Pickup module and voice acquisition device

Abstract

The invention discloses a pickup module and a voice acquisition device, and the pickup module comprises a remote speaking microphone sub-array which comprises at least two omnidirectional ECM microphones and is set to collect surrounding voice signals in a remote speaking pickup mode; the talkback microphone sub-array comprises at least two omnidirectional MEMS microphones and is set to acquire surrounding voice signals in a talkback pickup mode; the first signal processing module is set to perform beam forming and far-field noise reduction processing on the voice signals acquired by the far-talking microphone sub-array in a far-talking pickup mode, and then transmit the voice signals to a voice acquisition processing module of the voice acquisition device; and the second signal processing module is set to carry out near-field noise reduction and echo cancellation processing on the voice signals acquired by the talkback microphone sub-array in the talkback pickup mode and then transmit the voice signals to a voice acquisition processing module of the voice acquisition device. According to the technical scheme, high signal to noise ratio and low-distortion pickup effects can be ensured in both a far-speaking scene and a near-speaking scene.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of sound pickup, in particular to a sound pickup module and a voice collection device. BACKGROUND

[0002] In a voice collection device such as a translation machine, a smart screen, a robot or a smart conference terminal, its sound pickup module often needs to consider the sound pickup requirements in different distances and scenes. The microphone array (such as a beamforming array) of the traditional sound pickup module is generally designed only for the optimization of the far-talking scene with a sound pickup distance ≥ 3 m, and thus there are serious problems of near-field reverberation and plosive distortion for the sound pickup in the talkback scene with a sound pickup distance < 30 cm. However, when the sound pickup module adopts the single / dual-microphone structure used by the traditional talkback machine or handheld translation machine, although it has good sound pickup effect in the talkback scene, it has the problem of sharp decline of signal-to-noise ratio in the far-talking scene, which cannot support the sound pickup requirements in the far-talking scene. At this time, if the number of microphones is simply increased, although the sound pickup performance in the far-talking scene can be improved to a certain extent, the cost, power consumption and structure size of the device will be significantly increased, and the problems of howling, back-sampling distortion and other effects that affect the sound pickup effect in the talkback scene will occur. Therefore, the sound pickup module of the existing voice collection device cannot well meet the sound pickup requirements in the far-talking scene and the talkback scene at the same time. SUMMARY

[0003] The present application aims to provide a sound pickup module and a voice collection device, and aims to improve the problem that the sound pickup module of the existing voice collection device cannot well meet the sound pickup requirements in the far-talking scene and the talkback scene at the same time.

[0004] To achieve this purpose, the present application provides a sound pickup module applied to a voice collection device, which comprises a mode switching module, a far-talking microphone sub-array, a talkback microphone sub-array, a first signal processing module and a second signal processing module, wherein, The mode switching module is configured to switch the working mode of the sound pickup module, so that the sound pickup module is in a far-talking sound pickup mode or in a talkback sound pickup mode. The far-talking microphone sub-array comprises at least two omnidirectional ECM microphones, and the far-talking microphone sub-array is configured to collect the surrounding voice signal when the sound pickup module is switched to the far-talking sound pickup mode. The talkback microphone sub-array comprises at least two omnidirectional MEMS microphones, and the talkback microphone sub-array is configured to collect the surrounding voice signal when the sound pickup module is switched to the talkback sound pickup mode. The first signal processing module is electrically connected to the far-field microphone subarray and is configured to perform beamforming and far-field noise reduction processing on the voice signal acquired by the far-field microphone subarray when the pickup module switches to the far-field pickup mode, and then transmit it to the voice acquisition and processing module of the voice acquisition device. The second signal processing module is electrically connected to the intercom microphone subarray and is configured to perform near-field noise reduction and echo cancellation processing on the voice signal collected by the intercom microphone subarray when the pickup module switches to the intercom pickup mode, and then transmit it to the voice acquisition and processing module of the voice acquisition device.

[0005] Optionally, in some embodiments of this application, the far-field microphone subarray includes two omnidirectional ECM microphones, both of which are built into the housing of the voice acquisition device, and the first pickup holes of the two omnidirectional ECM microphones are respectively opened on the same side surface of the housing of the voice acquisition device. The distance between the first pickup holes of the two omnidirectional ECM microphones is greater than or equal to 50cm.

