Intelligent store conflict identification system, method, device and medium

Abstract

The embodiments of the present specification provide a smart store conflict identification system, method, device and medium, the system includes an environment collection module, the environment collection module is configured to collect environment data through an environment monitoring device deployed in the store; the environment collection module includes an audio collection unit, the audio collection unit is configured to collect first environment data through the environment monitoring device; and a processor, the processor is configured to: determine conflict judgment information based on the first environment data, and based on the conflict judgment information, judge whether a conflict occurs at a target location; in response to a conflict occurring at the target location, determine the target location as a conflict location, and send the conflict judgment information to a clerk terminal.

Description

Technical Field

[0001] The present invention relates to the field of intelligent analysis technology, and in particular to an intelligent store conflict identification system, method, device and medium. Background Art

[0002] With the rapid development of modern retail and service industries, store management has become increasingly complex and challenging. Ensuring the safety of customers and employees in stores and maintaining a good shopping or service experience have become the focus of managers. Traditionally, the detection and response of store conflicts rely on manual monitoring or customer complaints, which is not only inefficient, but also often difficult to capture and effectively handle conflict incidents in real time, which may lead to escalation of the situation and affect the store's reputation and customer satisfaction.

[0003] In order to overcome the above defects, intelligent conflict recognition systems based on audio monitoring technology have gradually emerged in recent years. Such systems can capture various sound information in the store in real time by integrating highly sensitive audio acquisition modules. However, the store environment is complex and changeable, and the audio data is often mixed with background noise, music, machine operation sound and other interference factors, which puts extremely high demands on audio signal processing and conflict recognition.

[0004] Therefore, the market is in urgent need of an intelligent store conflict identification system to address the limitations of traditional monitoring methods in conflict identification, achieve real-time, accurate identification and rapid response to store conflict events, provide strong technical support for store management, and significantly improve the safety and efficiency of store management. Summary of the invention

[0005] One or more embodiments of this specification provide a smart store conflict identification system. The system includes: an environment acquisition module, which is configured to collect environment data through an environment monitoring device deployed in the store; the environment acquisition module includes an audio acquisition unit, which is configured to collect first environment data through the environment monitoring device; and a processor, which is configured to: determine conflict judgment information based on the first environment data, and determine whether a target position conflicts based on the conflict judgment information; in response to a conflict at the target position, determine the target position as a conflict position, and send the conflict judgment information to the clerk terminal.

[0006] One or more embodiments of this specification provide a smart store conflict identification method. The method is applied to a smart store conflict identification system, the system includes an environment acquisition module, the environment acquisition module is configured to collect environment data through an environment monitoring device deployed in the store; the environment acquisition module includes an audio acquisition unit, the audio acquisition unit is configured to collect first environment data through the environment monitoring device; the method includes: based on the first environment data, determining conflict judgment information, and based on the conflict judgment information, judging whether a target position conflicts; in response to a conflict at the target position, determining the target position as a conflict position, and sending the conflict judgment information to a clerk terminal.

[0007] One or more embodiments of this specification provide a smart store conflict identification device. The device includes: a processor and at least one memory; the at least one memory stores computer instructions; the processor executes at least part of the computer instructions to implement any of the methods described above.

[0008] One or more embodiments of the present specification provide a computer-readable storage medium, wherein the storage medium stores computer instructions. When a computer reads the computer instructions in the storage medium, the computer executes any of the methods described above. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] This specification will be further described in the form of exemplary embodiments, which will be described in detail by the accompanying drawings. These embodiments are not restrictive, and in these embodiments, the same number represents the same structure, wherein:

[0010] Figure 1 is an exemplary application scenario diagram of the smart store conflict identification system according to some embodiments of this specification;

[0011] Figure 2 is an exemplary module diagram of a smart store conflict identification system according to some embodiments of this specification;

[0012] Figure 3 is an exemplary flow chart of determining whether a target position conflicts according to some embodiments of this specification;

[0013] Figure 4 is a schematic diagram of a conflict determination model according to some embodiments of this specification;

[0014] Figure 5 is a schematic diagram of a feature acquisition model according to some embodiments of this specification;

[0015] Figure 6is an exemplary flow chart of determining coordination parameters according to some embodiments of the present specification. DETAILED DESCRIPTION

[0016] In order to more clearly illustrate the technical solutions of the embodiments of this specification, the following is a brief introduction to the drawings required for the description of the embodiments. Obviously, the drawings described below are only some examples or embodiments of this specification. For ordinary technicians in this field, this specification can also be applied to other similar scenarios based on these drawings without creative work. Unless it is obvious from the language environment or otherwise explained, the same reference numerals in the figures represent the same structure or operation.

[0017] It should be understood that the "system", "device", "unit" and / or "module" used herein are a method for distinguishing different components, elements, parts, portions or assemblies at different levels. However, if other words can achieve the same purpose, the words can be replaced by other expressions.

[0018] As shown in this specification and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "comprise" and "include" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0019] Flowcharts are used in this specification to illustrate the operations performed by the system according to the embodiments of this specification. It should be understood that the preceding or following operations are not necessarily performed precisely in order. Instead, the steps may be processed in reverse order or simultaneously. At the same time, other operations may also be added to these processes, or one or more operations may be removed from these processes.

