A method for suppressing misidentification

Abstract

The present invention relates to a method for suppressing misidentification. The present invention determines whether the recognition result of a directional sound pickup output signal is caused by interference from a non-sound pickup area based on the direction of arrival and the energy suppression ratio, thereby suppressing misidentification caused by strong interference to a great extent. By combining the direction of arrival (DOA) information of the recognition sentence and the energy suppression ratio of the directional sound pickup to the recognition signal, the misidentification of the entire system is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of audio recognition, and in particular to a method for suppressing misrecognition. Background Art

[0002] In the prior art, in some specific scenarios, beamforming technology can be used to obtain target signals in a fixed direction (called directional sound pickup technology) while suppressing noise and interference in other directions. However, when there is strong interference in the non-sound pickup area, the misidentification phenomenon is more serious. Summary of the invention

[0003] The present invention provides a misidentification suppression method to solve the misidentification problem of directional sound pickup technology in a strong interference environment and further improve the directional sound pickup function.

[0004] The technical solution of the present invention to solve the above technical problems is as follows:

[0005] According to one aspect of an embodiment of the present invention, a method for suppressing misidentification is provided, comprising:

[0006] Step 1: The microphone array receives the signal and performs short-time Fourier transform;

[0007] Step 2: Estimate the direction of arrival of each frame of data based on the signal after short-time Fourier transform θ t , and select any microphone signal to calculate the energy of each frame of data, recorded as the initial energy And estimate the direction of arrival θ t and initial energy cache;

[0008] Step 3: Process the short-time Fourier transformed signal through directional sound pickup;

[0009] Step 4: Calculate the energy of each frame of data after directional sound pickup processing, and record it as output energy and cache;

[0010] Step 5: Send the processed data of directional sound pickup to the recognition engine, and obtain the start and end time point information of the recognition signal, which are recorded as the start time t1 and the end time t2 respectively;

[0011] Step 6: Calculate the initial energy E1 of the recognition signal and its corresponding output energy E2 according to the start and end time point information, and calculate the recognition signal energy ratio R;

[0012] Step 7: Count the time period θ from the start time t1 to the end time t2 t The mode of the identification signal is recorded as the direction of arrival of θ mode , calculate the deviation θ between the wave arrival direction and the directional sound pickup target direction error ;

[0013] Step 8: The judgment system makes a judgment based on the energy ratio and arrival direction of the identification signal, and determines whether the identification signal is interference based on the judgment result.

[0014] Preferably, the initial energy is

[0015] Preferably, the output energy is

[0016] Preferably, the signal energy ratio is

[0017] Preferably, the direction deviation is θ error =|θ mode -θ in |.

[0018] Preferably, the decision system is set to a strong suppression mode and / or a weak suppression mode.

[0019] Preferably, the strong suppression mode is implemented as the direction of arrival deviation θ error Greater than the angle deviation threshold θ th Or the recognition signal energy ratio R is less than the threshold R th , then the identification signal is a misidentification caused by interference.

[0020] Preferably, the weak suppression mode is implemented as the direction of arrival deviation θ error Greater than the threshold θ th And the recognition signal energy ratio R is greater than the threshold R th , then the identification signal is a misidentification caused by interference.

[0021] Preferably, the method further includes step nine: if the identification signal is determined to be interference, the system does not output any identification result.

[0022] According to another aspect of an embodiment of the present invention, a storage medium is provided, wherein the storage medium includes a stored program, wherein when the program is executed, the device where the storage medium is located is controlled to execute the aforementioned misrecognition suppression method.

[0023] Therefore, the present invention determines whether the recognition result of the directional sound pickup output signal is caused by interference in the non-sound pickup area according to the wave direction of arrival and the energy suppression ratio, and can suppress the misrecognition caused by strong interference to a great extent. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 is a schematic diagram of a microphone array of the present invention;

[0025] Figure 2 The present invention provides an algorithm flow chart. DETAILED DESCRIPTION

[0026] The principles and features of the present invention are described below in conjunction with the accompanying drawings. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention.

[0027] The embodiment of the present invention provides a method for suppressing misrecognition, so as to solve the misrecognition problem of directional sound pickup technology in a strong interference environment and further improve the directional sound pickup function.

[0028] Directional sound pickup can be achieved through beamforming technology, but when there is strong interference in the environment, the ability of directional sound pickup to suppress strong interference is relatively limited, so there will still be the problem of misidentification, especially when there is only strong interference, the misidentification will be more obvious. By combining the direction of arrival (DOA) information of the recognition sentence and the energy suppression ratio of directional sound pickup to the recognition signal, the misidentification of the entire system can be reduced.