[0006] Optionally, in some embodiments of this application, the far-field microphone subarray includes two omnidirectional ECM microphones, both of which are built into the housing of the voice acquisition device, and the first pickup holes of the two omnidirectional ECM microphones are respectively opened on different side surfaces of the housing of the voice acquisition device.

[0007] Optionally, in some embodiments of this application, the outer sides of the first pickup holes of both omnidirectional ECM microphones are covered with waterproof and dustproof mesh; and / or, The inner side of the first pickup hole of each of the two omnidirectional ECM microphones is provided with an acoustic damping material layer, which is configured to balance airflow and suppress standing wave interference in a preset frequency band; and / or, Both of the omnidirectional ECM microphones have a tapered, gradually expanding first pickup hole structure, such that the outer diameter of the opening of the first pickup hole of each omnidirectional ECM microphone is larger than the inner diameter of the opening.

[0008] Optionally, in some embodiments of this application, the intercom microphone subarray includes two omnidirectional MEMS microphones, both of which are built into the housing of the voice acquisition device, and the second pickup holes of the two omnidirectional MEMS microphones are respectively opened on different side surfaces of the housing of the voice acquisition device.

[0009] Optionally, in some embodiments of this application, the outer sides of the second pickup holes of the two omnidirectional MEMS microphones are covered with metal dustproof mesh; and / or, The inner side of the second pickup port of each of the two omnidirectional MEMS microphones integrates a miniature acoustic cavity structure. This miniature acoustic cavity structure is configured to adjust the frequency response characteristics of the corresponding omnidirectional MEMS microphone to improve the clarity of mid-to-high frequency speech pickup; and / or, The second pickup holes of the two omnidirectional MEMS microphones are arranged at an angle of 90 to 180 degrees. Optionally, in some embodiments of this application, the omnidirectional ECM microphone includes a first microphone body, a first acoustic conduit, and a first pickup hole. The first microphone body is built into the housing of the voice acquisition device, and the first pickup hole is opened on one side surface of the housing of the voice acquisition device. The pickup side of the first microphone body communicates with the first pickup hole through the first acoustic conduit; and / or, The omnidirectional MEMS microphone includes a second microphone body, a second acoustic conduit, and a second pickup hole. The second microphone body is built into the housing of the voice acquisition device, and the second pickup hole is opened on one side surface of the housing of the voice acquisition device. The pickup side of the second microphone body is connected to the second pickup hole through the second acoustic conduit.

[0010] Furthermore, to achieve this objective, this application embodiment also provides a voice acquisition device, which includes a device housing, a voice acquisition and processing module, and a pickup module of any of the above. The voice acquisition and processing module and the pickup module are respectively built into the device housing, and the voice acquisition and processing module is electrically connected to the first signal processing module and the second signal processing module respectively through the mode switching module.

[0011] Optionally, in some embodiments of this application, the voice acquisition device is a translation device, and the voice acquisition and processing module is a voice translation engine. The voice translation engine includes a voice recognition module, a neural machine translation module, and a voice synthesis module. The speech recognition module is configured to convert the speech signal transmitted from the first signal processing module or the second signal processing module into corresponding text information and transmit it to the neural machine translation module. The neural machine translation module is configured to automatically translate the text information transmitted from the speech recognition module from the original language into the target language and transmit it to the speech synthesis module. The speech synthesis module is configured to convert the text information transmitted from the speech synthesis module into artificial speech and play it.

[0012] Optionally, in some embodiments of this application, the neural machine translation module is located in a cloud server, and the voice acquisition device further includes a wireless communication module built into the device housing. The voice recognition module and the voice synthesis module are respectively wirelessly connected to the neural machine translation module through the wireless communication module.

