Conference sign-in method and system based on voiceprint recognition

Abstract

The invention discloses a conference sign-in method and system based on voiceprint recognition, and relates to the field of voiceprint recognition. The method comprises the steps that under the condition that conference sign-in is not started, a first sound source collected by a microphone array in a preset mode in a target area is acquired, the preset mode is adjusted to be a target mode based on the first sound source, and the preset mode comprises the arrangement mode and the array scale of the microphone array; and under the condition that conference sign-in begins, obtaining a second sound source collected by the microphone array in the target mode, obtaining a voiceprint recognition result of the second sound source, and performing conference sign-in based on the recognition result. By implementing the technical scheme provided by the invention, the reliability of the voiceprint recognition conference sign-in system in different conference environments can be improved.

Description

Technical Field

[0001] The present application relates to the field of voiceprint recognition, and specifically to a conference sign-in method and system based on voiceprint recognition. Background Art

[0002] As the level of intelligent conference management continues to improve, more and more companies and institutions are beginning to adopt automated conference sign-in systems. In large conference scenarios, due to the large number of participants and the complex and ever-changing conference environment, how to quickly and accurately complete the identity verification and sign-in of participants has become an important issue. Voiceprint recognition technology has gradually become an important technical choice for conference sign-in systems due to its non-contact, high efficiency and difficulty in forging characteristics.

[0003] At present, conference sign-in systems based on voiceprint recognition usually use fixed microphone arrays to collect voice signals from participants. When participants enter the conference room, such systems collect their voice samples through pre-deployed microphones, match the extracted voiceprint features with samples in the database, and thus achieve identity authentication and automatic sign-in.

[0004] However, existing voiceprint recognition conference sign-in systems often use fixed microphone configurations and cannot be dynamically adjusted according to the actual conference environment. When the conference environment changes, such as the number of participants, personnel distribution, or environmental noise, the fixed microphone configuration may cause the sound collection quality to deteriorate, affecting the accuracy of voiceprint recognition, thereby reducing the reliability of the sign-in system. Summary of the invention

[0005] The present application provides a conference sign-in method and system based on voiceprint recognition, which can improve the reliability of the voiceprint recognition conference sign-in system in different conference environments.

[0006] In a first aspect of the present application, the present application provides a conference check-in method based on voiceprint recognition, comprising:

[0007] When the conference check-in has not started, obtaining a first sound source collected by a microphone array in a preset mode in a target area, and adjusting the preset mode to a target mode based on the first sound source, wherein the preset mode includes an arrangement mode and an array size of the microphone array;

[0008] When the conference sign-in starts, a second sound source collected by the microphone array in the target mode is obtained, a voiceprint recognition result of the second sound source is obtained, and the conference sign-in is performed based on the recognition result.

[0009] In a second aspect of the present application, a conference check-in system based on voiceprint recognition is provided, comprising:

[0010] A microphone array adjustment module, used for obtaining a first sound source collected by a microphone array in a preset mode in a target area when the conference check-in has not started, and adjusting the preset mode to a target mode based on the first sound source, wherein the preset mode includes an arrangement mode and an array size of the microphone array;

[0011] The voiceprint recognition sign-in module is used to obtain a second sound source collected by the microphone array in the target mode when the meeting sign-in starts, obtain the voiceprint recognition result of the second sound source, and perform meeting sign-in based on the recognition result.

[0012] In a third aspect of the present application, a computer storage medium is provided, wherein the computer storage medium stores a plurality of instructions, wherein the instructions are suitable for being loaded by a processor and executing the above method steps.

[0013] In a fourth aspect of the present application, an electronic device is provided, comprising: a processor and a memory; wherein the memory stores a computer program, and the computer program is suitable for being loaded by the processor and executing the above-mentioned method steps.

[0014] In a fifth aspect of the present application, a computer program product is provided, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements any of the method steps described above.

[0015] In summary, one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0016] By obtaining the first sound source collected by the microphone array in a preset mode in the target area before the start of the meeting sign-in, and adjusting the preset mode to the target mode based on the first sound source, the arrangement and array size of the microphone array can adapt to the current meeting environment characteristics, and then at the start of the meeting sign-in, the microphone array in the adjusted target mode is used to collect the second sound source and perform voiceprint recognition. Compared with the method of using a fixed microphone configuration scheme in the prior art, the present application improves the sound collection quality by dynamically adjusting the microphone array configuration, so that the system can obtain better quality voiceprint features, thereby improving the reliability of the voiceprint recognition conference sign-in system in different meeting environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a flowchart of a conference sign-in method based on voiceprint recognition provided by an embodiment of the present application;

[0018] Figure 2 It is a structural diagram of a conference sign-in system based on voiceprint recognition provided in an embodiment of the present application;

[0019] Figure 3It is a schematic diagram of the structure of an electronic device disclosed in an embodiment of the present application. DETAILED DESCRIPTION

[0020] In order to enable technicians in this field to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments.

[0021] In the description of the embodiments of the present application, words such as "for example" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "for example" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "for example" or "for example" is intended to present related concepts in a specific way.

[0022] In the description of the embodiments of the present application, the meaning of the term "multiple" refers to two or more. For example, multiple systems refer to two or more systems, and multiple screen terminals refer to two or more screen terminals. In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. The terms "include", "comprise", "have" and their variations all mean "including but not limited to", unless otherwise specifically emphasized.

[0023] Please refer to Figure 1 , Figure 1 This is a flow chart of a conference sign-in method based on voiceprint recognition provided by an embodiment of the present application. The method can be implemented by a computer program, can be implemented by a single-chip microcomputer, or can be run on a conference sign-in system based on voiceprint recognition based on a von Neumann system. The computer program can be integrated into an application or run as an independent tool application. Specifically, the method may include the following steps:

[0024] Step 101: When the conference check-in has not started, obtain a first sound source collected by a microphone array in a preset mode in a target area, and adjust the preset mode to a target mode based on the first sound source. The preset mode includes an arrangement of the microphone array and an array size.

[0025] The target area refers to the specific space in the conference room used for conference sign-in. Specifically, the target area can be understood as an area surrounded by conference tables, chairs, podiums and other facilities, where participants can talk normally and the microphone array can effectively collect sound. For example, in a large conference room, the target area can be the space within 3 meters around the conference table; in a multi-functional conference hall, the target area can be the audience area with seats; in a lecture hall, the target area can include the podium area and the audience seating area.

