interference.

[0066] The storage module 23 is used to store the sample signal, and the sample signal is the user voice signal.

[0067] In an application scenario, the way to store the sample signal into the storage module 23 is that a technician pre-stores the user voice signal into the storage module 23 through other input devices; in another application scenario, the way to store the sample signal into the storage module 23 is that after the comparison module 22 compares the target analog signal with the sample signal and obtains a determination result that the target analog signal is different from the sample signal, the target analog signal is transmitted to the storage module 23 for storage as the sample signal.

[0068] The comparison module 22 is used to compare whether the target analog signal is the same as the sample signal stored in the storage module 23. If they are the same, the target analog signal is deleted; if they are different, the target analog signal is transmitted to the storage module 23 for storage as the sample signal.

[0069] In an application scenario, although the target analog signal contains human voices, the target analog signal may also be an echo generated from the audio signal output by the speaker 3. To distinguish whether the target analog signal is from the user's voice or the audio signal output by the speaker 3, the comparison module 22 compares the sample signal stored in the storage module 23 with the target analog signal. When the target analog signal is the same as the sample signal, it indicates that the target analog signal is an echo, and the target analog signal is deleted, thereby effectively removing the echo generated from the audio signal output by the speaker 3; when the target analog signal is different from the sample signal, it indicates that the target analog signal is not an echo, and the comparison module 22 transmits the target analog signal to the storage module 23 for storage as the sample signal. In this way, it can be ensured that the target analog signal obtained by the storage module 23 is the real user voice signal rather than an echo. Through the voiceprint recognition module 21 and the comparison module 22 in this embodiment, the user voice signal can be effectively obtained and the noise other than the user voice signal can be discarded.

[0070] It can be understood that through the two-stage screening of the analog signal by the voiceprint recognition module 21 and the comparison module 22 in this embodiment, it can be ensured that the analog signal obtained by the storage module 23 is the real user voice signal rather than an echo. The speaker 3 obtains the latest stored sample signal from the storage module 23, converts the latest stored sample signal into the corresponding audio signal and then outputs it, avoiding playing the echo through the speaker 3 again, thereby realizing echo filtering. In a preferred embodiment, the storage module 23 is a volatile memory. Each time the audio signal processing chip 2 is powered off and restarted, the data in the storage module 23 will be cleared. By timely clearing the storage space of the storage module 23 of the audio signal processing chip 2, the load of the audio signal processing chip 2 can be reduced, the resource utilization rate of the audio signal processing chip 2 can be improved, and the performance of the audio signal processing chip 2 can be enhanced.

[0071] However, since the data in the storage module 23 is cleared every time the audio signal processing chip 2 is powered off and restarted, this application provides an embodiment to solve the problem of how to quickly reuse the storage space in the storage module 23 after the data in the storage module 23 is cleared.

[0072] In this embodiment, the comparison module 22 is further configured to transmit the target analog signal to the storage module 23 for storage as a sample signal when the storage module 23 does not store a sample signal.

[0073] In an application scenario, when the audio signal processing device is started for the first time, the storage space of the storage module 23 of the audio signal processing chip 2 is cleared. The user inputs the user's voice through the pickup 1 for the first time. The pickup 1 converts the user's voice into an analog signal and transmits it to the voiceprint recognition module 21. The voiceprint recognition module 21 recognizes the human voice from the analog signal and transmits the analog signal to the comparison module 22 as the target analog signal. Since there is no sample signal stored in the storage module 23 at this time, the target analog signal belongs to a new sample signal. The comparison module 22 transmits the target analog signal to the storage module 23 as a sample signal, and the storage module 23 stores the sample signal. Then, the speaker 3 obtains the newly stored target analog signal from the storage module 23, converts the target analog signal into a corresponding audio signal and outputs it. The echo generated by the speaker 3 outputting the audio signal is obtained by the pickup 1 again. The pickup 1 converts the echo into an analog signal and transmits it to the voiceprint recognition module 21. The voiceprint recognition module 21 can still recognize the human voice from the analog signal and transmits the analog signal to the comparison module 22 as the target analog signal. At this time, since there is already a sample signal stored in the storage module 23, and the comparison module 22 compares the sample signal with the target analog signal and determines that the sample signal is the same as the target analog signal (because the target analog signal is an echo, so the sample signal is the same as the target analog signal), the comparison module 22 deletes the target analog signal.

