Method, device and storage medium for vehicle-mounted speech noise reduction

Abstract

The present invention relates to the field of intelligent voice interaction technology, and discloses a method, device and storage medium for vehicle-mounted voice noise reduction; the method comprises: locating the sound source position when the user wakes up the vehicle-mounted voice; if the sound source position meets the first preset condition, recording the wake-up energy of the user waking up the vehicle-mounted voice; performing energy detection on the user's voice interaction process; when it is detected that the audio energy and the wake-up energy of the current audio meet the second preset condition, judging whether the position of the current sound source meets the first preset condition; if the first preset condition is met, marking and storing the current audio as noise. The implementation of the present invention can dynamically capture noise, thereby improving the noise reduction effect and ensuring the effect of voice interaction.

Description

Technical Field

[0001] The present invention relates to the field of intelligent voice interaction technology, and in particular to a method, device and storage medium for vehicle-mounted voice noise reduction. Background Art

[0002] With the popularization and development of in-car voice, the use of in-car voice is a new and rapidly spreading model. At present, the main method of in-car voice noise reduction is: before the actual mass production of the vehicle, artificial noise is created and then some noise scene tests are carried out on some sections of the road to obtain the audio of the noise, and then the audio data is processed to obtain the characteristics of the noise, which are annotated into the algorithm to achieve the noise reduction process. However, the noise reduction effect of this method is not good enough. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide a method, device and storage medium for vehicle-mounted voice noise reduction, which can dynamically capture noise, thereby improving the noise reduction effect and ensuring the effect of voice interaction.

[0004] In order to solve the above technical problems, the first aspect of the present invention discloses a method for vehicle-mounted speech noise reduction, the method comprising:

[0005] Locate the sound source when the user wakes up the car voice;

[0006] If the sound source position meets the first preset condition, recording the wake-up energy of the user's wake-up vehicle voice;

[0007] Performing energy detection on the user's voice interaction process;

[0008] When it is detected that the audio energy of the current audio and the wake-up energy meet the second preset condition, determining whether the position of the current sound source meets the first preset condition;

[0009] If the first preset condition is met, the current audio is marked and stored as noise.

[0010] Optionally, the sound source position satisfies a first preset condition, including:

[0011] The sound source position is located within the set sound receiving area of ​​the microphone closest to it;

[0012] and / or,

[0013] The sound source position is located within a set area above any seat.

[0014] Optionally, the method further includes:

[0015] When it is detected that the audio energy of the current audio is less than the first threshold, the detection is terminated.

[0016] Optionally, after marking and storing the current audio as noise, the method further includes:

[0017] Process the marked audio data to obtain the characteristic value of the noise;

[0018] The noise reduction model is updated according to the characteristic value of the noise.

[0019] Optionally, the noise reduction model is obtained in the following manner:

[0020] Get artificially created audio of man-made noise;

[0021] Processing the data of the man-made noise audio to obtain a characteristic value of the man-made noise;

[0022] The characteristic values ​​of the man-made noise are used as an original training set for training to obtain the noise reduction model.

[0023] Optionally, updating the noise reduction model according to the characteristic value of the noise includes:

[0024] Using the characteristic values ​​of the noise as a supplementary training set;

[0025] The denoising model is further trained using the supplementary training set and the original training set to obtain an updated denoising model.

[0026] Optionally, after performing energy detection on the user's voice interaction process, the method further includes:

[0027] If the audio energy of the audio and the wake-up energy are not detected to meet the second preset condition within the preset time, the second threshold in the second preset condition is adjusted.

[0028] A second aspect of the present invention discloses a device for vehicle-mounted speech noise reduction, the device comprising:

[0029] The positioning module is used to locate the sound source when the user wakes up the in-vehicle voice;

[0030] A recording module, wherein if the sound source position satisfies a first preset condition, the recording module is used to record the wake-up energy of the user's wake-up vehicle voice;

[0031] A detection module, used for performing energy detection on the user's voice interaction process;

[0032] A judgment module, when it is detected that the audio energy of the current audio and the wake-up energy meet the second preset condition, used to judge whether the position of the current sound source meets the first preset condition;

[0033] A marking module, wherein if the first preset condition is met, the marking module is used to mark and store the current audio as noise.

[0034] A third aspect of the present invention discloses a device for vehicle-mounted speech noise reduction, the device comprising:

[0035] A memory storing executable program code;

[0036] a processor coupled to the memory;

[0037] The processor calls the executable program code stored in the memory to execute the method for in-vehicle speech noise reduction disclosed in the first aspect of the present invention.

[0038] The fourth aspect of the present invention discloses a computer-readable storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute the method for in-vehicle speech noise reduction disclosed in the first aspect of the present invention.