[0013] The microphone module and voice acquisition device provided in this application, through the aforementioned structural configuration, enable the microphone module to effectively enhance directional pickup capability and suppress far-field environmental noise when switched to far-field pickup mode. This is achieved by using a far-field microphone subarray composed of at least two omnidirectional ECM microphones, combined with beamforming and far-field noise reduction processing of the first signal processing module. This results in a high signal-to-noise ratio and low-distortion pickup effect in far-field pickup scenarios, effectively meeting the pickup requirements of such scenarios. When switched to intercom pickup mode, the microphone module can significantly improve near-field speech clarity and eliminate call echoes through an intercom microphone subarray composed of at least two omnidirectional MEMS microphones, combined with near-field noise reduction and echo cancellation processing of the second signal processing module. This also results in a high signal-to-noise ratio and low-distortion pickup effect in intercom scenarios, effectively meeting the pickup requirements of such scenarios. Thus, this technical solution enables a single microphone module to switch flexibly according to the usage scenario, ensuring high-quality voice signal acquisition while optimizing performance for both far-field and two-way communication scenarios, thereby improving the applicability and user experience of the voice acquisition device. It is evident that this technical solution effectively addresses the problem that existing voice acquisition devices' microphone modules cannot adequately meet the voice pickup requirements of both far-field and two-way communication scenarios simultaneously. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] The structures, proportions, sizes, etc., shown in the accompanying drawings are only for the purpose of assisting those skilled in the art in understanding and reading the content disclosed in the specification, and are not intended to limit the implementation conditions of this application. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportions, or adjustments to the size should still fall within the scope of the technical content disclosed in this application, provided that they do not affect the effects and purposes that this application can produce.

[0016] Figure 1This is a schematic diagram of the structure of the voice acquisition device according to an embodiment of this application; Figure 2 for Figure 1 The diagram shows the disassembled structure of the voice acquisition device. Figure 3 for Figure 1 The diagram shows the internal structure connection of the voice acquisition device.

[0017] Illustrations: 1. Voice acquisition device; 10. Sound pickup module; 11. Mode switching module; 12. Far-field microphone subarray; 121. Omnidirectional ECM microphone; 13. Intercom microphone subarray; 131. Omnidirectional MEMS microphone; 14. First signal processing module; 15. Second signal processing module; 20. Voice acquisition and processing module; 21. Speech recognition module; 22. Neural machine translation module; 23. Speech synthesis module; 30. Device casing. Detailed Implementation

[0018] To make the inventive objectives, features, and advantages of this application more apparent and understandable, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described below are only some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0019] In the description of this application, it should be understood that the terms "upper," "lower," "top," "bottom," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application. It should be noted that when a component is considered to be "connected" to another component, it can be directly connected to the other component or there may be a component centrally located at the same time.

[0020] The technical solution of this application will be further described below with reference to the accompanying drawings and specific embodiments.

[0021] Please see Figures 1 to 3As shown, in one embodiment, this application provides a microphone module 10, which includes a mode switching module 11, a far-field microphone subarray 12, an intercom microphone subarray, a first signal processing module 14, and a second signal processing module 15. The mode switching module 11 is configured to switch the operating mode of the microphone module 10, so that the microphone module 10 is in a far-field microphone mode or an intercom microphone mode. The far-field microphone subarray 12 includes at least two omnidirectional ECM microphones 121, and is configured to collect ambient voice signals when the microphone module 10 is switched to the far-field microphone mode. The intercom microphone subarray 13 includes at least two omnidirectional MEMS microphones 131, and is configured to collect ambient voice signals when the microphone module 10 is switched to the intercom microphone mode. The first signal processing module 14 is electrically connected to the far-field microphone subarray 12 and is configured to perform beamforming and far-field noise reduction processing on the voice signal acquired by the far-field microphone subarray 12 when the pickup module 10 switches to the far-field pickup mode, before transmitting it to the voice acquisition and processing module 20 of the voice acquisition device 1. The second signal processing module 15 is electrically connected to the intercom microphone subarray 13 and is configured to perform near-field noise reduction and echo cancellation processing on the voice signal acquired by the intercom microphone subarray 13 when the pickup module 10 switches to the intercom pickup mode, before transmitting it to the voice acquisition and processing module 20 of the voice acquisition device 1.

[0022] It should be noted that the microphone module 10 in this embodiment is mainly used in voice acquisition devices 1 such as translators, smart screens, robots, or smart conference terminals to realize the microphone function of these voice acquisition devices 1. Generally speaking, the intercom microphone mode mentioned above is mainly used in intercom scenarios with a microphone distance of <30 cm, but can also be extended to scenarios with a microphone distance of <1.5 m. The far-field microphone mode mentioned above is mainly used in far-field scenarios with a microphone distance of ≥3 m, but can also be extended to scenarios with a microphone distance of ≥1.5 m.