[0074] Compared with the method of manually retransmitting sample signals to the storage module 23 for storage, this embodiment more quickly transmits the target analog signal to the storage module 23 for storage as a sample signal through the comparison module 22, which is beneficial to accelerating the efficiency of storing the sample signal into the storage module 23.

[0075] To improve the discrimination ability of the comparison module 22, this application provides an embodiment as follows:

[0076] Please refer to Figure 2 , Figure 2 which is the second structural schematic diagram of the audio processing device provided by the embodiment of this application.

[0077] The audio signal processing device includes a pickup 1, a speaker 3, and an audio signal processing chip 2. The audio signal processing chip 2 includes a voiceprint recognition module 21, a comparison module 22, and a storage module 23.

[0078] On this basis, the comparison module 22 includes a plurality of comparison units 220; each comparison unit 220 compares the target analog signal with the sample signal according to a preset comparison method; if each comparison unit 220 determines that the target analog signal is different from the sample signal, then the comparison module 22 determines that the target analog signal is different from the sample signal, otherwise, the comparison module 22 determines that the target analog signal is the same as the sample signal.

[0079] In this embodiment, the comparison module 22 can implement multiple judgments on the analog signal and the sample signal through a plurality of comparison units 220. Compared with the comparison result obtained by a single comparison unit 220 for judging whether the analog signal is the same as the sample signal, the accuracy and accuracy of the comparison result obtained by the collaborative judgment of a plurality of comparison units 220 on whether the analog signal is the same as the sample signal are higher.

[0080] It should be noted that the researchers of this application found that a common situation that interferes with the audio signal processing device from obtaining the user's voice signal is that the pickup 1 often obtains an echo of the audio signal output by the speaker 3 as the audio signal, and converts the echo of the audio signal output by the speaker 3 into an analog signal. In order to accurately identify this analog signal as noise, an embodiment of this application is provided as follows:

[0081] Please refer to Figure 3 , Figure 3 which is the first structural schematic diagram of the comparison module 22 provided by the embodiment of this application.

[0082] At least one time comparison unit 221 is included in the plurality of comparison units 220.

[0083] The time comparison unit 221 compares the starting point of the target analog signal with the starting point of the sample signal; if the starting point of the target analog signal is different from the starting point of the sample signal, then the time comparison unit 221 determines that the target analog signal is different from the sample signal; otherwise, the time comparison unit 221 compares the ending point of the target analog signal with the ending point of the sample signal; if the ending point of the target analog signal is different from the ending point of the sample signal, then the time comparison unit 221 determines that the target analog signal is different from the sample signal, otherwise the time comparison unit 221 determines that the target analog signal is the same as the sample signal.

[0084] In this embodiment, the analog signal converted by the pickup 1 may be converted from the user's voice or may be an echo generated by the audio signal output by the speaker 3.

[0085] In an application scenario, the pickup 1 acquires the user's voice and converts the user's voice into an analog signal. The voiceprint recognition module 21 recognizes the analog signal, determines that the analog signal contains human voice, and transmits it as a target analog signal to the comparison module 22. Subsequently, the comparison module 22 acquires the target analog signal and a sample signal from the storage module 23, and compares the target analog signal with the sample signal. Through the time comparison unit 221 in the comparison module 22, the starting point of the target analog signal is compared with the starting point of the sample signal. Since the target analog signal is converted from the user's voice, the target analog signal is currently generated and is different from any previous sample signal. The starting point and the ending point of the target analog signal are different from the starting point and the ending point of the sample signal. Therefore, the time comparison unit 221 can easily determine that the starting point and the ending point of the target analog signal are different from those of the sample signal, and thus determines that the target analog signal is different from the sample signal.

[0086] In another application scenario, the pickup 1 acquires the echo generated by the audio signal output by the speaker 3 and converts the echo into an analog signal. Since the speaker 3 can also reproduce human voice, the voiceprint recognition module 21 also recognizes human voice from the analog signal and transmits the analog signal as a target analog signal to the comparison module 22 for further judgment. Through the time comparison unit 221 of the comparison module 22, the starting point of the target analog signal is compared with the starting point of the sample signal, and the ending point of the target analog signal is compared with the ending point of the sample signal. Since the sample signal is from the user's previous voice input through the pickup 1, and the target analog signal is from the echo of the user's first voice input through the pickup 1, that is, the sample signal and the target analog signal are from the same audio signal, the starting point of the target analog signal is the same as the starting point of the sample signal, and the ending point of the target analog signal is the same as the ending point of the sample signal. Thus, the time comparison unit 221 can determine that the target analog signal is the same as the sample signal.