[0026] In the embodiment of the present application, the delineation of the target area needs to comprehensively consider the following factors: first, the acoustic propagation characteristics within the area should be relatively stable to avoid severe acoustic attenuation or distortion caused by the spatial structure; second, the area should cover all possible locations of participants who need to sign in; finally, the scope of the area should be within the effective pickup range of the microphone array to ensure the collection quality of the voiceprint signal.

[0027] The target area is used to provide a spatial reference for the deployment of the microphone array, and also to provide boundary constraints for subsequent area division and sound source collection. By properly defining the target area, the adaptability of the system to the conference environment can be improved, and the accuracy of voiceprint recognition sign-in can be ensured. In actual applications, if the target area is too large, the collection quality of areas far away from the microphone array may be reduced; if the target area is too small, the sign-in audio of some participants may be missed.

[0028] Furthermore, a microphone array refers to an acoustic sensor system composed of multiple microphone units arranged in a target area in a specific geometric arrangement. Specifically, a microphone array can be understood as a combination of at least two microphone units through a preset spatial relationship, where these microphone units maintain a certain relative position and spacing, and work together to achieve functions such as sound source localization, sound enhancement, and noise suppression.

[0029] It should be noted that multiple independent microphone units can be set up in the target area, and some of the microphones in these microphone units can form one or more microphone arrays according to actual needs. For example, in a large conference room, there may be a total of 20 microphone units, of which 8 microphone units can form a linear microphone array for the podium area, and the other 12 microphone units can form multiple small arrays or remain in an independent working state as needed. This flexible configuration method enables the system to dynamically adjust the composition of the microphone array according to the specific conditions of the conference environment.

[0030] The microphone array in the embodiment of the present application is mainly used for the collection and processing of high-quality voiceprint information. Through the collaborative work of multiple microphone units, the microphone array can achieve spatially selective sound collection and effectively suppress the interference of environmental noise and reverberation. At the same time, by using the time difference and phase difference between the sound signals collected by each microphone unit in the microphone array, the system can accurately locate the sound source and separate the sound, providing high-quality audio input for subsequent voiceprint recognition.

[0031] Correspondingly, the sum of the acoustic signals in the target area collected by the microphone array in the preset mode. The first sound source can be understood as a composite acoustic signal that includes multiple acoustic components such as ambient noise, conversations of participants when entering the venue, operation sounds of equipment such as air conditioners, and room reverberation. For example, when participants enter the conference room and take their seats or wait in the queue area, all acoustic signals collected by the microphone array, including human voices, ambient noise, and all acoustic signals after reflection and reverberation in the conference room, are components of the first sound source.

[0032] In the embodiment of the present application, the first sound source is mainly used for the system to pre-evaluate and adapt the conference environment. By analyzing the acoustic characteristics such as the noise level, reverberation characteristics, and sound source distribution in the first sound source, the system can promptly discover whether the microphone array in the preset mode is suitable for the current conference environment. For example, by analyzing the noise components in the first sound source, the signal-to-noise ratio level of the current environment can be evaluated; by analyzing the reverberation characteristics therein, the acoustic characteristics of the conference room can be understood; by analyzing the spatial distribution of the human voice components, the main activity area of ​​the participants can be determined. This information provides a basis for subsequent adjustments to the arrangement and size of the microphone array.

[0033] The preset mode includes the arrangement and array size of the microphone array, which refers to the initial working configuration parameters of the microphone array before the meeting sign-in begins. Among them, the arrangement refers to the geometric distribution of the microphone units in space, and the array size refers to the number of microphone units involved in the array. Specifically, the preset mode can be understood as a set of default parameters pre-set by the system when it leaves the factory or by the system administrator. For example, the arrangement can be an evenly spaced linear arrangement with a spacing of 10 cm, and the array size can be a standard configuration consisting of 8 omnidirectional microphone units.

[0034] In specific application scenarios, the arrangement of preset modes may include but is not limited to: linear arrangement (microphone units are evenly distributed along a straight line), circular arrangement (microphone units are arranged in a circle), rectangular arrangement (microphone units are arranged in a rectangular array), random arrangement (irregular arrangement according to the layout of the conference room), etc. The array size can be preset according to the size of the conference room and the expected number of participants, and can usually be set to different numbers of microphone units such as 4, 8, 16, etc. The preset mode usually selects more general configuration parameters to adapt to most common meeting scenarios.

[0035] The preset mode setting is used to provide the system with an initial working state and serve as a benchmark configuration for subsequent optimization and adjustment. By analyzing the characteristics of the first sound source collected in the preset mode, the system can evaluate whether the current preset mode is suitable for a specific meeting environment. For example, if the sound source localization accuracy is found to be insufficient under the preset linear arrangement, the system may need to be adjusted to a circular arrangement; if the noise suppression effect is found to be unsatisfactory under the preset 8-unit scale, the number of microphone units may need to be increased. This adaptive optimization mechanism based on the preset mode enables the system to find the optimal working configuration based on the characteristics of the actual meeting environment.

[0036] Compared with the traditional fixed-mode microphone array system, the present invention significantly improves the system's adaptability to different conference environments by using a preset mode as the initial configuration and combining it with an adaptive adjustment mechanism.

[0037] Based on the above embodiment, as an optional embodiment, in step 101: the step of obtaining a first sound source collected by a microphone array in a preset mode in the target area may further include the following steps:

[0038] Step 201: Obtain the distribution locations of participants in the target area.

[0039] Specifically, obtaining the distribution position of participants in the target area means that before the meeting check-in begins, the system determines the spatial distribution form of participants in the target area. The distribution position here emphasizes the organizational form and activity characteristics of participants in space, rather than simple location coordinates or personnel density. For example, during a meeting, participants usually have a relatively fixed seated state, and the distribution position in this case can be described as "static distribution"; while during the meeting check-in stage, participants may need to queue up in a designated area, and the distribution position at this time is "dynamic distribution".

[0040] Furthermore, the purpose of obtaining the distribution location of participants is to identify the application characteristics of the current conference scene and provide a basis for the subsequent selection of sound source collection strategies. The system can collect images or videos through visual sensors deployed in the conference room and analyze the distribution pattern of participants through image processing algorithms; it can also determine the spatial distribution characteristics of sound sources by analyzing acoustic signals from microphone arrays; or determine the overall distribution pattern through signal interaction between smart devices carried by participants and positioning base stations in the conference room.