[0039] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0040] In the embodiment of the present invention, by locating the sound source position when the user wakes up the vehicle voice, if the sound source position meets the first preset condition, the wake-up energy of the user's wake-up voice is recorded; the user's voice interaction process is energy detected; when it is detected that the audio energy and the wake-up energy of the current audio meet the second preset condition, it is determined whether the position of the current sound source meets the first preset condition; if the first preset condition is met, the current audio is marked and stored as noise. It can be seen that the implementation of the present invention can dynamically capture noise, thereby improving the noise reduction effect and ensuring the effect of voice interaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0042] Figure 1 It is a flow chart of a method for vehicle-mounted speech noise reduction disclosed in an embodiment of the present invention;

[0043] Figure 2 It is a structural schematic diagram of a device for vehicle-mounted speech noise reduction disclosed in an embodiment of the present invention;

[0044] Figure 3 It is a structural schematic diagram of another device for vehicle-mounted speech noise reduction disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0045] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] The terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish different objects rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, device, product or end including a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units inherent to these processes, methods, products or ends.

[0047] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0048] Embodiment 1

[0049] Figure 1 It is a flow chart of a method for vehicle-mounted speech noise reduction disclosed in an embodiment of the present invention.

[0050] like Figure 1 As shown, the method for vehicle-mounted speech noise reduction includes:

[0051] Step 101: Locate the sound source position when the user wakes up the vehicle voice.

[0052] Step 102: If the sound source position meets the first preset condition, the wake-up energy of the user's wake-up vehicle voice is recorded.

[0053] Step 103: Perform energy detection on the user's voice interaction process.

[0054] Step 104: When it is detected that the audio energy and the wake-up energy of the current audio meet the second preset condition, it is determined whether the position of the current sound source meets the first preset condition.

[0055] In the embodiment of the present disclosure, the second preset condition refers to the condition that the audio energy and the wake-up energy need to meet in terms of numerical values. If the audio energy and the wake-up energy of the audio meet the second preset condition, it indicates that noise occurs.

[0056] Optionally, the second preset condition is set as: the difference between the audio energy and the wake-up energy is greater than a second threshold.

[0057] Optionally, the second preset condition is set as: a ratio of audio energy to wake-up energy is greater than a second threshold.

[0058] Optionally, the second threshold is 2.

[0059] Step 105: If the first preset condition is met, the current audio is marked and stored as noise.

[0060] In the disclosed embodiment, the in-vehicle voice is installed on the car. No matter it is 2mics (microphones) or 4mics, the angle of each mic equivalent to the driving position can be known. For example, the angle of the main mic from the main driver can be calculated, and the algorithm can be set to make the wake-up effect better.

[0061] Optionally, the sound source position satisfies the first preset condition, including: the sound source position is located within the set sound receiving area of ​​the microphone closest to it. In this way, it can be determined that the sound source is normally emitted by the user and is not noise.

[0062] Optionally, the sound source position satisfies the first preset condition, including: the sound source position is located within a set area above any seat. In this way, it can be determined that the sound source is emitted by a person in the car, not noise.

[0063] Optionally, the method further includes: when it is detected that the audio energy of the current audio is less than the first threshold, the detection is terminated. In this way, when the audio energy is too low, the noise can be ignored, and therefore no further detection is required.

[0064] Optionally, after the current audio is marked and stored as noise, the method further includes: processing the marked audio data to obtain a characteristic value of the noise; and updating the noise reduction model according to the characteristic value of the noise. In this way, the noise reduction effect can be improved.

[0065] Optionally, the noise reduction model is obtained by: obtaining artificially produced artificial noise audio; processing the data of the artificial noise audio to obtain the characteristic value of the artificial noise; using the characteristic value of the artificial noise as the original training set for training to obtain the noise reduction model. In this way, the original training method of the noise reduction model is given.

[0066] Optionally, the noise reduction model is updated according to the characteristic value of the noise, including: using the characteristic value of the noise as a supplementary training set; and continuing to train the noise reduction model using the supplementary training set and the original training set to obtain an updated noise reduction model. In this way, the noise reduction effect is improved and the effect of voice interaction is guaranteed.

[0067] Optionally, after performing energy detection on the user's voice interaction process, the method further includes: if the audio energy and the wake-up energy of the audio are not detected to meet the second preset condition within the preset time, adjusting the second threshold in the second preset condition. In this way, if the audio energy and the wake-up energy of the audio are not detected to meet the second preset condition within the preset time, it means that the setting of the second threshold in the second preset condition is not appropriate, usually too large, so the second threshold needs to be adjusted before detection.

[0068] The method for in-vehicle voice noise reduction provided by the embodiment of the present disclosure is adopted. By locating the sound source position when the user wakes up the in-vehicle voice, if the sound source position meets the first preset condition, the wake-up energy of the user's wake-up voice is recorded; the user's voice interaction process is energy detected; when it is detected that the audio energy and the wake-up energy of the current audio meet the second preset condition, it is determined whether the position of the current sound source meets the first preset condition; if the first preset condition is met, the current audio is marked and stored as noise. It can be seen that the implementation of the present invention can dynamically capture noise, thereby improving the noise reduction effect and ensuring the effect of voice interaction.