[0087] It can be understood that in this embodiment, the time comparison module 22 can quickly and accurately determine whether the sample signal and the target analog signal are from the same audio signal, thereby improving the ability of the comparison module 22 to distinguish the user's voice from the echo.

[0088] It should be noted that the researchers of this application found that in some cases, the starting point of the analog signal converted from the user's voice is the same as the starting point of the sample signal, and the ending point of the analog signal is also the same as the ending point of the sample signal. To accurately identify the user's voice, another embodiment of this application is provided as follows:

[0089] Please refer to Figure 4 , Figure 4It is the second structural schematic diagram of the comparison module 22 provided by the embodiments of the present application.

[0090] At least one of the multiple comparison units 220 is a waveform comparison unit 222.

[0091] The waveform comparison unit 222 compares the waveform of the target analog signal with the waveform of the sample signal; if the similarity between the waveform of the target analog signal and the waveform of the sample signal reaches a preset threshold, the waveform comparison unit 222 determines that the target analog signal is the same as the sample signal, otherwise the waveform comparison unit determines that the target analog signal is different from the sample signal.

[0092] It can be understood that generally, the speaker 3 is far from the microphone (pickup 1), and the audio signal output by the speaker 3 is continuously reflected and transmitted to the microphone, that is, the echo is collected by the microphone. At this time, the amplitude of the audio signal collected by the microphone will be attenuated compared with the amplitude of the audio signal just output by the speaker 3. Based on this principle, in this embodiment, the waveform comparison unit 222 of the comparison module 22 compares the waveform of the analog signal converted from the audio signal output by the speaker 3 with the waveform of the sample signal, and can effectively identify the difference between the waveform of the analog signal and the waveform of the sample signal.

[0093] However, since the attenuation degree of the amplitude of the audio signal output from the speaker 3 varies greatly, the error between the waveform of the analog signal converted from the audio signal output by the speaker 3 and the waveform of the sample signal will also be very large. Based on this, in this embodiment, when the waveform comparison unit 222 determines whether the target analog signal is the same as the sample signal, it only needs to determine whether the similarity between the waveform of the target analog signal and the waveform of the sample signal reaches a preset threshold. If so, it determines that the target analog signal is the same as the sample signal, otherwise, it determines that the target analog signal is different from the sample signal. Thus, by determining whether the similarity between the waveform of the target analog signal and the waveform of the sample signal reaches a preset threshold through the waveform comparison unit 222, the error tolerance rate of the waveform comparison unit 222 for determining whether the target analog signal is the same as the sample signal can be improved. And even in some cases where the starting point of the analog signal converted from the user's voice is the same as the starting point of the sample signal, and the ending point of the analog signal is also the same as the ending point of the sample signal, the waveform comparison unit 222 can identify the difference between the waveform of the target analog signal and the waveform of the sample signal, thereby determining that the target analog signal is different from the sample signal, and further effectively improving the accuracy of the comparison module 22 for determining the target analog signal and the sample signal.

[0094] The researchers found that the waveform comparison unit 222 compares the waveform of the analog signal with the waveform of the sample signal. However, the preset threshold for judging similarity is fixed. In some application scenarios, the audio signal obtained by the pickup 1 is amplified by an amplifier. After the pickup 1 converts the amplified audio signal into an amplified analog signal, the pickup 1 transmits this amplified analog signal to the waveform comparison unit 222. The waveform comparison unit 222 determines whether the similarity between the waveform of the amplified analog signal and the waveform of the sample signal reaches the preset threshold. Since the waveform of this amplified analog signal is very different from the waveform of the sample signal, it affects the accuracy of the waveform comparison unit 222 in judging whether the similarity reaches the preset threshold, resulting in an error in the judgment of the waveform comparison unit 222.

[0095] To solve the above technical problems, the present application provides another embodiment as follows:

[0096] Please refer to Figure 5 , Figure 5 which is the third structural schematic diagram of the audio processing device provided by the embodiment of the present application.

[0097] In this embodiment, the audio signal processing chip 2 further includes an adjustment module 25.