[0041] Specifically, the system first needs to establish a scene recognition model. For static distribution scenes, the system focuses on whether the seating positions of participants are relatively stable and whether participants in different areas are regularly distributed. For dynamic distribution scenes, the system needs to identify the location of the queue channel, the direction of the team formation, and the movement pattern of the flow of people. The recognition process of this scene feature is continuous, and the system will confirm the changes in the distribution pattern through real-time monitoring.

[0042] Step 202: When the distribution positions of the participants are statically distributed, the target area is divided into at least one sub-area based on the distribution positions of the participants and the distribution positions of the microphones in the target area, and the first sound source collected by the microphone array in a preset mode in each sub-area is obtained.

[0043] Specifically, when it is determined that the participants are statically distributed, the system needs to divide the target area into sub-areas based on the spatial distribution relationship between the participants and the microphones, and collect sound sources on this basis. The static distribution here usually occurs in the formal stage of the meeting, when the participants are already seated in the meeting room and their positions are relatively fixed, such as sitting around the conference table or in the seating area of ​​the lecture hall.

[0044] Furthermore, the purpose of dividing the target area into sub-areas is to achieve more targeted sound source collection. Since the positions of participants are relatively fixed in the case of static distribution, the system can divide the target area into several sub-areas with relatively independent acoustic characteristics according to the distribution characteristics of the participants and the deployment positions of the microphones. For example, in a multi-functional conference hall, the target area can be divided into a podium area, a central conference area, and a surrounding audience area; in a large conference room, the area can be divided into several seating areas according to the layout of the conference table. This area division method based on actual scene characteristics can better adapt to the acoustic characteristics of the conference environment.

[0045] Specifically, the system first needs to determine the sub-area division scheme. A region division algorithm based on the Voronoi diagram can be used, with the microphone position as the center point, to divide the target area into multiple sub-areas based on the spatial distance relationship. The following factors will be considered during the division: First, the number of participants in each sub-area should be relatively balanced to avoid the situation where there are too many or too few people in a sub-area; second, the boundaries of the sub-areas should be coordinated with the physical layout of the conference room (such as the placement of tables and chairs, the location of aisles, etc.); finally, each sub-area should have sufficient microphone coverage to ensure the quality of sound source collection.

[0046] After completing the area division, the system will use the microphone array in the preset mode to collect sound sources in each sub-area. Since the acoustic environment and the distribution characteristics of participants in each sub-area may be different, the system will adjust the working parameters of the microphone array in a targeted manner. For example, for sub-areas with denser participants, it may be necessary to improve the directional performance of the microphone array; and for sub-areas with more dispersed participants, it may be necessary to expand the pickup range of the microphone. The system will continuously monitor the quality of the first sound source collected in each sub-area, including parameters such as signal-to-noise ratio, sound source directionality, and reverberation level, to provide a basis for subsequent mode optimization.

[0047] For example, in a static distribution scenario, assume that the microphone positions in the target area are M1, M2, ..., M N , the positions of the participants are P1, P2, ...P k ,The goal is to divide the target area into several sub-areas so that each microphone is responsible for collecting the sound in its sub-area.

[0048] Furthermore, in order to achieve the best region division effect, the present invention adopts a region division algorithm based on the Voronoi diagram. The algorithm divides each point in space into the region corresponding to the microphone closest to it. Specifically, for each microphone M i , and its corresponding Voronoi cell V i Defined by the following conditions:

[0049]

[0050] Among them, d(P,M j ) represents the point P and the microphone M i The Euclidean distance between:

[0051]

[0052] In the above manner, the system can divide the target area into N non-overlapping sub-areas, and the distance from all points in each sub-area to its corresponding microphone is smaller than the distance to other microphones.

[0053] The above-mentioned area division method based on the Voronoi diagram can ensure that each participant is divided into the nearest microphone collection area, minimizing the attenuation during sound propagation; secondly, since the division of the Voronoi diagram is based on the natural division of spatial distance, the sub-area boundaries formed are often well adapted to the physical layout of the conference room; finally, this division method is simple to calculate and the result is unique, and it can quickly respond to changes in the distribution of participants. In this way, the system can establish a clear microphone responsible area, provide a clear spatial division basis for subsequent sound source collection, and thus improve the work efficiency of the entire conference check-in system.

[0054] Step 203: When the distribution positions of the participants are dynamically distributed, based on the distribution positions of the participants and microphones in the target area, a queuing area in the target area is determined, and a first sound source collected by the microphone array in a preset mode in the queuing area is obtained.

[0055] Specifically, when it is determined that the participants are dynamically distributed, the system needs to determine the queuing area in the target area based on the spatial distribution relationship between the participants and the microphones, and collect sound sources in the area. The dynamic distribution here usually occurs in the initial stage of conference check-in, when the participants have not yet taken their seats, but are queuing in a specific area to check in, such as queuing at the entrance of the conference room, waiting in turn in front of the check-in desk, etc.

[0056] Furthermore, the purpose of determining the queuing area is to accurately capture the sound source information in dynamic scenes. Since participants will continue to move during the queuing process and the number of people is relatively dense, the traditional fixed-area sound source collection method may not be able to effectively cope with such dynamically changing scenes. By identifying and determining the queuing area, the system can focus on the sound source collection within the spatial range of the participants' main activities, thereby improving the collection efficiency. For example, one or more queues may form at the entrance of a conference room. The system needs to accurately identify the location and direction of these queues and adjust the working mode of the microphone array accordingly.

[0057] Specifically, the system first needs to determine the scope and characteristics of the queuing area through real-time monitoring. A method based on spatiotemporal clustering can be used to identify the queuing area by analyzing the movement trajectory and aggregation characteristics of the participants. The identification process will take into account the following factors: First, the system will analyze the movement direction of the participants and identify the flow trajectory with obvious directionality; second, it will pay attention to the changes in the density distribution of the participants and find areas with dense population and linear or strip distribution; finally, combined with the physical layout of the conference room, the final queuing area scope is determined.

[0058] After determining the queuing area, the system will arrange and adjust the microphone array in the area in a targeted manner. Since the queuing area is characterized by dense crowds and frequent movement, the system uses the microphone array in the preset mode to collect sound sources.

[0059] It should be noted that the coverage of the microphone array needs to match the length and width of the queue; secondly, considering the mobility of the participants, the system may need to enable the sound source tracking function to ensure the continuity of sound collection; finally, in response to the higher ambient noise that may occur in the queuing area, the system will appropriately adjust the noise reduction parameters of the microphone array.