[0020] Figure 1 This is an exemplary application scenario diagram of the smart store conflict identification system shown in some embodiments of this specification.

[0021] like Figure 1 As shown, the application scenario may include an environment monitoring device 110 , a processor 120 , and a store clerk terminal 130 .

[0022] In some embodiments, one or more components of the smart store conflict identification system 100 (hereinafter referred to as the system 100) may transmit data to other components of the system 100 via a network. For example, the processor 120 may obtain information and / or data from the environment monitoring device 110 and the store clerk terminal 130 via a network, or may send information and / or data to the environment monitoring device 110 and the store clerk terminal 130 via a network.

[0023] In some embodiments, the environment monitoring device 110 and the store clerk terminal 130 , as well as other possible system components, may include a storage device.

[0024] In some instances, different functions may be performed on different devices, such as data screening, querying, preprocessing, model training, model execution, etc., and this specification does not limit this.

[0025] Figure 2 It is an exemplary module diagram of the smart store conflict identification system shown in some embodiments of this specification.

[0026] In some embodiments, Figure 2 As shown, the system 100 may include an environment acquisition module 140 and a processor 120 .

[0027] The environment collection module 140 is used to collect the environment data in the physical store in real time. The environment data refers to the relevant data reflecting the environment conditions in the store, such as the temperature in the store, the humidity in the store, the audio in the store, the image in the store, and other data. In some embodiments, the environment collection module 140 can be configured to collect the environment data in real time through the environment monitoring device 110 deployed in the store. In some embodiments, the environment collection module 140 includes an audio collection unit 141.

[0028] The environmental monitoring device 110 refers to a device used to monitor and record environmental data in real time. Figure 1 As shown, in some embodiments, the environment monitoring device 110 may include but is not limited to an audio monitoring device 111, an image acquisition device 112, a human flow monitoring device, a temperature and humidity monitoring device, a light intensity monitoring device, an air quality monitoring device, and the like.

[0029] The audio collection unit 141 refers to a unit in the environment collection module 140 for collecting audio data in the store. In some embodiments, the audio collection unit 141 is configured to collect the first environment data through the audio monitoring device 111 of the environment monitoring device 110.

[0030] like Figure 1 As shown, the environment monitoring device 110 may include an audio monitoring device 111, which refers to a device for real-time monitoring and recording of audio. In some embodiments, the audio monitoring device 111 may be one of devices such as a microphone, an array microphone, or an audio sensor, or any combination thereof.

[0031] For more information about the first environment data, please see below.

[0032] In some embodiments, Figure 2As shown, the environment monitoring device 110 includes an image acquisition device 112, and the environment acquisition module 140 includes an image acquisition unit. The image acquisition device 112 refers to a device for converting actual scene acquisition into digital images or videos. In some embodiments, the image acquisition device 112 may include but is not limited to one or any combination of a camera, a scanner, an infrared camera, etc. The image acquisition unit refers to a unit in the environment acquisition module 140 for collecting image data in the store. The image acquisition unit can convert analog signal graphics into digital image data. In some embodiments, the image acquisition unit is configured to collect environmental image data through the image acquisition device 112. In some embodiments, the image acquisition unit can convert the analog image signal (such as an optical signal captured by a camera, infrared photography, etc.) collected by the image acquisition device 112 into a digital image signal, and perform preliminary processing (such as format conversion, storage, etc.).

[0033] The processor 120 is used to process data and / or information obtained from other devices or system components. The processor 120 can execute program instructions based on these data, information and / or processing results to perform one or more functions described in this application, such as determining conflict judgment information, determining whether a conflict occurs, determining a conflict location, etc.

[0034] In some embodiments, processor 120 may include one or more sub-processing devices (e.g., a single-core processing device or a multi-core multi-core processing device). As an example only, processor 120 may include a central processing unit (CPU), an application specific integrated circuit (ASIC), a microprocessor, etc., or any combination thereof.

[0035] In some embodiments, the processor 120 is configured to determine conflict judgment information based on the first environment data, and determine whether a conflict occurs at the target location based on the conflict judgment information; in response to a conflict at the target location, determine the target location as a conflict location, and send the conflict judgment information to the clerk terminal 130 (see Figure 2 For more information about the functions of processor 120, see Figure 3-6 Description.

[0036] The clerk terminal 130 refers to one or more terminal devices or software used by the clerk. The clerk terminal 130 may include a processing unit, a display unit, an input / output unit, a storage unit, etc. For example, the display device of the clerk terminal 130 may display the current conflicting position (e.g., cashier area, display area, dining area), etc. to the clerk.

[0037] In some embodiments, the system 100 further includes a clerk information acquisition module, which is used to acquire clerk information. The clerk information includes clerk number, clerk location, working hours, job level, etc. In some embodiments, the clerk information acquisition module can be configured to acquire clerk information based on information tags. Information tags include Bluetooth beacons, RFID tags, etc. installed in clerk badges. Clerk information can be acquired through Bluetooth beacons, RFID tags, etc. installed in clerk badges. For example, the clerk location and employee number are acquired through information tags, and then the working hours and job level are queried based on the employee number.