[0098] Among them, the adjustment module 25 is used to obtain the sample signal from the storage module 230 when the time comparison unit 221 determines that the target analog signal is the same as the sample signal, and compare the amplitude of the target analog signal with the amplitude of the sample signal; if the amplitude of the target analog signal is greater than the amplitude of the sample signal, then reduce the amplitude of the target analog signal; if the amplitude of the target analog signal is less than the amplitude of the sample signal, then increase the amplitude of the target analog signal; and transmit the adjusted target analog signal to the waveform comparison unit.

[0099] In an application scenario, the audio signal obtained by the pickup 1 is an amplified echo. After the pickup 1 converts the amplified echo into an amplified analog signal, even if the voiceprint recognition module 21 transmits this analog signal to the waveform comparison unit 222 as the target analog signal, due to the huge difference between the waveform of the amplified analog signal and the waveform of the sample signal, it affects the accuracy of the waveform comparison unit 222 in judging whether the similarity between the waveform of the amplified analog signal and the waveform of the sample signal reaches the preset threshold, resulting in the waveform comparison unit 222 being prone to make mistakes in judgment. In view of this situation, in this embodiment, after the pickup 1 obtains the amplified audio signal and converts it into an amplified analog signal, the pickup 1 transmits this analog signal to the voiceprint recognition module 21 for recognition. After being determined as the target analog signal, it is transmitted to the time comparison unit 221 for comparison. The time comparison unit 221 compares the starting point of this target analog signal with the starting point of the sample signal and the ending point of this target analog signal with the ending point of the sample signal. After obtaining the result that this target analog signal is the same as the sample signal, it transmits this target analog signal to the adjustment module 25. After obtaining this target analog signal, the adjustment module 25 compares the amplitude of this target analog signal with the amplitude of the sample signal. By comparison, it can be seen that the amplitude of this target analog signal is greater than the amplitude of the sample signal. Based on this, the adjustment module 25 adjusts this target analog signal, that is, reduces the amplitude of this target analog signal, which is beneficial to reducing the difference between the waveform of this target analog signal and the waveform of the sample signal. Then the adjustment module 25 transmits the adjusted target analog signal to the waveform comparison unit 222. The waveform comparison unit 222 compares this adjusted target analog signal with the sample signal. When the similarity between the waveform of this adjusted target analog signal and the waveform of the sample signal reaches the preset threshold, the waveform comparison unit 222 determines that this adjusted target analog signal is the same as the sample signal.

[0100] In another application scenario, due to reflection, the amplitude of the echo will be greatly reduced. If the pickup 1 transmits the analog signal with reduced amplitude converted from this echo to the waveform comparison unit 222, it will also affect the accuracy of the waveform comparison unit 222 in judging whether the similarity reaches the preset threshold. In view of this situation, in this embodiment, after the pickup 1 obtains the echo with reduced amplitude and converts it into an analog signal with reduced amplitude, the pickup 1 transmits this analog signal to the voiceprint recognition module 21 for recognition. After being determined as the target analog signal, it is transmitted to the time comparison unit 221 for comparison. The time comparison unit 221 compares the starting point of this target analog signal with the starting point of the sample signal and the ending point of this target analog signal with the ending point of the sample signal. After obtaining the result that this target analog signal is the same as the sample signal, it transmits this target analog signal to the adjustment module 25.

[0101] After the adjustment module 25 obtains the target analog signal, it compares the amplitude of the target analog signal with the amplitude of the sample signal. Through comparison, it can be obtained that the amplitude of the target analog signal is smaller than the amplitude of the sample signal. Based on this, the adjustment module 25 adjusts the target analog signal to increase the amplitude of the target analog signal, which is beneficial to reducing the waveform difference between the target analog signal and the sample signal. Then, the adjustment module 25 transmits the adjusted target analog signal to the waveform comparison unit 222. The waveform comparison unit 222 compares the adjusted target analog signal with the sample signal. When the similarity between the waveform of the adjusted target analog signal and the waveform of the sample signal reaches the preset threshold, the waveform comparison unit 222 determines that the adjusted target analog signal is the same as the sample signal.

[0102] It can be understood that in some application scenarios, the amplitude of the target analog signal obtained by the adjustment module 25 is the same as the amplitude of the sample signal. In this case, the adjustment module 25 does not need to adjust the target analog signal and transmits the target analog signal to the waveform comparison unit.

[0103] It can be seen that in this embodiment, the adjustment module 25 adjusts the target analog signal, which is beneficial to reducing the large difference between the waveform of the target analog signal and the waveform of the sample signal, and further avoiding the influence on the comparison between the waveform of the target analog signal and the waveform of the sample signal by the waveform comparison unit 222, and improving the judgment accuracy of the waveform comparison unit 222.