[0060] In an optional implementation, when the distribution positions of the participants are dynamically distributed, the system needs to determine the queuing area based on the real-time movement characteristics of the participants. Specifically, assuming that the real-time positions of the participants in the target area are P1(t), P1(t), ..., P k (t), since participants will continue to move during the queuing process, the traditional fixed area division method is difficult to adapt to this dynamically changing scene. Therefore, the present invention adopts a dynamic clustering algorithm based on K-means to determine the queuing area, and adaptively identifies and updates the scope of the queuing area by real-time tracking and analyzing the position changes of participants.

[0061] Furthermore, the system first needs to set the cluster centers C1, C2, ..., C m The value of m can be predetermined based on the size of the meeting and the layout of the venue. In the initialization phase of the algorithm, the system randomly selects m participant locations as the initial clustering centers.

[0062] This random initialization method can prevent the algorithm from falling into the local optimal solution and improve the global optimality of the clustering results. After that, the system enters the iterative optimization stage: for each participant position P updated in real time k (t), find the nearest cluster center by calculating the Euclidean distance, that is, solve:

[0063]

[0064] After completing a round of allocation, the system will recalculate the center position of each cluster, and the update formula is:

[0065]

[0066] Among them, S j Belongs to cluster C j The above steps are repeated for all participant position sets until the cluster centers converge. The queuing area can be defined by the clustering results.

[0067] The above dynamic clustering method can reflect the movement patterns of participants in real time. When the position of the participants changes, the cluster center will be adjusted accordingly to ensure that the division of the queuing area is always consistent with the actual situation. Secondly, by calculating the cluster center, the system can effectively identify areas with dense crowds, which are usually where queues occur. Finally, the iterative optimization characteristics of the algorithm make the clustering results continue to converge to the optimal solution, providing a reliable basis for the accurate division of the queuing area.

[0068] In actual applications, the system will continuously monitor the changing trend of the cluster center. When the position of the cluster center tends to be stable, the algorithm can be considered to have converged, and the clustering results at this time can accurately reflect the spatial distribution of the queuing area. Based on the final clustering results, the system can determine the specific range of each queuing area and adjust the working parameters of the microphone array accordingly to ensure the accuracy of voiceprint collection. For example, for the identified long queue area, the system can deploy a linear microphone array along the queue direction; for more dispersed waiting areas, a ring or matrix microphone layout can be used. This queuing area recognition method based on dynamic clustering can effectively improve the adaptability and reliability of the system in dynamic scenarios.

[0069] Step 301: Obtain the noise level in the first sound source and the density of participants.

[0070] Among them, the noise level in the first sound source refers to the measurement value of the intensity of all background sounds except the effective voice signal in the target area. In the embodiment of the present application, it can be understood as the acoustic interference intensity formed by the superposition of multiple noise sources such as the operating sound of environmental equipment, room reverberation, external interference, etc., which is usually quantified in decibels (dB). For example, the operating noise of the air-conditioning system, the heat dissipation sound of the projection equipment, the traffic noise introduced from the outside, and the reverberation effect of these noises in the conference room, etc., are all components of the noise level of the first sound source.

[0071] The noise level in the first sound source is used in the embodiment of the present application to evaluate the acoustic quality of the conference environment and predict possible interference with voiceprint collection. By measuring and analyzing the noise level, the system can determine whether it is necessary to adjust the signal processing parameters of the microphone array, such as increasing the signal filtering strength in a high-noise environment, or reducing the signal gain in a low-noise environment to avoid distortion. At the same time, the changing trend of the noise level can also reflect the dynamic characteristics of the conference environment, providing an important reference for the real-time optimization of the microphone array.

[0072] The density of participants refers to the number of participants accommodated per unit area in the target area. In the embodiments of the present application, it can be understood as the degree of concentration of participants in the conference space, usually expressed in units of people per square meter. For example, in a large conference room, there may be only a few speakers in the podium area, showing a low density of people; while in the check-in waiting area, due to the concentrated queueing of participants, there may be a higher density of people.

[0073] The density of participants is used in the embodiment of the present application to evaluate the spatial distribution characteristics of the sound source and predict the possible degree of interference from the sound source. By calculating and analyzing the density of people, the system can determine the optimal coverage and spatial resolution requirements of the microphone array. For example, in areas with high population density, the system needs to improve the spatial selectivity of the microphone array to accurately separate the voices of adjacent participants; while in areas with low population density, the pickup range of a single microphone can be appropriately expanded to improve the coverage efficiency of the system.

[0074] For example, in a conference scenario, the acoustic environment in the target area is usually complex, including voice signals of multiple participants and various environmental noises. In order to accurately obtain the voiceprint information of the participants, the system uses a microphone array to collect and process the sound signals. Specifically, when the microphone M i When collecting sound signals in the target area, the acquired signal S i (t) contains the superposition of all sound sources in the area, which can be expressed as:

[0075]

[0076] In the formula, A k is the sound source intensity of the kth participant, ω k is the frequency of the kth sound source, is the phase of the kth sound source, N i (t) is the microphone M i The noise collected.

[0077] In order to improve the quality of sound source collection, the system introduces beamforming technology. By weighted superposition of the signals collected by each microphone in the microphone array, the output signal of beamforming can be expressed as:

[0078]

[0079] In the formula, w i It is a microphone M i The weight, τ i It is a microphone M i The time delay of the acquired signal relative to the reference microphone.

[0080] The system can specifically enhance sound signals from a specific direction while suppressing interference from other directions. The weight coefficient determines the contribution of each microphone to the final output signal, while the delay reflects the difference in the time required for sound to propagate to different microphone positions.

[0081] Among them, the delay τ i The distance d from the sound source to the microphone can be i calculate:

[0082]

[0083] Where c is the speed of sound.

[0084] In practical applications, the system needs to accurately evaluate the acoustic environment characteristics of the target area. First, by analyzing the background noise signal N(t) collected by the microphone array, the noise level L N It can be calculated by the following formula:

[0085]

[0086] The attendee density ρ is defined as the number of attendees per unit area and is calculated as:

[0087]

[0088] Where K is the total number of participants and A is the area of ​​the target area.