[0038] In some embodiments of this specification, the intelligent store conflict identification system can capture the environmental changes in the store in real time through the environment collection module, especially the audio collection unit, and more accurately and efficiently identify and judge conflict events to ensure the safety of customers and employees and the order of the store; the system makes a preliminary judgment based on the audio features, and combines the store space characteristics, environmental image data and other multi-dimensional information for comprehensive analysis to improve the accuracy and reliability of conflict identification. For situations where the preliminary judgment is doubtful, the false alarm rate is reduced through further verification of the image data, ensuring the accuracy of conflict identification; the system can automatically evaluate the severity of the conflict, and set the alarm level and collaborative monitoring parameters accordingly to achieve intelligent conflict management. According to the development trend of the conflict, the monitoring and response measures are dynamically adjusted, such as raising the alarm level, sending assistance instructions to specific mediation clerks, etc., which greatly improves the efficiency and effect of conflict handling; the system continuously optimizes the conflict judgment model and regulation strategy. This personalized conflict prevention strategy not only improves the adaptability of the system, but also effectively predicts and prevents the occurrence of potential conflicts, creating a more harmonious and safe shopping environment for the store; the system can reduce the adverse effects of conflicts on customer shopping experience and employee working environment. At the same time, intelligent conflict management also reduces the workload of store clerks, allowing them to focus more on improving service quality and customer satisfaction.

[0039] It should be understood that Figure 2 The system and its modules shown can be implemented in various ways. For example, in some embodiments, the image acquisition unit can be written in different programming languages.

[0040] It should be noted that the above description of the smart store conflict identification system and its modules is only for the convenience of description, and this specification cannot be limited to the scope of the embodiments cited. It can be understood that for those skilled in the art, after understanding the principle of the system, it is possible to arbitrarily combine the various modules, or form a subsystem to connect with other modules without deviating from this principle. In some embodiments, the environment acquisition module and processor disclosed in the figure can be different modules in a system, or a module can implement the functions of two or more modules mentioned above. For example, each module can share a storage module, or each module can have its own storage module. Such variations are within the scope of protection of this specification.

[0041] Some embodiments of this specification provide a smart store conflict identification method, which is applied to the aforementioned system 100. For more information about the system 100, see Figure 1 and Figure 2 Related description.

[0042] In some embodiments, the processor 120 determines conflict judgment information based on the first environment data, and determines whether a conflict occurs at the target location based on the conflict judgment information.

[0043] The first environment data refers to the original environment data collected by the environment collection module 110. For example, the (unprocessed) original sound data collected by the audio collection unit 111. The duration of the original sound data can be preset, for example, 20s, 30, 1min, etc. In some embodiments, the original sound data can be the original sound data of the installation location of the audio monitoring device 110-1 in the store.

[0044] In some embodiments, the processor 120 may obtain the first environment data in a variety of ways. For example, the processor may send a computer instruction to obtain the first environment data to the environment acquisition module 110 to obtain the first environment data from the environment acquisition module 110. For another example, the processor 120 may also receive the first environment data actively uploaded by the environment acquisition module 110.

[0045] The target location refers to the location area corresponding to the first environment data, for example, the area within the range covered by the first environment data.

[0046] In some embodiments, the target location can be determined in a variety of ways. For example, the target location can be determined by the user by pre-dividing the store, such as into a cashier area, a dining area, etc., and each divided area is used as the target location, and an environmental monitoring device can be installed in each area. For another example, the target location can also be determined based on the installation location of the audio monitoring device, such as using the area within a preset radius centered on the environmental monitoring device as the target location corresponding to the first environmental data collected by the environmental monitoring device. Among them, the preset radius can be set manually as needed.

[0047] The conflict judgment information refers to information used to judge whether a conflict occurs at a target location. In some embodiments, the conflict judgment information may include the volume, duration, target location, etc. of the original sound data in the first environment data. The volume is represented by a decibel value of the volume.

[0048] In some embodiments, the processor 120 may perform audio feature extraction on the original sound data in the first environment data to obtain conflict judgment information corresponding to the original sound data. As an example only, the audio feature extraction may include converting the original sound data into a digital signal and analyzing the amplitude value of the sound wave in the digital signal; and calculating the volume according to the amplitude value. Example calculation formula: dB = 20*log 10 (A / Aref); dB is the decibel value, which indicates the intensity or loudness of the sound; A is the amplitude of the sound, usually a specific measurement value, the unit can be Pascal (Pa) or micro Pascal (μPa), etc.; Aref is the reference amplitude, usually 20 micro Pascal (μPa).

[0049] In some embodiments, the processor 120 can determine whether a conflict occurs at the target location based on the conflict judgment information in a variety of ways. For example, if the decibel level corresponding to the original sound data in the conflict judgment information exceeds a preset decibel threshold, and the duration of the decibel level exceeding the decibel threshold exceeds a preset time threshold, then it is determined that a conflict occurs at the target location corresponding to the first environment data; otherwise, it is determined that no conflict occurs.