[0023] The aforementioned omnidirectional ECM microphone 121 typically possesses high acoustic sensitivity and a wide dynamic range, effectively capturing low-volume speech signals from a distance. This ensures that even in far-field scenarios (such as conference rooms or living rooms), the speaker's speech details are clearly picked up, avoiding signal attenuation and resulting signal-to-noise ratio degradation. Furthermore, the omnidirectional ECM microphone 121 can uniformly receive sound waves from all directions. Therefore, a far-field microphone subarray 12 formed by at least two omnidirectional ECM microphones 121 can achieve wide-range spatial sound field coverage without pre-setting the sound source direction. This characteristic allows for flexible adaptation to changes in the speaker's position in far-field pickup mode, avoiding blind spots caused by directional limitations. Additionally, the aforementioned omnidirectional ECM microphone 121 typically exhibits good noise suppression characteristics in the low-frequency range. Its signal characteristics facilitate optimization and noise reduction by the first signal processing module 14 for typical far-field diffuse noise and steady-state noise, further improving the pickup effect of far-field speech. Therefore, by forming a far-field microphone subarray 12 with at least two omnidirectional ECM microphones 121, and cooperating with the first signal processing module 14 to perform beamforming processing on the voice signal collected by the far-field microphone subarray 12, a directional beam can be formed at the software level. This enhances the voice signal in the target direction while suppressing environmental noise and reverberation interference in other directions. This combination of "hardware omnidirectional + software directional" takes into account both the pickup flexibility and the far-field anti-interference capability, so that it can well meet the pickup needs in far-field scenarios.

[0024] The aforementioned omnidirectional MEMS microphone 131 typically exhibits a wide and flat frequency response within the typical distance (a few centimeters to tens of centimeters) of near-field conversations, enabling it to capture rich details of human voice with high fidelity, especially mid-to-high frequency speech. Furthermore, the high signal-to-noise ratio of the omnidirectional MEMS microphone 131 ensures effective speech pickup even at low volumes. In addition, the high manufacturing precision of the aforementioned omnidirectional MEMS microphone 131 results in extremely consistent frequency response and phase characteristics among individual microphones, providing a reliable hardware foundation for multi-microphone collaborative processing (such as beamforming and echo cancellation) without the need for complex individual calibrations. Therefore, by forming a walkie-talkie subarray 13 with at least two omnidirectional MEMS microphones 131, and in conjunction with the second signal processing module 15, multi-channel echo cancellation can be achieved. The algorithm of the second signal processing module 15 can use the spatial correlation of multiple signals to more accurately distinguish between "echoes from far-field speakers" and "speech from near-field mouths", which greatly improves the echo cancellation effect. In particular, it can better protect near-field speech from being accidentally deleted during walkie-talkies, so that it can well meet the sound pickup needs in walkie-talkie scenarios.

[0025] Thus, the pickup module 10 of this embodiment, through the above-described structural configuration, when switched to far-field pickup mode, can effectively enhance directional pickup capability and suppress far-field environmental noise through a far-field microphone subarray 12 composed of at least two omnidirectional ECM microphones 121, combined with beamforming and far-field noise reduction processing of the first signal processing module 14. This results in a high signal-to-noise ratio and low-distortion pickup effect in far-field scenarios, effectively meeting the pickup requirements of far-field scenarios. When switched to intercom pickup mode, the pickup module 10 can significantly improve near-field speech clarity and eliminate call echo through an intercom microphone subarray 13 composed of at least two omnidirectional MEMS microphones 131, combined with near-field noise reduction and echo cancellation processing of the second signal processing module 15. This results in a high signal-to-noise ratio and low-distortion pickup effect in intercom scenarios, effectively meeting the pickup requirements of intercom scenarios. Thus, this technical solution enables a single microphone module 10 to switch flexibly according to the usage scenario, ensuring high-quality acquisition of voice signals while taking into account performance optimization under different application scenarios of far-field communication and two-way communication, thereby improving the applicability and user experience of the voice acquisition device 1.