[0104] It should be noted that the researchers found that the speed of finding the sample signal from the storage module 23 will affect the efficiency of the comparison module 22 in comparing the target analog signal with the sample signal. In order to improve the speed of finding the sample signal from the storage module 23, the following embodiments are provided in this application:

[0105] Please refer to Figure 6 , Figure 6 which is the structural schematic diagram of the storage module 23 provided by the embodiment of this application.

[0106] In this embodiment, the storage module 23 includes a plurality of storage units 230, and each storage unit 230 stores a corresponding sample signal. Each storage unit 230 is numbered in sequence, and each number is recorded in the mapping table. Among them, the mapping table is stored in one of the storage units 230 of the storage module 23.

[0107] The storage module 23 searches for the corresponding storage unit 230 according to the number recorded in the mapping table; the storage module 23 obtains the corresponding sample signal from the found storage unit 230 and transmits the corresponding sample signal to the comparison module 22.

[0108] It can be understood that when the storage module 23 stores multiple sample signals, extracting the multiple sample signals one by one will slow down the extraction speed of the sample signals and reduce the processing efficiency of the comparison module 22. In this regard, the storage module 23 of this embodiment stores multiple different sample signals through multiple storage units 230. Each storage unit 230 is numbered in sequence, and each number is recorded in the mapping table. Among them, the mapping table is stored in one of the storage units 230 of the storage module 23. For example, the storage unit 230 numbered 1 stores the first sample signal, the storage unit 230 numbered 2 stores the second sample signal, the storage unit 230 numbered 3 stores the third sample signal, and so on. When it is necessary to extract the sample signal, the storage module 23 searches for the corresponding storage unit 230 according to the number recorded in the mapping table, and obtains the corresponding sample signal from the found storage unit 230.

[0109] In an application scenario, the voice just output by the user is converted into a new analog signal by the pickup 1, and after the new analog signal is recognized by the voiceprint recognition module 21 as containing human voices, it is transmitted to the comparison module 22 for comparison. The comparison module 22 needs to compare the new analog signal with the sample signal. Therefore, the storage module 23 quickly finds the storage unit 230 with the latest number according to the mapping table and extracts the latest stored sample signal from this storage unit 230 (the new analog signal and the latest stored sample signal are most likely to be the same), so that the comparison module 22 compares the new analog signal with the latest stored sample signal. It can be seen that this embodiment effectively speeds up the extraction speed of the sample signal and is conducive to improving the comparison efficiency of the comparison module 22.

[0110] In one embodiment, the storage module 23 is used to delete the sample signal when the comparison module 22 determines that the analog signal is the same as the sample signal.

[0111] It can be understood that after the comparison module 22 compares the analog signal converted from the echo with the sample signal and obtains the comparison result, if the analog signal converted from the echo is the same as the sample signal, it indicates that the audio signal identical to the echo has been acquired by the audio signal processing device, and the chance for the audio signal processing device to acquire the echo or an audio signal identical to the echo again in the future is very small. The comparison module 22 no longer needs the sample signal for comparison. Therefore, in order to save storage space, the storage module 23 can delete the sample signal that is the same as the analog signal converted from the echo. In this way, the storage space of the storage module 23 can be saved, and at the same time, since the deletion of the sample signal reduces the total amount of sample signals stored in the storage module 23, the speed of the storage module 23 to search for and extract the sample signal is also accelerated.

[0112] This application provides an audio signal processing method. Please refer to Figure 7 , Figure 7It is a schematic flowchart of the audio signal processing method provided by an embodiment of the present application.

[0113] This audio signal processing method is used for an audio signal processing device. The audio signal processing device includes a pickup 1, a speaker 3, and an audio signal processing chip 2. The audio signal processing chip 2 includes a voiceprint recognition module 21, a comparison module 22, and a storage module 23. And the audio signal processing method includes the following steps:

[0114] S101. The pickup 1 acquires an audio signal, converts the audio signal into an analog signal, and then transmits it to the voiceprint recognition module.

[0115] S102. The voiceprint recognition module detects whether the analog signal contains human voice. If so, it transmits the analog signal as a target analog signal to the comparison module. Otherwise, it deletes the analog signal.

[0116] S103. The comparison module compares whether the target analog signal is the same as the sample signal stored in the storage module. If they are the same, it deletes the target analog signal. If they are different, it transmits the target analog signal as a sample signal to the storage module for storage.