[0089] Through the above methods, the system can achieve accurate positioning of the sound source and high-quality collection. In areas with strong noise, the spatial filtering effect can be enhanced by adjusting the weight coefficient; in areas with high density of participants, the spatial resolution can be improved by optimizing the delay parameters. This adaptive adjustment mechanism based on environmental characteristics improves the system's voiceprint collection performance in complex conference environments and provides high-quality audio input for subsequent voiceprint recognition.

[0090] Step 302: Based on the noise level and the density of people, the arrangement of the microphone array is adjusted to a target arrangement and the array size is adjusted to a target array size.

[0091] Among them, the target arrangement mode and target array size refer to the spatial layout form and unit number configuration of the microphone array after the system dynamically adjusts according to the characteristics of the actual meeting environment. In the embodiment of the present application, it can be understood that: the target arrangement mode includes the specific geometric distribution form of the microphone units in space and the relative spacing between the units; the target array size refers to the total number of microphone units participating in the work after optimization.

[0092] For example, in a specific scenario, the target arrangement may be a matrix arrangement with a spacing of 15 cm and a target array size of 16 microphone units; in another scenario, it may be adjusted to a circular arrangement with a spacing of 30 cm and a target array size of 8 microphone units.

[0093] Furthermore, the target arrangement mode and target array size are mainly used to achieve the optimal adaptation of the microphone array to the conference environment. By adjusting these two parameters, the system can select the most suitable spatial sampling strategy for different noise levels and population density characteristics.

[0094] The target arrangement determines the spatial selectivity of the array, affecting the system's ability to distinguish sound sources from different directions; the target array size is directly related to the system's signal processing capabilities, affecting the effects of noise suppression and signal enhancement. This dynamic optimization mechanism of parameter combinations ensures that the system can maintain stable voiceprint collection quality in various conference scenarios.

[0095] Based on the above embodiment, as an optional embodiment, in step 302: adjusting the arrangement of the microphone array to the target arrangement and the array size to the target array size based on the noise level and the personnel density may further include the following steps:

[0096] Step 401: Determine a target arrangement of a microphone array based on population density.

[0097] Specifically, determining the target arrangement of the microphone array based on the density of people means that the system selects the optimal spatial arrangement pattern of the microphone units according to the density of the spatial distribution of the participants in the target area. In an embodiment of the present application, the density ρ of the participants directly reflects the spatial distribution characteristics of the sound source. Under different conditions of personnel density, the system needs to adopt different microphone arrangements to achieve the optimal sound source collection effect. For example, in the check-in queue area where the participants are densely populated, the crosstalk problem between adjacent sound sources is more prominent, and a more sophisticated spatial sampling strategy is needed at this time; in areas where the participants are dispersed, it is more necessary to consider the uniformity of the sound field coverage.

[0098] Furthermore, the system has designed three typical microphone arrangement patterns, targeting different crowd density scenarios.

[0099] When the density of people ρ≥ρ1, it indicates that the sound sources in the target area are extremely densely distributed. The system adopts a square grid arrangement mode, and the microphone units are arranged in an evenly spaced square grid. The spacing between adjacent microphones is usually set to 0.1 to 0.2 meters. This compact orthogonal arrangement can provide the maximum spatial sampling density and effectively improve the system's ability to resolve adjacent sound sources.

[0100] When ρ2≤ρ<ρ1, it indicates that the sound source distribution is in a medium-density state. The system selects a triangular grid arrangement mode, and the microphone units are arranged according to a regular triangular grid. This honeycomb arrangement can provide more uniform sound field coverage while maintaining good spatial resolution.

[0101] When ρ<ρ2, it indicates that the sound source distribution is relatively dispersed. The system adopts a linear arrangement mode. The microphone units are evenly arranged along a straight line. The unit spacing can be appropriately increased to 0.3 to 0.5 meters. This arrangement simplifies the system structure while ensuring the effective collection of dispersed sound sources.

[0102] Step 402: Determine a target array size of the microphone array in a target arrangement based on the noise level.

[0103] After determining the arrangement, it is necessary to further determine the microphone arrangement density, which refers to the physical distance between adjacent microphone units in the microphone array. Its value directly affects the system's spatial sampling accuracy and resolution of the sound source. In a conference scenario, since the distribution density of participants will affect the spatial distribution characteristics of the sound source, the system needs to dynamically adjust the microphone arrangement density according to the actual density of people to obtain the optimal sound source collection effect.

[0104] Furthermore, the embodiment of the present application adopts an adaptive arrangement density adjustment method based on personnel density. By establishing a mathematical relationship model between microphone arrangement density and participant density, the system can automatically calculate the most suitable microphone spacing according to the personnel distribution characteristics in the current conference environment. Specifically, the target arrangement density can be expressed as:

[0105]

[0106] Where, d target is the target arrangement density, i.e., the spacing between microphones, d base is the initial arrangement density in the preset mode, ρ0 is the baseline value of the participant density, ρ is the current participant density, and β is the adjustment coefficient, which indicates the degree of influence of the participant density on the arrangement density.

[0107] The above-mentioned dynamic arrangement density adjustment mechanism based on personnel density establishes an inverse relationship between the arrangement density and the personnel density. The system can automatically improve the spatial sampling accuracy when the participants are dense, and expand the coverage when the participants are sparse, thereby achieving the optimization of resource utilization; secondly, the introduction of the adjustment coefficient β enables the system to flexibly adjust the response characteristics according to actual application needs. When the β value is large, the system is more sensitive to changes in personnel density. When the β value is small, the arrangement density is maintained relatively stable; finally, this adaptive adjustment mechanism can continuously optimize the spatial layout of the microphone array to ensure that ideal voiceprint collection effects can be obtained in different meeting scenarios.

[0108] After determining the arrangement and density of the microphones, the next step is to adjust the size of the microphone array based on the noise level. The size of the microphone array refers to the total number of microphone units involved in the work, and its value directly affects the signal processing capability and noise resistance of the system. In an actual meeting environment, due to the presence of multiple noise sources such as air conditioning systems, equipment operation, and personnel activities, environmental noise will significantly affect the quality of voiceprint signal collection. Therefore, the system needs to dynamically adjust the size of the microphone array according to the current noise level to ensure that the signal-to-noise ratio of the collected signal meets the requirements of voiceprint recognition.

[0109] Furthermore, the embodiment of the present application adopts an adaptive scale adjustment method based on noise level. By establishing a mathematical relationship model between the number of microphones and environmental noise, the system can automatically calculate the required number of microphones according to the noise level monitored in real time.