[0026] In some examples, such as Figure 2 and Figure 3 As shown, the far-field microphone subarray 12 includes two omnidirectional ECM microphones 121. Both omnidirectional ECM microphones 121 are built into the housing 30 of the voice acquisition device 1, and the first pickup holes of the two omnidirectional ECM microphones 121 are respectively opened on the same side surface of the housing 30 of the voice acquisition device 1. The distance between the first pickup holes of the two omnidirectional ECM microphones 121 is greater than or equal to 50cm. In this way, through the above structural arrangement, an effective directional pickup surface can be formed on one side of the housing 30 of the voice acquisition device 1. Combined with the wideband response characteristics of the omnidirectional ECM microphones 121, it can maintain high sensitivity pickup of speech in the target direction at a long distance, while effectively expanding the pickup angle to adapt to flexible changes in the speaker's position. In addition, the larger microphone spacing can also effectively increase the physical aperture of the array, enabling the beamforming algorithm to form a sharper and more directional beam, thereby more accurately distinguishing the target sound source from interference in complex acoustic environments and enhancing the ability to suppress room reverberation. Furthermore, in this example, a far-field microphone subarray 12 can be formed using only two omnidirectional ECM microphones 121 to meet the sound pickup requirements in far-field scenarios, which can significantly reduce the number of microphones and reduce the cost, power consumption and structural size of the device.

[0027] It should be noted that the first pickup hole in this example can specifically be... Figure 2The mesh structure shown effectively balances air pressure and suppresses airflow impact, thereby significantly reducing wind noise and noise interference from breathing and microphone popping. This structure ensures efficient sound wave transmission while also providing physical protection, enhancing the pickup stability and reliability of the omnidirectional ECM microphone 121 in complex environments, particularly in far-field communication scenarios. In this example, the two first pickup holes can be simultaneously located on the top surface of the device housing 30.

[0028] In some examples, such as Figure 2 and Figure 3 As shown, the far-field microphone subarray 12 includes two omnidirectional ECM microphones 121, both of which are built into the housing 30 of the voice acquisition device 1. The first pickup holes of the two omnidirectional ECM microphones 121 are respectively located on different side surfaces of the housing 30. Thus, by placing the first pickup holes of the two omnidirectional ECM microphones 121 on different side surfaces of the housing 30, the effective pickup space coverage of the far-field microphone subarray 12 can be effectively expanded, enabling signal acquisition from multi-directional or surround sound fields. This layout enhances the sound source orientation adaptability of the pickup module 10 in far-field mode, facilitating more comprehensive capture of target speech in complex acoustic environments.

[0029] In some examples, such as Figure 2 As shown, the outer sides of the first pickup holes of the two omnidirectional ECM microphones 121 are covered with waterproof and dustproof mesh (not shown). This structural design effectively prevents external dust, moisture, and small impurities from entering the interior of the omnidirectional ECM microphone 121, thus protecting the delicate acoustic components from contamination and corrosion and ensuring their long-term stable pickup performance.

[0030] In some examples, such as Figure 2 As shown, the inner side of the first pickup hole of each of the two omnidirectional ECM microphones 121 is provided with an acoustic damping material layer (not shown). The acoustic damping material layer is designed to balance airflow and suppress standing wave interference in a preset frequency band. Thus, through the above structural arrangement, the acoustic damping material layer can effectively balance airflow pressure, buffer wind noise, and specifically suppress the formation of standing waves in the preset frequency band. This design optimizes the working environment of the diaphragm of the omnidirectional ECM microphone 121, helping to improve the purity and clarity of the acquired voice signal, reduce sound quality degradation caused by airflow impact or cavity resonance, and thereby enhance the overall acoustic performance of far-field sound pickup.

[0031] It should be noted that the acoustic damping material layer in this example can specifically be a structural layer made of porous or fibrous material that adheres to the inside of the pickup hole, used to absorb specific sound wave energy and regulate airflow. Its structure typically includes a substrate layer and a surface filter. The substrate layer has a specific thickness, density, and porosity to specifically attenuate the target frequency band (such as low-frequency standing waves) and achieve a smooth airflow transition.

[0032] In some examples, such as Figure 2 As shown, the first pickup holes of both omnidirectional ECM microphones 121 are tapered and gradually expanding (not shown), such that the outer diameter of the opening of the first pickup hole of each omnidirectional ECM microphone 121 is larger than the inner diameter. This structural design effectively expands the effective incident angle of sound waves, enhancing the omnidirectional ECM microphone 121's ability to capture sound signals. Furthermore, this structure helps improve high-frequency response characteristics, enhancing the clarity and naturalness of the speech signal, while also reducing interference from broadband environmental noise such as wind noise to a certain extent, optimizing the pickup effect in far-field speaking scenarios.