[0117] Among them, the storage module is used to store sample signals, and the sample signals are user voice signals.

[0118] S104. The speaker acquires the latest stored sample signal from the storage module, converts the latest stored sample signal into a corresponding audio signal, and then outputs it.

[0119] In one embodiment, the comparison module 22 includes a plurality of comparison units 220.

[0120] The comparison module compares the sample signal with the target analog signal, including: each comparison unit compares the target analog signal with the sample signal according to a preset comparison method; if each comparison unit determines that the target analog signal is different from the sample signal, then the comparison module determines that the target analog signal is different from the sample signal. Otherwise, the comparison module determines that the target analog signal is the same as the sample signal.

[0121] In one embodiment, when the storage module 23 does not store sample signals, the comparison module 22 transmits the target analog signal as a sample signal to the storage module 23 for storage.

[0122] In one embodiment, at least one time comparison unit 221 is included in the plurality of comparison units 220.

[0123] The comparison unit compares the target analog signal with the sample signal according to a preset comparison method, including: the time comparison unit obtains the starting point of the target analog signal and the starting point of the sample signal and compares them; if the starting point of the target analog signal is different from the starting point of the sample signal, the time comparison unit determines that the target analog signal is different from the sample signal; otherwise, the time comparison unit obtains the ending point of the target analog signal and the ending point of the sample signal and compares them; if the ending point of the target analog signal is different from the ending point of the sample signal, the time comparison unit determines that the target analog signal is different from the sample signal, otherwise the time comparison unit determines that the target analog signal is the same as the sample signal.

[0124] In one embodiment, at least one waveform comparison unit 222 is included in the multiple comparison units 220.

[0125] The comparison unit compares the target analog signal with the sample signal according to a preset comparison method, including: the waveform comparison unit compares the waveform of the target analog signal with the waveform of the sample signal; if the similarity between the waveform of the target analog signal and the waveform of the sample signal reaches a preset threshold, the waveform comparison unit determines that the target analog signal is the same as the sample signal, otherwise the waveform comparison unit determines that the target analog signal is different from the sample signal.

[0126] In one embodiment, the audio signal processing chip 2 further includes an adjustment module 25; the audio signal processing method further includes: when the time comparison unit determines that the target analog signal is the same as the sample signal, the adjustment module obtains the sample signal from the storage module, and compares the amplitude of the target analog signal with the amplitude of the sample signal; if the amplitude of the target analog signal is greater than the amplitude of the sample signal, the amplitude of the target analog signal is decreased; if the amplitude of the target analog signal is less than the amplitude of the sample signal, the amplitude of the target analog signal is increased; the adjusted target analog signal is transmitted to the waveform comparison unit.

[0127] In one embodiment, the storage module 23 includes multiple storage units 230, and each storage unit 230 stores a corresponding sample signal.

[0128] The storage module transmits the sample signal to the comparison module, including: the storage module numbers in sequence each storage unit storing a corresponding sample signal, records each number in a mapping table, wherein the mapping table is stored in one of the storage units of the storage module; the storage module searches for the corresponding storage unit according to the number recorded in the mapping table; the storage module searches for the corresponding storage unit according to the number recorded in the mapping table; the storage module obtains the corresponding sample signal from the found storage unit and transmits the corresponding sample signal to the comparison module.

[0129] In one embodiment, the audio signal processing method further includes the following steps:

[0130] When the comparison module determines that the target analog signal is the same as the sample signal, the storage module deletes the sample signal.

[0131] In a preferred embodiment, the storage module 23 is a volatile memory. Each time the audio signal processing chip 2 is powered off and restarted, the data in the storage module 23 will be cleared.

[0132] An embodiment of the present application provides a storage medium, on which a computer program is stored.

[0133] In this embodiment, when the computer program stored in the storage medium is executed by a processor, the steps of the audio signal processing method in any of the above embodiments are implemented.