[0069] Embodiment 2

[0070] See also Figure 2 , Figure 2 Schematic diagram of a device for vehicle-mounted speech noise reduction disclosed in an embodiment of the present invention. Figure 2 As shown, the device for vehicle-mounted voice noise reduction includes:

[0071] Positioning module 201, used to locate the sound source position when the user wakes up the vehicle voice;

[0072] The recording module 202 is used to record the wake-up energy of the user's wake-up voice when the sound source position meets the first preset condition;

[0073] The detection module 203 is used to perform energy detection on the user's voice interaction process;

[0074] The judging module 204 is used to judge whether the position of the current sound source satisfies the first preset condition when it is detected that the audio energy and the wake-up energy of the current audio meet the second preset condition;

[0075] The marking module 205 is used to mark and store the current audio as noise if the first preset condition is met.

[0076] The device for in-vehicle voice noise reduction provided by the embodiment of the present disclosure can dynamically capture noise through the cooperation between the positioning module 201, the recording module 202, the detection module 203, the judgment module 204, and the marking module 205, thereby improving the noise reduction effect and ensuring the effect of voice interaction.

[0077] Embodiment 3

[0078] See also Figure 3 , Figure 3 Schematic diagram of a device for vehicle-mounted speech noise reduction disclosed in an embodiment of the present invention. Figure 3 As shown, the device for vehicle-mounted speech noise reduction may include:

[0079] A memory 301 storing executable program codes;

[0080] a processor 302 coupled to the memory 301;

[0081] The processor 302 calls the executable program code stored in the memory 301 to execute the steps of the method for vehicle-mounted speech noise reduction described in the foregoing embodiment of the present invention.

[0082] Embodiment 4

[0083] An embodiment of the present invention discloses a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to enable a computer to execute the steps in the method for in-vehicle speech noise reduction described in the above embodiment.

[0084] The device embodiments described above are only illustrative, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, i.e., they may be located in one place, or they may be distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative work.

[0085] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0086] Through the specific description of the above embodiments, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the above technical solution can be essentially or partly contributed to the prior art in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable rewritable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, a magnetic disk storage, a magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0087] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0088] Finally, it should be noted that the method, device and storage medium for vehicle-mounted voice noise reduction disclosed in the embodiments of the present invention are only preferred embodiments of the present invention, which are only used to illustrate the technical solutions of the present invention, rather than to limit them. Although the present invention has been described in detail with reference to the aforementioned embodiments, it should be understood by those skilled in the art that the technical solutions described in the aforementioned embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for vehicle-mounted speech noise reduction, characterized in that: The method comprises: Locate the sound source when the user wakes up the car voice; If the sound source position meets the first preset condition, recording the wake-up energy of the user's wake-up vehicle voice; Performing energy detection on the user's voice interaction process; When it is detected that the audio energy of the current audio and the wake-up energy meet the second preset condition, determining whether the position of the current sound source meets the first preset condition; If the first preset condition is met, the current audio is marked and stored as noise; After the current audio is marked and stored as noise, the method further includes: Process the marked audio data to obtain the characteristic value of the noise; Updating a noise reduction model according to the characteristic value of the noise; The updating of the noise reduction model according to the characteristic value of the noise includes: Using the characteristic values ​​of the noise as a supplementary training set; The denoising model is further trained using the supplementary training set and the original training set to obtain an updated denoising model.

2. The method according to claim 1, characterized in that The sound source position satisfies a first preset condition, including: The sound source position is located within the set sound receiving area of ​​the microphone closest to it; and / or, The sound source position is located within a set area above any seat.

3. The method according to claim 1, characterized in that The method further comprises: When it is detected that the audio energy of the current audio is less than the first threshold, the detection is terminated.

4. The method according to claim 1, characterized in that: The denoising model is obtained in the following way: Get artificially created audio of man-made noise; Processing the data of the man-made noise audio to obtain a characteristic value of the man-made noise; The characteristic values ​​of the man-made noise are used as an original training set for training to obtain the noise reduction model.

5. The method according to claim 1, characterized in that After performing energy detection on the user's voice interaction process, the method further includes: If the audio energy of the audio and the wake-up energy are not detected to meet the second preset condition within the preset time, the second threshold in the second preset condition is adjusted.

6. A device for vehicle-mounted speech noise reduction, characterized in that: The device comprises: The positioning module is used to locate the sound source when the user wakes up the in-vehicle voice; A recording module, wherein if the sound source position satisfies a first preset condition, the recording module is used to record the wake-up energy of the user's wake-up vehicle voice; A detection module, used for performing energy detection on the user's voice interaction process; A judgment module, when it is detected that the audio energy of the current audio and the wake-up energy meet the second preset condition, used to judge whether the position of the current sound source meets the first preset condition; A marking module, wherein if the first preset condition is met, the marking module is used to mark and store the current audio as noise; After marking and storing the current audio as noise, processing the data of the marked audio to obtain a characteristic value of the noise; Updating a noise reduction model according to the characteristic value of the noise; The updating of the noise reduction model according to the characteristic value of the noise includes: Using the characteristic values ​​of the noise as a supplementary training set; The denoising model is further trained using the supplementary training set and the original training set to obtain an updated denoising model.

7. A device for vehicle-mounted speech noise reduction, characterized in that: The device comprises: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the method for in-vehicle speech noise reduction as described in any one of claims 1-5.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, which, when called, are used to execute the method for in-vehicle speech noise reduction as described in any one of claims 1-5.

Images