[0033] In some examples, such as Figure 1 and Figure 3 As shown, the intercom microphone subarray 13 includes two omnidirectional MEMS microphones 131, both of which are built into the housing 30 of the voice acquisition device 1. The second pickup holes of the two omnidirectional MEMS microphones 131 are respectively located on different side surfaces of the housing 30. Thus, by placing the second pickup holes of the two omnidirectional MEMS microphones 131 on different side surfaces of the housing 30, the effective pickup range of the intercom microphone subarray 13 can be significantly expanded, ensuring stable and clear near-field voice acquisition under different orientations and postures. This layout enhances the flexibility and inclusiveness of intercom voice pickup, while the spatial isolation of different sides helps to initially separate the target voice from environmental noise at the physical level, providing a better signal foundation for subsequent near-field noise reduction and echo cancellation processing. Furthermore, in this example, only two omnidirectional MEMS microphones 131 are needed to form a walkie-talkie microphone subarray 13 to meet the sound pickup requirements in walkie-talkie scenarios, which can significantly reduce the number of microphones and reduce the cost, power consumption and structural size of the device.

[0034] It should be noted that, in this example, one of the two second microphone holes can be specifically set on the back surface of the device housing 30 (and further, adjacent to the camera module of the voice acquisition device 1), and the other can be specifically set on the bottom surface of the device housing 30, so that the two can be set on different side surfaces of the device housing 30.

[0035] In some examples, such as Figure 1As shown, the outer sides of the second pickup holes of the two omnidirectional MEMS microphones 131 are covered with metal dustproof mesh (not shown). This structural design effectively prevents dust and small foreign objects from entering the interior of the omnidirectional MEMS microphones 131, protecting the delicate internal MEMS sensors from contamination and ensuring their long-term stable sound pickup performance. Furthermore, while providing physical protection, the metal mesh structure also has minimal impact on sound signal attenuation, thus maintaining high-frequency response characteristics and meeting the requirements for voice clarity and high fidelity in intercom scenarios.

[0036] In some examples, such as Figure 1 As shown, the inner side of the second pickup hole of each of the two omnidirectional MEMS microphones 131 integrates a miniature acoustic cavity structure (not shown). This miniature acoustic cavity structure is configured to adjust the frequency response characteristics of the corresponding omnidirectional MEMS microphone 131 to improve the clarity of mid-to-high frequency speech pickup. Thus, through this structural configuration, the frequency response curve of each omnidirectional MEMS microphone 131 can be precisely adjusted via the miniature acoustic cavity structure, particularly enhancing the sensitivity to mid-to-high frequency speech signals (typically the 2000Hz-8000Hz key frequency band). This design significantly improves the clarity, intelligibility, and naturalness of human voices in intercom scenarios, helping the voice acquisition device 1 to more accurately capture and reproduce the speech details of near-field speakers, providing a higher-quality signal source for subsequent noise reduction and echo cancellation processing.

[0037] In some examples, such as Figure 1 As shown, the axial direction of the second pickup holes of the two omnidirectional MEMS microphones 131 (i.e., Figure 1 The microphones (in the directions indicated by arrows X and Y) are arranged at an angle of 90 to 180 degrees. This structural arrangement effectively expands the coverage of the intercom microphone subarray 13, enabling wider spatial sound field sampling. This design enhances the directional tolerance and spatial robustness of sound pickup in intercom scenarios, ensuring stable and clear capture of sound sources at close range, regardless of their angle relative to the device, thereby improving user experience and the reliability of voice interaction.

[0038] In some examples, such as Figure 2As shown, the omnidirectional ECM microphone 121 includes a first microphone body, a first acoustic conduit, and a first pickup hole. The first microphone body is built into the housing 30 of the voice acquisition device 1. The first pickup hole is located on one side surface of the housing 30 of the voice acquisition device 1. The pickup side of the first microphone body is connected to the first pickup hole through the first acoustic conduit. Thus, through the above structural arrangement, the first microphone body is connected to the first pickup hole on the surface of the housing 30 via the first acoustic conduit, achieving a physically separate layout between the first microphone body and the first pickup hole. This design allows the first microphone body to be placed more flexibly inside the housing 30 to optimize the overall structure or avoid interference sources. Simultaneously, the longer first acoustic conduit can create a certain physical filtering effect, helping to attenuate some noise of specific frequencies during transmission. Furthermore, this structure increases the design freedom of the housing 30, facilitating better sealing and protection.