[0134] The audio signal processing device, method and storage medium provided by the embodiments of the present application, wherein the audio signal processing method is used for the audio signal processing device, the audio signal processing device includes a pickup 1, a speaker 3 and an audio signal processing chip 2, the audio signal processing chip 2 includes a voiceprint recognition module 21, a comparison module 22, a storage module 23 and a comparison module 22, and the audio signal processing method includes the following steps: the pickup acquires an audio signal and converts the audio signal into an analog signal and then transmits it to the voiceprint recognition module; the voiceprint recognition module determines whether the analog signal contains human voices, if so, the analog signal is used as a target analog signal and transmitted to the comparison module, otherwise the analog signal is deleted, so as to eliminate noises other than human voices and initially screen out human voices, which is beneficial to accurately collect user voice signals; the comparison module 22 compares the target analog signal with the sample signal stored in the storage module, if the comparison module 22 determines that the target analog signal is different from the sample signal, the target analog signal is transmitted to the storage module 23 for storage, otherwise, the target analog signal is deleted.

[0135] It can be understood that since the sample signal stored in the storage module is the user voice signal input last time, if the target analog signal is an echo, the target model signal must be the same as the sample signal. On the contrary, if the target analog signal is a newly input user voice signal, the target analog signal is different from the sample signal. Therefore, the comparison module 22 compares the sample signal stored in the storage module 23 with the target analog signal. When the target analog signal is the same as the sample signal, it indicates that the target analog signal is an echo, and the comparison module 22 deletes the target analog signal. When the target analog signal is different from the sample signal, it indicates that the target analog signal is not an echo, and the comparison module 22 transmits the target analog signal to the storage module 23 as a sample signal for storage. In this way, it can be ensured that the target analog signal obtained by the storage module 23 is a real user voice signal rather than an echo. The speaker 3 obtains the latest stored sample signal from the storage module 23, converts the latest stored sample signal into a corresponding audio signal and outputs it, avoiding playing the echo through the speaker again, thereby realizing echo filtering.

[0136] It should be understood that each part of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following technologies well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0137] In addition, each functional unit in various embodiments of the present application can be integrated in a processing module, or each unit can exist physically alone, or two or more units can be integrated in one module. The above integrated module can be implemented in the form of hardware or in the form of a software functional module. When the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. The storage medium can be a read-only memory, a magnetic disk or an optical disc, etc.

[0138] The above is the preferred embodiment of the present application. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present application, several improvements and refinements can be made, and these improvements and refinements are also regarded as the protection scope of the present application.

Claims

1. An audio signal processing device, characterized in that, It includes a pickup, a speaker, and an audio signal processing chip. The audio signal processing chip includes a voiceprint recognition module, a comparison module, an adjustment module, and a storage module; The pickup is configured to obtain an audio signal, convert the audio signal into an analog signal, and transmit the analog signal to the voiceprint recognition module; The voiceprint recognition module is configured to detect whether the analog signal contains human voice. If so, the analog signal is used as a target analog signal and transmitted to the comparison module. Otherwise, the analog signal is deleted; The storage module is configured to store a plurality of sample signals, and the sample signals are user voice signals; The comparison module is configured to compare whether the target analog signal is the same as the sample signals stored in the storage module. If they are the same, the target analog signal is deleted. If they are different, the target analog signal is transmitted to the storage module as a sample signal for storage; And when the storage module does not store the sample signal, the target analog signal is transmitted to the storage module as a sample signal for storage; The comparison module includes a plurality of comparison units. At least one time comparison unit and at least one waveform comparison unit are included in the plurality of comparison units. Each comparison unit compares the target analog signal with the sample signal according to a preset comparison method. If each comparison unit determines that the analog signal is different from the sample signal, the comparison module determines that the target analog signal is different from the sample signal. Otherwise, the comparison module determines that the target analog signal is the same as the sample signal; The adjustment module is configured to compare the amplitude of the target analog signal with the amplitude of the sample signal when the time comparison unit determines that the target analog signal is the same as the sample signal. If the amplitude of the target analog signal is greater than the amplitude of the sample signal, the amplitude of the target analog signal is decreased. If the amplitude of the target analog signal is less than the amplitude of the sample signal, the amplitude of the target analog signal is increased, and the adjusted target analog signal is transmitted to the waveform comparison unit; The speaker is configured to obtain the latest stored sample signal from the storage module, convert the latest stored sample signal into a corresponding audio signal, and output it.

2. The audio signal processing apparatus according to claim 1, wherein The comparison unit compares the target analog signal with the sample signal according to a preset comparison method, including: The time comparison unit compares the starting point of the target analog signal with the starting point of the sample signal; If the starting point of the target analog signal is different from the starting point of the sample signal, the time comparison unit determines that the target analog signal is different from the sample signal. Otherwise, The time comparison unit compares the ending point of the analog signal with the ending point of the sample signal; If the ending point of the target analog signal is different from the ending point of the sample signal, the time comparison unit determines that the target analog signal is different from the sample signal. Otherwise, it is determined that the target analog signal is the same as the sample signal.