[0110] Exemplarily, the scale of the target microphone array may be determined using the following formula:

[0111]

[0112] Where N target is the target microphone array size, N base is the initial microphone array size, L0 is the noise threshold preset by the system, and L N is the current noise level, and α is the adjustment coefficient related to the noise level, which indicates the degree of influence of noise on the number of microphones.

[0113] Through the dynamic scale adjustment mechanism based on the noise level, a linear relationship between the number of microphones and the noise level can be established. The system can automatically increase the signal collection points when the noise is strong and reduce redundant units when the noise is weak, thereby realizing on-demand allocation of system resources; secondly, the introduction of the adjustment coefficient α enables the system to adjust the response strength according to actual application requirements. When the α value is large, the system is more sensitive to noise changes. When the α value is small, the array size is maintained relatively stable; finally, this adaptive adjustment mechanism can continuously optimize the system configuration to ensure sufficient signal quality in different noise environments.

[0114] As an optional embodiment, when the arrangement of the microphones has been determined, the area coverage of the microphone array based on different arrangements can be further determined. The area coverage refers to the proportion of the effective collection range of the microphone array in the target area, and its value is directly related to the system's acoustic coverage of the conference space. In an actual conference environment, since different microphone arrangements will form different spatial sampling patterns, it is necessary to calculate the total number of microphones required based on the specific arrangement to ensure effective coverage of the entire target area.

[0115] The total number of target microphones can be calculated by the following formula:

[0116]

[0117] In the formula, A room is the total area of ​​the target region, A cell is the area of ​​each microphone unit.

[0118] For different arrangements, the unit area covered by the microphone can be calculated by the following formula:

[0119] Square grid arrangement:

[0120] Triangular grid arrangement:

[0121] Linear arrangement: A cell =d target l;

[0122] Where l is the length of the microphone coverage area in the linear arrangement.

[0123] The above-mentioned arrangement-based microphone quantity calculation mechanism can establish area coverage models under different arrangement methods. The system can select the optimal arrangement method for different conference space layouts to maximize the spatial sampling efficiency; secondly, the introduction of the unit area calculation formula enables the system to accurately evaluate the coverage characteristics of each arrangement method, thereby optimizing the number of microphones used while ensuring the collection effect; finally, this calculation mechanism can be combined with the aforementioned arrangement density adjustment method to achieve overall optimization of the microphone array deployment.

[0124] Step 102: When the conference sign-in starts, obtain a second sound source collected by the microphone array in target mode, obtain a voiceprint recognition result of the second sound source, and perform conference sign-in based on the recognition result.

[0125] Among them, the second sound source refers to the voice signal of the participant collected by the microphone array in the target mode during the conference check-in stage. In the embodiment of the present invention, it can be understood as the sound information containing personal voice characteristics emitted by the participant at the designated location. The sound information includes but is not limited to the participant reading out his or her own name, work number or preset password. The second sound source is different from the first sound source collected by the system before the check-in starts to adjust the microphone array. It is mainly used for voiceprint recognition and verification of the identity of the participant, and provides voiceprint feature data for subsequent automatic check-in. The sound source has unique acoustic characteristics, including the frequency characteristics, volume, duration, etc. of the sound source. These characteristics together constitute the voiceprint information that can be used for identity recognition.

[0126] Furthermore, the collection process of the second sound source is carried out after the system completes the optimization configuration of the microphone array. At this time, the microphone array has been adjusted to the target mode that best suits the current conference environment and can collect the voice signals of the participants in the best state. By collecting the second sound source, the system can obtain higher quality and more distinctive sound signals, thereby improving the accuracy of subsequent voiceprint recognition, and ultimately achieving reliable identity authentication and automatic sign-in.

[0127] Based on the above embodiment, as an optional embodiment, in step 102: the step of obtaining the voiceprint recognition result of the second sound source may further include the following steps:

[0128] Step 501: Acquire an audio signal corresponding to a second sound source;

[0129] Since the second sound source collected by the microphone array is an analog acoustic signal, it needs to be converted into an audio signal that can be digitally processed for subsequent voiceprint feature extraction and recognition matching. Specifically, the system converts the collected acoustic vibration signal into an electrical signal through each microphone in the microphone array, and then converts the analog electrical signal into a digital audio signal through an analog-to-digital converter.

[0130] During the conversion process, the system uses a suitable sampling frequency and quantization bit number for audio sampling. Considering that the frequency range of human voice is mainly concentrated between 20Hz and 20kHz, this embodiment selects a sampling frequency of 44.1kHz to ensure that the collected audio signal can fully retain the acoustic characteristics of human voice. At the same time, using 16-bit quantization accuracy for sampling can provide sufficient dynamic range to express subtle changes in sound and ensure accurate extraction of voiceprint features.

[0131] In the process of acquiring audio signals, this embodiment also combines the beamforming technology of the microphone array in the target mode. By adjusting the beam direction of each microphone, the system can enhance the sound signal from the target direction while suppressing noise interference from other directions. The audio signal acquired in this way has a higher signal-to-noise ratio, which is conducive to improving the accuracy of subsequent voiceprint recognition.

[0132] Step 502: Separate the audio signal based on the time difference between the microphones in the microphone array to obtain a plurality of audio sub-signals.

[0133] Since multiple participants may speak at the same time in a conference scene, it is necessary to separate the mixed audio signals collected by the microphone array to obtain the independent sound signal of each participant. The system uses the time difference characteristics between the microphones in the microphone array to separate the audio signals. This separation method based on time difference can effectively solve the problem of sound source aliasing.

[0134] Specifically, assuming that the signal emitted by the sound source is S(t), the sound propagation time varies due to the different distances between the participants and the microphones. i The collected signal S i (t) is actually the result of the original signal after time delay, that is:

[0135] S i (t) = S(t-τ i );

[0136] In order to accurately separate the independent audio signals, this embodiment uses the delay and summation method for signal processing. First, the system performs delay compensation on the signal collected by each microphone and adjusts it to:

[0137] S′ i (t) = S i (t+τ i ) = S(t);

[0138] This allows the signals from all microphones to be time synchronized. This delay compensation ensures that signals from the same source can be coherently added, while signals from different sources will be weakened due to time differences.

[0139] After completing the delay compensation, the system performs weighted superposition processing on the synchronized signal to obtain the enhanced signal:

[0140]

[0141] where w i It is a microphone M i The weights are determined based on the performance characteristics of each microphone, mainly considering factors such as the microphone's signal-to-noise ratio and frequency response characteristics. For microphones with higher signal-to-noise ratios, the system will assign greater weights to improve the quality of the separated signal.