[0039] In some examples, such as Figure 1 As shown, the omnidirectional MEMS microphone 131 includes a second microphone body, a second acoustic conduit, and a second pickup hole. The second microphone body is built into the housing 30 of the voice acquisition device 1, and the second pickup hole is located on one side surface of the housing 30. The pickup side of the second microphone body is connected to the second pickup hole via the second acoustic conduit. Thus, through this structural arrangement, the second microphone body is connected to the second pickup hole on the surface of the housing 30 via the second acoustic conduit, achieving a physically separate layout between the second microphone body and the second pickup hole. This design allows the second microphone body to be placed more flexibly inside the housing 30 to optimize the overall structure or avoid interference sources. Furthermore, this structure, through the acoustic filtering characteristics of the second acoustic conduit, can effectively suppress environmental noise of specific frequencies (such as internal fan or circuit noise) during sound wave transmission, helping to improve the signal-to-noise ratio and purity of the voice signal in intercom scenarios.

[0040] In one embodiment, such as Figures 1 to 3 As shown, this application embodiment also provides a voice acquisition device 1, which includes a voice acquisition and processing module 20 and a pickup module 10 as described above. The voice acquisition and processing module 20 is electrically connected to the first signal processing module 14 and the second signal processing module 15 via a mode switching module 11. Thus, since the voice acquisition device 1 of this application embodiment uses the pickup module 10 of the above embodiment, it can achieve the same technical effects as the pickup module 10 described above, and will not be repeated here.

[0041] In some examples, such as Figures 1 to 3As shown, the voice acquisition device 1 can specifically be a translation device, and the voice acquisition and processing module 20 can specifically be a voice translation engine. The voice translation engine includes a speech recognition (Automatic Speech Recognition, or "ASR") module, a neural machine translation (NMT) module, and a speech synthesis (Text-To-Speech, or "TTS") module. The speech recognition module 21 is configured to convert the voice signal transmitted from the first signal processing module 14 or the second signal processing module 15 into corresponding text information and transmit it to the neural machine translation module 22. The neural machine translation module 22 is configured to automatically translate the text information transmitted from the speech recognition module 21 from the original language into the target language and transmit it to the speech synthesis module 23. The speech synthesis module 23 is configured to convert the text information transmitted from the speech synthesis module 23 into artificial speech and play it. Thus, through the aforementioned structural design, the sound pickup module 10 is integrated into the translation device and equipped with a complete speech translation engine. This allows the device to adaptively switch between far-field speaking and two-way audio pickup modes, optimizing the speech acquisition quality in different scenarios, thereby providing high-fidelity audio input for subsequent recognition, translation, and synthesis. This design enables the translation device to not only accurately process speech in scenarios such as long-distance speeches and meetings, but also clearly capture close-range conversations, achieving high-quality, real-time cross-language translation and speech playback across all scenarios, significantly improving user experience and product usability.

[0042] In some examples, the neural machine translation module 22 is located in a cloud server. The speech acquisition device 1 also includes a wireless communication module. The speech recognition module 21 and the speech synthesis module 23 are wirelessly connected to the neural machine translation module 22 via the wireless communication module. Thus, by deploying the neural machine translation module 22 in a cloud server and connecting it via a wireless communication module, the translation device can utilize the powerful computing resources and continuously updated translation models of the cloud to achieve more complex and accurate real-time machine translation, while significantly reducing the hardware load and power consumption of local devices. This architecture ensures the scalability and updability of translation capabilities, enabling the translation device to provide high-quality translation services anytime, anywhere, and supporting longer battery life.

[0043] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A sound pickup module, used in a voice acquisition device, characterized in that, The microphone pickup module includes a mode switching module, a far-field microphone subarray, a two-way microphone subarray, a first signal processing module, and a second signal processing module. The mode switching module is configured to switch the working mode of the microphone module so that the microphone module is in far-field microphone mode or in intercom microphone mode. The far-field microphone subarray includes at least two omnidirectional ECM microphones, and the far-field microphone subarray is configured to collect ambient voice signals when the pickup module switches to the far-field pickup mode. The intercom microphone subarray includes at least two omnidirectional MEMS microphones, and the intercom microphone subarray is configured to collect ambient voice signals when the pickup module switches to the intercom pickup mode; The first signal processing module is electrically connected to the far-field microphone subarray and is configured to perform beamforming and far-field noise reduction processing on the voice signal acquired by the far-field microphone subarray when the pickup module switches to the far-field pickup mode, and then transmit it to the voice acquisition and processing module of the voice acquisition device. The second signal processing module is electrically connected to the intercom microphone subarray and is configured to perform near-field noise reduction and echo cancellation processing on the voice signal collected by the intercom microphone subarray when the pickup module switches to the intercom pickup mode, and then transmit it to the voice acquisition and processing module of the voice acquisition device.