[0050] In some embodiments, the decibel threshold is positively correlated to the daily average decibel of the target location. For example, the processor 120 can calculate the decibel average of the original sound data in the plurality of first environmental data corresponding to the target location within a preset period as the daily average decibel; and adjust the value based on the daily average decibel as the decibel threshold. The higher the daily average decibel, the higher the decibel threshold. The preset period can be a recent period without conflict.

[0051] For other methods of determining the conflict judgment information based on the first environment data and judging whether a conflict occurs at the target position, see Figure 3 and its related description.

[0052] In some embodiments, in response to a conflict at the target location, the processor 120 determines the target location as a conflict location and sends conflict determination information to the clerk terminal.

[0053] The conflict location refers to the location where the conflict occurs, such as the specific location where a verbal conflict or a physical conflict occurs.

[0054] In some embodiments, the processor 120 may determine that a target position where a conflict occurs as a conflict position.

[0055] In some embodiments, the processor 120 sends the conflicting position to the clerk terminal. After the clerk terminal receives the information, the clerk can take appropriate processing measures based on the conflict judgment information.

[0056] In some embodiments of the present specification, by determining whether a conflict occurs through the first environmental data, the conflict location in the store can be discovered in time so that the store staff can handle it in time, thereby ensuring the normal operation of the store.

[0057] Figure 3 FIG. 1 is an exemplary flow chart of determining whether a target position conflicts with another target position in some embodiments of this specification. Figure 3 As shown, the process 300 includes the following steps. In some embodiments, the process 300 may be executed by the processor 120 of the system 100 .

[0058] Step 310: pre-process the first environment data to obtain second environment data.

[0059] The second environment data refers to data obtained after preprocessing the first environment data. In some embodiments, the preprocessing may include noise reduction processing, enhancement processing, etc.

[0060] Denoising refers to a process of reducing or eliminating noise in sound data. In some embodiments, the processor may perform denoising on the original sound data in the first environment data by using a denoising algorithm such as frequency domain filtering, wavelet transform, or spectral subtraction.

[0061] Enhancement processing refers to a processing method for enhancing the sound in the sound data. Enhancement may include enhancing the amplitude of the sound, etc. In some embodiments, the processor may enhance the original sound data in the first environment data through enhancement algorithms such as dynamic range compression, gain adjustment, and echo cancellation.

[0062] In some embodiments, the processor may also determine the second environment data through a preprocessing layer of the feature acquisition model. For relevant descriptions of the preprocessing layer of the feature acquisition model, see Figure 5 and related instructions.

[0063] Step 320: extracting environmental audio features based on the second environmental data.

[0064] The environmental audio features refer to the audio features of the sound in the environment, such as the frequency, amplitude and other related features of the sound. In some embodiments, the environmental audio features may include an audio waveform diagram, a spectrum diagram and the like.

[0065] An audio waveform is an image that shows how sound amplitude changes over time.

[0066] In some embodiments, the processor 120 may draw an audio waveform graph based on the second environment data. For example, the processor 120 may convert the audio in the second environment data into a digital signal; perform spectrum analysis on the digital signal to obtain the amplitude of the audio at each time point; and draw an audio waveform graph based on the amplitude of the audio at each time point.

[0067] A spectrogram is an image that shows how the energy of different frequencies in a sound changes over time.

[0068] In some embodiments, the processor 120 may draw a spectrogram based on the second environment data. For example, the processor 120 may perform windowing on the audio in the second environment data to obtain multiple windows, wherein the windowing method may include a Hamming window, a Hanning window, etc.; perform a Fourier transform on each window to convert each window from a time domain signal to a frequency domain signal; and draw a spectrogram based on the Fourier transformed window. The spectrogram may include a static spectrogram of a single window, a dynamic spectrogram over time, etc.

[0069] In some embodiments, the processor 120 may also determine the ambient audio features through an extraction layer of a feature acquisition model. For a description of the extraction layer of the feature acquisition model, see Figure 5 and its related description.

[0070] Step 330 , in response to the environmental audio features satisfying the preset judgment conditions, conflict judgment information is determined based on the environmental audio features and the store space features.

[0071] The preset judgment condition refers to a condition used to determine whether a conflict judgment is required. For example, the average amplitude in the ambient audio feature exceeds the preset amplitude threshold. The average amplitude can be obtained based on the audio waveform graph, and the preset amplitude threshold can be set by the user or technician.

[0072] It can be understood that when the ambient audio characteristics meet the preset judgment conditions, the possibility of a conflict at the target position is relatively high, and it is necessary to determine the conflict judgment information to determine whether a conflict occurs.

[0073] Store space features are features that characterize the store space layout. For example, the size information, partition information (such as cashier area, display area, dining area, etc.) and the spatial location of the target location in the store. The spatial location of the target location refers to the relative location information of the target location in the store space.

[0074] In some embodiments, the store space characteristics can be measured by the user or technician. For example, the size information of the store can be measured by the technician when installing the system 100 and input into the system 100; the partition information can be obtained by the user or technician manually dividing; the spatial position of the target location can be determined by the user or technician manually marking the installation position of the environmental monitoring device corresponding to the sound data.