[0142] In the laboratory test environment, when three participants spoke at the same time, the system was able to separate the mixed signal into three independent audio sub-signals, and the signal-to-noise ratio of the separated signals increased by about 8dB. Even in a noisy environment, the separated audio sub-signals still maintained good quality, and the voiceprint features were clearly discernible. This separation effect provides a reliable signal basis for subsequent voiceprint recognition, significantly improving the recognition accuracy when multiple people sign in at the same time.

[0143] Through this time difference-based signal separation method, this embodiment not only solves the aliasing problem of the traditional conference sign-in system when multiple people sign in at the same time, but also improves the real-time performance of the system in processing audio signals. The processing delay of the entire signal separation process is controlled within 100 milliseconds, meeting the real-time requirements of the conference sign-in system. At the same time, the computational complexity of this method is relatively low, suitable for implementation in embedded systems, and has good engineering practicality.

[0144] Step 503: Match the voiceprint recognition results corresponding to each audio sub-signal.

[0145] In order to achieve accurate identity recognition, the system needs to extract and match the voiceprint features of each separated audio sub-signal. The core of voiceprint recognition is to convert the audio signal into a feature vector that can uniquely represent the speaker's identity characteristics, and match it with the pre-stored voiceprint features.

[0146] Optionally, this embodiment uses the MFCC method to extract voiceprint features. First, the audio sub-signal is framed, and the length of each frame is set to 25ms, and the frame shift is 10ms to ensure the continuity of signal analysis and the local characteristics of the time domain. After each frame of the signal is processed by Hamming windowing, the spectrum is obtained by fast Fourier transform (FFT). Taking into account the perception characteristics of the human ear to sounds of different frequencies, the system converts the spectrum to a Mel frequency scale that is more in line with the auditory characteristics of the human ear. Subsequently, the Mel spectrum is logarithmically processed to simulate the nonlinear perception characteristics of the human ear to sound intensity. Finally, the logarithmic Mel spectrum is converted into the MFCC feature vector V through discrete cosine transform (DCT).i , the feature vector contains the speaker's vocal tract feature information.

[0147] When performing feature matching, the system adopts a dual judgment mechanism, that is, it uses both Euclidean distance and cosine similarity to calculate the similarity of feature vectors. i , the system calculates its feature vector V with the M participants registered in the voiceprint database j Euclidean distance and cosine similarity between (j=1,2,…,M).

[0148] Among them, the Euclidean distance expression is:

[0149]

[0150] Where d is the dimension of the feature vector, V i,k and V j,k They are the eigenvectors V i and V j The kth component of .

[0151] The cosine similarity expression is:

[0152]

[0153] Where V i ·V j is the dot product of two vectors, ||V i ||||V j || is V i ·V j The norm of .

[0154] Furthermore, this embodiment sets a strict double-threshold decision rule: only when the Euclidean distance is less than the preset threshold θ d And the cosine similarity is greater than the preset threshold θ s This double judgment mechanism can effectively reduce the misrecognition rate and improve the reliability of the system. When a feature vector match that meets the conditions is found, the system can determine the speaker's identity and automatically complete the sign-in record.

[0155] Reference Figure 2 , the present application also provides a conference check-in system based on voiceprint recognition, including:

[0156] A microphone array adjustment module, used for obtaining a first sound source collected by a microphone array in a preset mode in a target area when the conference check-in has not started, and adjusting the preset mode to a target mode based on the first sound source, wherein the preset mode includes an arrangement mode and an array size of the microphone array;

[0157] The voiceprint recognition sign-in module is used to obtain a second sound source collected by the microphone array in the target mode when the meeting sign-in starts, obtain the voiceprint recognition result of the second sound source, and perform meeting sign-in based on the recognition result.

[0158] On the basis of the above embodiment, as an optional embodiment, the microphone array adjustment module is also used to obtain the distribution positions of participants in the target area; when the distribution positions of the participants are statically distributed, based on the distribution positions of the participants and the distribution positions of the microphones in the target area, the target area is divided into at least one sub-area, and the first sound source collected by the microphone array in a preset mode in each sub-area is obtained; when the distribution positions of the participants are dynamically distributed, based on the distribution positions of the participants and the microphones in the target area, a queuing area in the target area is determined, and the first sound source collected by the microphone array in a preset mode in the queuing area is obtained.

[0159] Based on the above embodiments, as an optional embodiment, the microphone array adjustment module is also used to obtain the noise level in the first sound source and the density of participants; based on the noise level and the density of participants, the arrangement of the microphone array is adjusted to a target arrangement and the array size is adjusted to a target array size; the target arrangement and the target array size are determined as a target mode, and the preset mode is adjusted to the target mode; wherein the noise level in the third sound source collected by the microphone array in the target mode is less than the corresponding threshold, and the time difference between the microphones in the microphone array is greater than the corresponding threshold.

[0160] Based on the above embodiment, as an optional embodiment, the microphone array adjustment module is further used to determine the target arrangement of the microphone array based on the personnel density; and determine the target array size of the microphone array under the target arrangement based on the noise level.

[0161] Based on the above embodiments, as an optional embodiment, the voiceprint recognition sign-in module is also used to obtain the audio signal corresponding to the second sound source; separate the audio signal based on the time difference between each microphone in the microphone array to obtain multiple audio sub-signals; and match the voiceprint recognition results corresponding to each audio sub-signal.

[0162] It should be noted that: when the device provided in the above embodiment realizes its function, only the division of the above functional modules is used as an example. In actual application, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the device and method embodiments provided in the above embodiment belong to the same concept, and the specific implementation process is detailed in the method embodiment, which will not be repeated here.

[0163] The embodiment of the present application also provides a computer storage medium, which can store multiple instructions. The instructions are suitable for being loaded by a processor and executing the conference sign-in method based on voiceprint recognition as in the above embodiment. The specific execution process can refer to the specific description of the embodiment shown, which will not be repeated here.

[0164] The present application also discloses an electronic device. Figure 3 , Figure 3 The electronic device 300 may include: at least one processor 301 , at least one network interface 304 , a user interface 303 , a memory 305 , and at least one communication bus 302 .

[0165] The communication bus 302 is used to realize the connection and communication between these components.

[0166] The user interface 303 may include a display interface and a camera interface. The optional user interface 303 may also include a standard wired interface and a wireless interface.