2. The pickup module according to claim 1, characterized in that, The far-field microphone subarray includes two omnidirectional ECM microphones, both of which are built into the housing of the voice acquisition device, and the first pickup holes of the two omnidirectional ECM microphones are respectively opened on the same side surface of the housing of the voice acquisition device. The distance between the first pickup holes of the two omnidirectional ECM microphones is greater than or equal to 50cm.

3. The pickup module according to claim 1, characterized in that, The far-field microphone subarray includes two omnidirectional ECM microphones, both of which are built into the housing of the voice acquisition device, and the first pickup holes of the two omnidirectional ECM microphones are respectively opened on different side surfaces of the housing of the voice acquisition device.

4. The pickup module according to claim 2 or 3, characterized in that, The outer sides of the first pickup holes of both omnidirectional ECM microphones are covered with waterproof and dustproof mesh; and / or, The inner side of the first pickup hole of each of the two omnidirectional ECM microphones is provided with an acoustic damping material layer, which is configured to balance airflow and suppress standing wave interference in a preset frequency band; and / or, Both of the omnidirectional ECM microphones have a tapered, gradually expanding first pickup hole structure, such that the outer diameter of the opening of the first pickup hole of each omnidirectional ECM microphone is larger than the inner diameter of the opening.

5. The pickup module according to claim 1, characterized in that, The intercom microphone subarray includes two omnidirectional MEMS microphones, both of which are built into the housing of the voice acquisition device, and the second pickup holes of the two omnidirectional MEMS microphones are respectively opened on different side surfaces of the housing of the voice acquisition device.

6. The pickup module according to claim 5, characterized in that, The outer sides of the second pickup holes of the two omnidirectional MEMS microphones are covered with metal dustproof mesh; and / or, The inner side of the second pickup hole of each of the two omnidirectional MEMS microphones is integrated with a micro acoustic cavity structure. The micro acoustic cavity structure is configured to adjust the frequency response characteristics of the corresponding omnidirectional MEMS microphone to improve the pickup clarity of mid-to-high frequency speech. And / or, The second pickup holes of the two omnidirectional MEMS microphones are arranged at an angle of 90 to 180 degrees.

7. The pickup module according to claim 1, characterized in that, The omnidirectional ECM microphone includes a first microphone body, a first acoustic conduit, and a first pickup hole. The first microphone body is built into the housing of the voice acquisition device. The first pickup hole is formed on one side surface of the housing of the voice acquisition device. The pickup side of the first microphone body communicates with the first pickup hole through the first acoustic conduit; and / or, The omnidirectional MEMS microphone includes a second microphone body, a second acoustic conduit, and a second pickup hole. The second microphone body is built into the housing of the voice acquisition device, and the second pickup hole is opened on one side surface of the housing of the voice acquisition device. The pickup side of the second microphone body is connected to the second pickup hole through the second acoustic conduit.

8. A voice acquisition device, characterized in that, The voice acquisition device includes a device housing, a voice acquisition and processing module, and a pickup module as described in any one of claims 1-7. The voice acquisition and processing module and the pickup module are respectively built into the device housing. The voice acquisition and processing module is electrically connected to the first signal processing module and the second signal processing module respectively through the mode switching module.

9. The voice acquisition device according to claim 8, characterized in that, The voice acquisition device is a translation device, and the voice acquisition and processing module is a voice translation engine. The voice translation engine includes a voice recognition module, a neural machine translation module, and a voice synthesis module. The speech recognition module is configured to convert the speech signal transmitted from the first signal processing module or the second signal processing module into corresponding text information and transmit it to the neural machine translation module. The neural machine translation module is configured to automatically translate the text information transmitted from the speech recognition module from the original language into the target language and transmit it to the speech synthesis module. The speech synthesis module is configured to convert the text information transmitted from the speech synthesis module into artificial speech and play it.

10. The voice acquisition device according to claim 9, characterized in that, The neural machine translation module is located in a cloud server, and the voice acquisition device also includes a wireless communication module built into the device housing. The voice recognition module and the voice synthesis module are wirelessly connected to the neural machine translation module through the wireless communication module.

Images