[0075] In some embodiments, the conflict judgment information may also include the conflict type of the target location and its corresponding confidence level. The conflict type may include verbal conflict, physical conflict, etc. The confidence level refers to a numerical value that characterizes the degree of credibility of the actual occurrence of the conflict type. The confidence level may correspond to the conflict type, that is, the confidence level may include the confidence level corresponding to the verbal conflict, the confidence level corresponding to the physical conflict, etc.

[0076] In some embodiments, the processor 120 can construct a sound feature vector based on environmental audio features and store space features; retrieve a reference feature vector with the highest similarity in a first vector database based on the sound feature vector; use the reference conflict type corresponding to the reference feature vector with the highest similarity as the conflict type that may occur at the target location, and use the similarity between the sound feature vector and the reference conflict vector as the confidence level of the conflict type.

[0077] The first vector database includes a large number of reference feature vectors and their corresponding reference conflict types. In some embodiments, the first vector database can be constructed based on historical conflict data, and the historical conflict data includes historical conflict types when conflicts actually occurred in historical records, as well as historical first environmental data and historical store space characteristics collected when conflicts occurred. For example, the processor 120 can determine historical conflict judgment information based on the historical first environmental data according to the methods of steps 310 to 320 above; and construct a reference feature vector based on the historical conflict judgment information and historical store space characteristics, and use the historical conflict type as the reference conflict type corresponding to the reference feature vector.

[0078] In some embodiments, the processor 120 may also determine the conflict judgment information through a conflict judgment model. For more information about the conflict judgment model, see Figure 4 and its related description.

[0079] Step 340: Based on the conflict judgment information, determine whether a conflict occurs at the target position.

[0080] In some embodiments, the processor 120 may determine whether a conflict occurs at the target location based on at least one of a confidence level corresponding to a verbal conflict and a confidence level corresponding to a physical conflict.

[0081] In some embodiments, in response to the confidence level corresponding to the conflict type in the conflict judgment information being greater than a preset threshold, for example, the confidence level corresponding to the language conflict is greater than the language threshold and / or the confidence level corresponding to the physical conflict is greater than the physical threshold, the processor 120 determines that a conflict occurs at the target location. The preset threshold includes a physical threshold and a language threshold, and the preset threshold is determined based on the store type and feedback information.

[0082] The physical threshold refers to the threshold for judging whether a physical conflict has occurred. The verbal threshold refers to the threshold for judging whether a verbal conflict has occurred.

[0083] In some embodiments, the processor 120 may determine the preset threshold by querying a preset threshold table based on the store type. The store type refers to the type of store business content, such as a coffee shop, a comprehensive supermarket, a fast food restaurant, etc. The preset threshold table includes body threshold information and language thresholds corresponding to different store types.

[0084] In some embodiments, the preset threshold table can be determined based on the confidence corresponding to the conflict type when physical conflicts or verbal conflicts actually occur in different store types within a preset period of time. For example, the processor 120 can count the confidence of the conflict type determined based on the sound feature vector when physical conflicts or verbal conflicts actually occur in different store types within a month, and use the minimum value of the confidence corresponding to the physical conflicts and the minimum value of the confidence corresponding to the verbal conflicts as the physical threshold and the verbal threshold of the store type, respectively, to construct the preset threshold table.

[0085] In some embodiments, the processor 120 may also update and adjust the preset threshold based on feedback information. Feedback information refers to information that provides feedback on the judgment and handling of conflict situations, such as information that provides feedback on the timeliness and sensitivity of the judgment and handling of conflict situations. As an example only, when a customer provides feedback that no store clerk handles the conflict in a timely manner, or a store clerk provides feedback that no reminder of the conflict has been received in a timely manner, the processor 120 may lower the preset threshold to more sensitively monitor conflicts that occur in the store.

[0086] In some embodiments of the present specification, by determining the preset threshold according to the store type, the environmental conditions and customer behavior patterns of different stores can be taken into account, so that the intelligent store conflict judgment system can better adapt to the specific needs of different stores and improve the accuracy of conflict judgment; by updating and adjusting the preset threshold through feedback information, the system can be updated according to the actual situation, improve the sensitivity of conflict identification, so as to better take countermeasures.

[0087] In some embodiments, in response to the conflict judgment information satisfying the verification condition, the processor 120 may determine the verification judgment information based on the environmental image data; and determine whether a conflict occurs at the target location based on the verification judgment information.

[0088] The verification condition refers to a condition for further verifying whether a conflict occurs at the target location. In some embodiments, the verification condition may include that the confidence level corresponding to the conflict type is less than a preset threshold and greater than a verification threshold. When the verification condition is met, although it cannot be directly determined that a conflict occurs at the target location, there is still a greater possibility of a conflict, so it is necessary to further verify whether a conflict occurs at the target location. The verification threshold may also include a body verification threshold and a language verification threshold, respectively. The verification threshold may be preset, and the verification threshold may be set to be less than the preset threshold.

[0089] Environmental image data refers to image data collected by the environmental monitoring module, for example, image data in the area corresponding to the target location.