[0029] Set the target direction of the sound pickup to θ in , the specific steps are as follows:

[0030] Step 1: The microphone array receives the signal and performs short-time Fourier transform;

[0031] Step 2: Estimate the direction of arrival of each frame of data based on the signal after short-time Fourier transform θ t , and select any microphone signal to calculate the energy of each frame of data, recorded as the initial energy And estimate the direction of arrival θ t and initial energy cache;

[0032] Step 3: Process the short-time Fourier transformed signal through directional sound pickup;

[0033] Step 4: Calculate the energy of each frame of data after directional sound pickup processing, and record it as output energy and cache;

[0034] Step 5: Send the processed data of directional sound pickup to the recognition engine, and obtain the start and end time point information of the recognition signal, which are recorded as the start time t1 and the end time t2 respectively;

[0035] Step 6: Calculate the initial energy E1 of the recognition signal and its corresponding output energy E2 according to the start and end time point information, and calculate the recognition signal energy ratio R;

[0036] Step 7: Count the time period θ from the start time t1 to the end time t2 t The mode of the identification signal is recorded as the direction of arrival of θ mode , calculate the deviation θ between the wave arrival direction and the directional sound pickup target direction error ;

[0037] Step 8: The judgment system makes a judgment based on the energy ratio and arrival direction of the identification signal, and determines whether the identification signal is interference based on the judgment result.

[0038] If the recognition signal is within the pickup area of ​​directional pickup, the energy of the directional pickup output signal and the initial signal energy will only change slightly. On the contrary, if it is outside the pickup area, its energy will be attenuated more. At the same time, whether its wave direction is within the pickup area is also an important reference for judging whether it is a signal or interference. The present invention judges whether the recognition result of the directional pickup output signal is caused by interference from the non-pickup area based on the wave direction and energy suppression ratio, which can greatly suppress the misrecognition caused by strong interference.

[0039] In some embodiments, the initial energy is

[0040] In some embodiments, the output energy is

[0041] In some embodiments, the signal energy ratio is

[0042] In some embodiments, the directional deviation is θ error =|θ mode -θ in |.

[0043] In some embodiments, the decision system is set to a strong suppression mode and / or a weak suppression mode.

[0044] In some embodiments, the strong suppression mode is implemented as a direction of arrival deviation θ error Greater than the angle deviation threshold θ th Or the recognition signal energy ratio R is less than the threshold R th , then the identification signal is a misidentification caused by interference.

[0045] In some embodiments, the weak suppression mode is implemented as a direction of arrival deviation θ error Greater than the threshold θ th And the recognition signal energy ratio R is greater than the threshold R th , then the identification signal is a misidentification caused by interference.

[0046] In some embodiments, step nine is also included: if the identification signal is determined to be interference, the system does not output any identification result.

[0047] A method for suppressing misidentification provided by an embodiment of the present invention can be implemented in the form of a software function module and sold or used as an independent product, and can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0048] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for suppressing misidentification, characterized in that: include: Step 1: The microphone array receives the signal and performs short-time Fourier transform; Step 2: Estimate the direction of arrival of each frame of data based on the signal after short-time Fourier transform θ t , and select any microphone signal to calculate the energy of each frame of data, recorded as the initial energy And estimate the direction of arrival θ t and initial energy cache; Step 3: Process the short-time Fourier transformed signal through directional sound pickup; Step 4: Calculate the energy of each frame of data after directional sound pickup processing, and record it as output energy and cache; Step 5: Send the processed data of directional sound pickup to the recognition engine, and obtain the start and end time point information of the recognition signal, which are recorded as the start time t1 and the end time t2 respectively; Step 6: Calculate the initial energy E1 of the recognition signal and its corresponding output energy E2 according to the start and end time point information, and calculate the recognition signal energy ratio R; Step 7: Count the time period θ from the start time t1 to the end time t2 t The mode of the identification signal is recorded as the direction of arrival of θ mode , calculate the deviation θ between the wave arrival direction and the directional sound pickup target direction error ; Step 8: The judgment system makes a judgment based on the energy ratio and the direction of arrival of the recognition signal, and determines whether the recognition signal is interference based on the judgment result; The initial energy is The output energy is The directional deviation is θ error =|θ mode -θ in |θ in The target direction for directional sound pickup; The decision system is set to a strong suppression mode and / or a weak suppression mode; The strong suppression mode is implemented as the direction of arrival deviation θ error Greater than the angle deviation threshold θ th Or the recognition signal energy ratio R is less than the threshold R th , then the identification signal is a misidentification caused by interference; The weak suppression mode is implemented as the direction of arrival deviation θ error Greater than the threshold θ th And the recognition signal energy ratio R is greater than the threshold R th , then the identification signal is a misidentification caused by interference.

2. The method for suppressing misrecognition according to claim 1, characterized in that: Also includes, Step 9: If the identification signal is determined to be interference, the system does not output any identification results.

3. A storage medium, characterized in that: The storage medium includes a stored program, wherein when the program is executed, the device where the storage medium is located is controlled to execute the misrecognition suppression method according to any one of claims 1 to 2.

Images