3. The audio signal processing apparatus according to claim 2, wherein The comparison unit compares the target analog signal with the sample signal according to a preset comparison method, including: The waveform comparison unit compares the waveform of the target analog signal with the waveform of the sample signal; If the similarity between the waveform of the target analog signal and the waveform of the sample signal reaches a preset threshold, the waveform comparison unit determines that the target analog signal is the same as the sample signal; otherwise, it determines that the target analog signal is different from the sample signal.

4. The audio signal processing device according to claim 3, wherein The storage module includes a plurality of storage units, each storage unit stores a corresponding sample signal, each storage unit is numbered in sequence, and each number is recorded in a mapping table, wherein the mapping table is stored in one of the storage units of the storage module; The storage module transmits the sample signal to the comparison module, including: The storage module looks up the corresponding storage unit according to the number recorded in the mapping table; The storage module obtains the corresponding sample signal from the found storage unit and transmits the corresponding sample signal to the comparison module.

5. The audio signal processing device according to claim 1, characterized in that, The storage module is a volatile memory. After each power-off and restart of the audio signal processing chip, the data in the storage module will be cleared.

6. An audio signal processing method, characterized in that, For an audio signal processing device, the audio signal processing device includes a pickup, a speaker, and an audio signal processing chip. The audio signal processing chip includes a voiceprint recognition module, a comparison module, an adjustment module, and a storage module, and the audio signal processing method includes the following steps: The pickup obtains an audio signal and converts the audio signal into an analog signal and then transmits it to the voiceprint recognition module; The voiceprint recognition module detects whether the analog signal contains human voices. If so, it transmits the analog signal as a target analog signal to the comparison module; otherwise, it deletes the analog signal; The comparison module compares whether the target analog signal is the same as the sample signal stored in the storage module. If they are the same, it deletes the target analog signal; if they are different, it transmits the target analog signal to the storage module for storage as a sample signal; And for when the storage module does not store the sample signal, transmitting the target analog signal to the storage module for storage as a sample signal; Wherein, the storage module is used to store a plurality of sample signals, and the sample signals are user voice signals; The comparison module includes a plurality of comparison units. At least one time comparison unit and at least one waveform comparison unit are included in the plurality of comparison units. Each comparison unit compares the target analog signal with the sample signal according to a preset comparison method. If each comparison unit determines that the analog signal is different from the sample signal, the comparison module determines that the target analog signal is different from the sample signal; otherwise, the comparison module determines that the target analog signal is the same as the sample signal; The adjustment module is configured to compare the amplitude of the target analog signal with the amplitude of the sample signal when the time comparison unit determines that the target analog signal is the same as the sample signal. If the amplitude of the target analog signal is greater than the amplitude of the sample signal, the amplitude of the target analog signal is decreased; if the amplitude of the target analog signal is less than the amplitude of the sample signal, the amplitude of the target analog signal is increased, and the adjusted target analog signal is transmitted to the waveform comparison unit. The loudspeaker obtains the latest stored sample signal from the storage module, converts the latest stored sample signal into a corresponding audio signal, and outputs the audio signal.

7. The audio signal processing method according to claim 6, wherein The comparison unit compares the target analog signal with the sample signal according to a preset comparison method, including: The time comparison unit compares the starting point of the target analog signal with the starting point of the sample signal. If the starting point of the target analog signal is different from the starting point of the sample signal, the time comparison unit determines that the target analog signal is different from the sample signal; otherwise, The time comparison unit compares the ending point of the analog signal with the ending point of the sample signal. If the ending point of the target analog signal is different from the ending point of the sample signal, the time comparison unit determines that the target analog signal is different from the sample signal; otherwise, it determines that the target analog signal is the same as the sample signal.

8. The audio signal processing method according to claim 7, wherein The comparison unit compares the target analog signal with the sample signal according to a preset comparison method, including: The waveform comparison unit compares the waveform of the target analog signal with the waveform of the sample signal. If the similarity between the waveform of the target analog signal and the waveform of the sample signal reaches a preset threshold, the waveform comparison unit determines that the target analog signal is the same as the sample signal; otherwise, it determines that the target analog signal is different from the sample signal.

9. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, the steps of the audio signal processing method according to any one of claims 6 to 8 are implemented.

Images