[0090] In some embodiments, the processor 120 may obtain the environmental image data in a variety of ways. For example, the processor 120 may directly obtain the environmental image data from the image acquisition unit. For another example, the processor 120 may also accept the environmental image data actively uploaded by the image acquisition unit.

[0091] Verification judgment information refers to information used to further judge whether a target location conflicts, such as the facial expression of the guest. The facial expression of the guest may include anger, fear, surprise, indifference, happiness, etc.

[0092] In some embodiments, the processor 120 can recognize the facial expression of the guest in the environmental image data in a variety of ways to obtain verification judgment information. For example, the processor 120 can recognize the environmental image data through Microsoft Azure, Face2Exp model, GFFT model, etc., to obtain the facial expression of the guest as verification judgment information.

[0093] In some embodiments, if the guest's facial expression in the verification judgment information shows a preset expression, the processor may determine that a conflict occurs in the target position corresponding to the verification judgment information. The preset expressions may include anger, fear, sadness, surprise, disgust, etc.

[0094] In some embodiments of the present specification, for conflict judgment information that meets the verification conditions, the verification judgment information is determined through environmental image data to further determine whether a conflict occurs, which can improve the system's accuracy in identifying conflicts and avoid some conflicts being missed that may have adverse effects on store operations.

[0095] In some embodiments of the present specification, by processing the first environment to obtain environmental audio features and using them to determine conflict judgment information, the accuracy and timeliness of the conflict judgment results can be improved, helping store clerks to promptly and reasonably select and effectively handle the situation, further ensuring the normal operation of the store.

[0096] In some embodiments, the processor 120 may also determine the conflict judgment information through a conflict judgment model based on environmental audio features, store space features, store environment features, clerk information, and historical conflict judgment information.

[0097] Store environment features are features related to the store environment, such as temperature, humidity, table distribution, customer flow, etc.

[0098] In some embodiments, the temperature and humidity of the store can be obtained through the environmental monitoring device respectively; the table distribution can be pre-entered and set by the user, such as inputting the layout diagram of the store, etc.; the customer flow can be obtained through the counting device installed at the entrance of the store, and the counting device can include a counter, a camera, etc. It can be understood that the counting device can be considered to be included in the environmental monitoring device.

[0099] Historical conflict judgment information refers to the conflict judgment information determined in the historical period. For example, the conflict judgment information determined in the past week. For more information about conflict judgment information, see Figure 3 In some embodiments, the conflict determination information also includes conflict trend data. The conflict trend data refers to data on how the conflict changes over time, such as conflict escalation, conflict de-escalation, conflict continuation, etc.

[0100] For more information about ambient audio characteristics, store space characteristics, and store clerk information, see Figure 1 , Figure 3 and its related description.

[0101] The conflict judgment model refers to a model used to determine the conflict judgment information. In some embodiments, the conflict judgment model may include a machine learning model, such as any one or combination of a recurrent neural network (RNN) model or other custom model structures.

[0102] Figure 4 It is a schematic diagram of a conflict judgment model shown in some embodiments of this specification.

[0103] like Figure 4As shown, the input of the conflict judgment model 420 includes environmental audio features 411, store space features 412, store environment features 413, clerk information 414, and historical conflict judgment information 415, and the output of the conflict judgment model 420 includes conflict judgment information 430.

[0104] In some embodiments, the processor 120 can train the initial conflict judgment model through a large number of first samples with first labels. The first sample may include sample environmental audio features, sample store space features, sample store environment features, sample clerk information, and sample historical conflict judgment information. The first label may include the corresponding actual collected conflict judgment information under the first sample condition. For example, the processor can use the historical environmental audio features, historical store space features, historical store environment features, historical clerk information, and historical conflict judgment information of different types of stores in historical data when conflicts occur as the first sample, and construct the corresponding first label according to the actual conflict that occurred.

[0105] In some embodiments, the processor 120 inputs a plurality of first samples with first labels into the initial conflict judgment model, constructs a loss function through the first labels and the prediction results of the initial conflict judgment model, updates the initial conflict judgment model based on the iteration of the loss function, and completes the conflict judgment model training when the loss function satisfies a preset iteration condition. The preset iteration condition may be that the loss function converges, the number of iterations reaches a set value, etc.

[0106] In some embodiments of the present specification, conflict judgment information is determined through a conflict training model, so that the determined conflict judgment information can be made more accurate; by incorporating conflict trend data into the conflict judgment information, the conflict judgment information can be made more detailed and more referenceable, making it easier for store clerks to make appropriate handling measures based on the conflict judgment information.

[0107] In some embodiments, the first environmental data is preprocessed to obtain the second environmental data, including: based on a preprocessing layer of a feature acquisition model, the first environmental data and store environmental features are processed to obtain the second environmental data and a processing effect score.

[0108] In some embodiments, extracting environmental audio features based on the second environmental data includes: processing the second environmental data and the processing effect score based on an extraction layer of a feature acquisition model to obtain environmental audio features.

[0109] Figure 5 It is a schematic diagram of a feature acquisition model shown in some embodiments of this specification.