[0167] The network interface 304 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).

[0168] Among them, the processor 301 may include one or more processing cores. The processor 301 uses various interfaces and lines to connect various parts in the entire server, and executes various functions of the server and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 305, and calling data stored in the memory 305. Optionally, the processor 301 can be implemented in at least one hardware form of digital signal processing (Digital Signal Processing, DSP), field programmable gate array (Field-Programmable Gate Array, FPGA), and programmable logic array (Programmable Logic Array, PLA). The processor 301 can integrate one or a combination of a central processing unit (Central Processing Unit, CPU), a graphics processing unit (Graphics Processing Unit, GPU) and a modem. Among them, the CPU mainly processes the operating system, user interface diagrams and applications, etc.; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 301, and it can be implemented separately through a chip.

[0169] Among them, the memory 305 may include a random access memory (Random Access Memory, RAM) and may also include a read-only memory (Read-Only Memory). Optionally, the memory 305 includes a non-transitory computer-readable storage medium. The memory 305 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 305 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 305 may optionally be at least one storage device located away from the aforementioned processor 301. Refer to Figure 3 , the memory 305 as a computer storage medium may include an operating system, a network communication module, a user interface module and an application program of a conference sign-in method based on voiceprint recognition.

[0170] exist Figure 3In the electronic device 300 shown, the user interface 303 is mainly used to provide an input interface for the user and obtain the data input by the user; and the processor 301 can be used to call the memory 305 to store an application for a conference check-in method based on voiceprint recognition. When executed by one or more processors 301, the electronic device 300 executes one or more of the methods described in the above embodiments. It should be noted that for the aforementioned method embodiments, for the sake of simple description, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited to the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required for the present application.

[0171] On the other hand, the present application also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the conference check-in method based on voiceprint recognition provided by the above methods.

[0172] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0173] In the several embodiments provided in this application, it should be understood that the disclosed devices can be implemented in other ways. For example, the device embodiments described above are only schematic, such as the division of units, which is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some service interfaces, and the indirect coupling or communication connection of devices or units can be electrical or other forms.

[0174] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0175] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0176] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a memory and includes several instructions for a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the various embodiments of the present application. The aforementioned memory includes: various media that can store program codes, such as USB flash drives, mobile hard drives, magnetic disks or optical disks.

[0177] The above is only an exemplary embodiment of the present disclosure and cannot be used to limit the scope of the present disclosure. That is, any equivalent changes and modifications made according to the teachings of the present disclosure are still within the scope of the present disclosure. After considering the disclosure of the specification and the truth of practice, those skilled in the art will easily think of other embodiments of the present disclosure.

[0178] This application is intended to cover any variation, use or adaptation of the present disclosure, which follows the general principles of the present disclosure and includes common knowledge or customary technical means in the art not described in the present disclosure. The description and examples are to be regarded as exemplary only, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. A conference sign-in method based on voiceprint recognition, characterized in that: include: When the conference check-in has not started, obtaining a first sound source collected by a microphone array in a preset mode in a target area, and adjusting the preset mode to a target mode based on the first sound source, wherein the preset mode includes an arrangement mode and an array size of the microphone array; When the conference sign-in starts, a second sound source collected by the microphone array in the target mode is obtained, a voiceprint recognition result of the second sound source is obtained, and the conference sign-in is performed based on the recognition result.

2. The conference sign-in method based on voiceprint recognition according to claim 1 is characterized in that: The step of obtaining a first sound source collected by a microphone array in a preset mode in a target area includes: Get the distribution locations of participants in the target area; In the case where the distribution positions of the participants are statically distributed, based on the distribution positions of the participants and the distribution positions of the microphones in the target area, the target area is divided into at least one sub-area, and a first sound source collected by a microphone array in a preset mode in each sub-area is obtained; In the case where the distribution positions of the participants are dynamically distributed, a queuing area in the target area is determined based on the distribution positions of the participants and microphones in the target area, and a first sound source collected by the microphone array in a preset mode in the queuing area is obtained.

3. The conference sign-in method based on voiceprint recognition according to claim 1 is characterized in that: The adjusting the preset mode to a target mode based on the first sound source includes: Acquire the noise level in the first sound source and the density of participants; Adjusting the arrangement of the microphone array to a target arrangement and the array size to a target array size based on the noise level and the density of people; Determining the target arrangement mode and the target array size as a target mode, and adjusting the preset mode to the target mode; The noise level in the third sound source collected by the microphone array in the target mode is less than the corresponding threshold, and the time difference between the microphones in the microphone array is greater than the corresponding threshold.

4. The conference sign-in method based on voiceprint recognition according to claim 3 is characterized in that: The method of adjusting the arrangement of the microphone array to a target arrangement and the array size to a target array size based on the noise level and the density of people includes: Determining a target arrangement of a microphone array based on the personnel density; A target array size of the microphone array in the target arrangement is determined based on the noise level.

5. The conference sign-in method based on voiceprint recognition according to claim 1 is characterized in that: After acquiring the second sound source collected by the microphone array in the target mode, the method further includes: Constructing sound field information corresponding to the second sound source, and adjusting beam directions of microphones in the microphone array in the target mode based on the sound field information.

6. The conference sign-in method based on voiceprint recognition according to any one of claims 1-5, characterized in that: The obtaining the voiceprint recognition result of the second sound source includes: Acquire an audio signal corresponding to the second sound source; Separate the audio signal based on the time difference between the microphones in the microphone array to obtain a plurality of audio sub-signals; Match the voiceprint recognition results corresponding to each of the audio sub-signals.

7. A conference check-in system based on voiceprint recognition, characterized in that: include: A microphone array adjustment module, used for obtaining a first sound source collected by a microphone array in a preset mode in a target area when the conference check-in has not started, and adjusting the preset mode to a target mode based on the first sound source, wherein the preset mode includes an arrangement mode and an array size of the microphone array; The voiceprint recognition sign-in module is used to obtain a second sound source collected by the microphone array in the target mode when the meeting sign-in starts, obtain the voiceprint recognition result of the second sound source, and perform meeting sign-in based on the recognition result.

8. An electronic device, characterized in that: It includes a processor, a memory, a user interface and a network interface, the memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device executes the method as described in any one of claims 1-6.

9. A computer storage medium, characterized in that The computer storage medium stores instructions, and when the instructions are executed, the method according to any one of claims 1 to 6 is performed.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images