[0110] The feature acquisition model is a model that users use to acquire the features of the ambient audio. The feature acquisition model is a machine learning model. Figure 5As shown, in some embodiments, the feature acquisition model 520 may include a preprocessing layer 521 and an extraction layer 522 .

[0111] The preprocessing layer 521 refers to a model layer for preprocessing the first environment data. In some embodiments, the preprocessing layer 521 can be any one or combination of a convolutional recurrent neural network (CRNN) model, a residual network (ResNet) model, or other custom model structures.

[0112] like Figure 5 As shown, in some embodiments, the input of the preprocessing layer 521 may include first environmental data 511 , store environmental characteristics 512 , and store space characteristics 513 ; the output may include second environmental data 531 and a processing effect score 532 .

[0113] The processing effect score is a score that characterizes the preprocessing effect. The preprocessing effect can indicate the effectiveness of the audio in the second environment data relative to the audio in the first environment data. For example, the less background noise in the audio in the second environment data and the louder the guest's voice, the higher the processing effect score.

[0114] For more information about the first environment data, see Figure 2 and its related descriptions; for the store space characteristics and second environment data, see Figure 3 and its related descriptions; for relevant content on store environment characteristics, see Figure 4 and its related description.

[0115] In some embodiments, the preprocessing layer can be obtained by training the initial preprocessing layer with a large number of second samples with second labels. The second samples may include sample first environment data, sample store environment characteristics, and sample store space characteristics; the second label may include the actual second environment data corresponding to the second sample and the processing effect score.

[0116] In some embodiments, the second sample and the second label can be obtained based on experimental data. As an example only, the experimenter can simulate various types of conflicting or non-conflicting speaking sounds in different sample stores; at the same time, the sample first environment data is collected through the installed environmental monitoring device, and the sample first environment data, sample store environmental characteristics and sample store space characteristics obtained during the experimental simulation are used as the second sample; the sound data is collected by a microphone worn on the experimenter's mouth as the actual second environment data; the similarity between the sample first environment data and the actual second environment data is calculated. The lower the similarity, the more noise is removed in the first environment data, the more obvious the enhancement effect of the guest's voice is, and the higher the processing effect score is. The calculation method can be preset. The actual second environment data collected in the experiment and the calculated processing effect score are used as the second label.

[0117] The training process of the preprocessing layer is similar to that of the conflict judgment model. For details about the training process of the conflict judgment model, see Figure 4 and its related description.

[0118] In some embodiments, the input to the pre-processing layer also includes device parameter data of the audio monitoring device and background noise characteristics.

[0119] Device parameter data refers to data related to the audio monitoring device. For example, the device type information, status information, and device location information of the audio monitoring device. Among them, the status information may include the power of the audio monitoring device, connection stability, etc. The connection stability can be determined based on the fluctuation amplitude and fluctuation frequency of the data transmission speed between the audio monitoring device and the processor. The larger the fluctuation amplitude and the higher the fluctuation frequency, the lower the connection stability. Device location information refers to the installation location of the audio monitoring device in the store. The device parameter data can be input and determined by the technician after installation.

[0120] Background noise features refer to features related to background noise in a store, such as audio waveforms and spectrograms of equipment operating noise, background music noise, or other continuous noise in a store.

[0121] In some embodiments, the processor 120 can directly read the background music data being played, determine the corresponding audio waveform graph and spectrum graph based on the background music data, and obtain the background noise characteristics.

[0122] In some embodiments, the processor 120 may also perform spectrum analysis on the historical first environment data through a spectrum analysis tool (such as Fast Fourier Transform, FFT), for example, identifying background noise frequency bands in the spectrum graph and extracting their audio features as background noise features.

[0123] It is understandable that background noise continuously appears in the store, and the processor 120 can continuously monitor the first environmental data and identify the sounds that have appeared in the continuously monitored first environmental data as background noise.

[0124] For the relevant description and determination method of audio waveform and spectrum diagram, see Figure 3 and its related description.

[0125] In some embodiments, when the input of the preprocessing layer includes device parameter data and background noise characteristics, the preprocessing sample includes sample device parameter data and sample background noise characteristics. The sample device parameter data and sample background noise characteristics can be determined according to the above method in an experimental scenario.

[0126] In some embodiments of the present specification, using device parameter data and background noise characteristics as inputs of the processing layer can improve the preprocessing effect of the preprocessing layer, thereby improving the accuracy of the environmental audio characteristics extracted by the subsequent extraction layer.

[0127] The extraction layer 522 is a model layer for extracting ambient audio features. In some embodiments, the extraction layer 522 can be any one or combination of a convolutional neural network (CNN) model or other custom model structures.

[0128] like Figure 5 As shown, the input of the extraction layer 522 may include the second environment data 531, the processing effect score 532 and the store type 540; the output of the extraction layer 522 includes the environment audio feature 550. Among them, the second environment data 531 and the processing effect score 532 are the outputs of the pre-processing layer 521.

[0129] For relevant data on the second environment data, processing effect score, store type, and environmental audio characteristics, please refer to the previous text and its related instructions.

[0130] In some embodiments, the extraction layer can be obtained by training the initial extraction layer based on a large number of third samples with third labels. The third samples may include sample second environment data, sample processing effect scores, and sample store types; and the third labels may include environmental audio features corresponding to the third samples.

[0131] In some embodiments, the third sample and the third label can be obtained based on experimental data. For example, the processor 120 can directly use the actual second environment data and the processing effect score in the second label as the sample second environment data and the sample processing effect score, and use the experimental store type as the sample store type, thereby forming the third sample; and use the environmental audio features actually collected based on the actual second environment data as the third label. For the relevant content of the second label, see the description of the training data of the preprocessing layer; for the content of determining the environmental audio features based on the second environment data, see Figure 3 and related instructions.

[0132] The training process of the extraction layer is similar to that of the conflict judgment model. For details about the training process of the conflict judgment model, see Figure 4 and its related description.

[0133] In some embodiments, the processor 120 can determine the misjudgment frequency based on historical conflict information and historical verification information; in response to the misjudgment frequency and the processing effect score satisfying the control conditions, determine the device control parameters based on the misjudgment frequency and the processing effect score.

[0134] Historical conflict information refers to the conflict judgment information in the historical conflict judgment record. Historical verification information refers to the verification judgment information in the historical conflict judgment record. For relevant instructions on conflict judgment information and verification judgment information, see Figure 3 and its related description.

[0135] The misjudgment frequency refers to the frequency of misjudgment when judging whether a conflict occurs, for example, the frequency of judging that no conflict occurs according to the conflict judgment information, but judging that a conflict occurs according to the verification judgment information.

[0136] In some embodiments, the processor 120 can obtain, based on historical information statistics, the total number of times that it is determined that no conflict has occurred based on the conflict judgment information, but the total number of times that a conflict has occurred based on the verification judgment information as the first number, and the total number of times judgments are made based on the verification judgment information based on historical information statistics as the second number, and the ratio between the first number and the second number is used as the misjudgment frequency.

[0137] The control condition refers to the condition used to determine whether the deployment of the audio monitoring device needs to be adjusted. In some embodiments, the control condition may include that the misjudgment frequency is greater than the misjudgment threshold, and the processing effect score is less than the score threshold. Among them, the misjudgment threshold can be determined by querying the preset table based on the partition type of the store partition where the audio monitoring device is currently located. The preset table includes misjudgment thresholds corresponding to different partition types. The preset table can be preset by a technician. For example, the misjudgment threshold of the cashier area in the preset table can be less than the misjudgment threshold of the dining area. The score threshold can be preset.

[0138] Device control parameters refer to parameters for adjusting the deployment of audio monitoring devices, such as parameters for adjusting the number of audio monitoring devices, parameters for adjusting the distance between audio monitoring devices, etc.

[0139] In some embodiments, for stores that meet the control conditions, the processor can determine the device control parameters based on the misjudgment frequency and the processing effect score. The higher the misjudgment frequency and the lower the processing effect score, the more audio monitoring devices are added and the shorter the distance between the audio monitoring devices.

[0140] In some embodiments of the present specification, by determining device control parameters based on misjudgment frequency and processing effect score to adjust the deployment of the audio monitoring device, the recognition sensitivity and accuracy of the smart store conflict identification system can be improved, and the timeliness and effect of conflict processing can be improved.

[0141] In some embodiments of the present specification, the second environmental data and the processing effect score are determined through the preprocessing layer of the feature acquisition model; and the environmental feature data are determined through the extraction layer of the feature acquisition model, which can improve the accuracy of the acquired environmental feature data and thereby improve the system's recognition sensitivity and accuracy for conflicts.

[0142] Figure 6 is an exemplary flow chart of determining collaborative monitoring parameters according to some embodiments of this specification. Figure 6 As shown, the process 600 includes the following steps. In some embodiments, the process 600 may be executed by the processor 120 of the system 100 .

[0143] In some embodiments, in response to a conflict in the target location, the processor 120 may execute the following steps 610 to 640 .

[0144] Step 610, determining the severity of the conflict based on the conflict judgment information and the conflict position characteristics.

[0145] The conflict location features refer to the features related to the location where the conflict occurs, such as the store division where the conflict location is located, the flow of people at the conflict location, the distribution of obstacles, etc.

[0146] In some embodiments, the processor 120 may determine the conflicting location characteristics in a variety of ways. For example, the processor 120 may compare the location information corresponding to the conflicting location with the location of the store partition to determine the store partition where the conflicting location is located. For another example, the processor may determine the conflicting location characteristics such as the flow of people, the distribution of obstacles, etc. based on the store environment characteristics. For more information about store environment characteristics, see Figure 4 and its related description.

[0147] The conflict severity is a parameter that characterizes the intensity of the conflict. In some embodiments, the conflict severity can be represented by a numerical value. For example, the conflict severity can be divided into 1 to 5 levels, and the higher the numerical value, the more intense and serious the conflict.

[0148] In some embodiments, the processor 120 can construct a conflict vector based on conflict judgment information and conflict location features; retrieve the historical conflict vector with the highest similarity in the second vector database as a reference conflict vector based on the conflict vector; and use the historical severity corresponding to the reference conflict vector as the current conflict severity.

[0149] The second vector database can be constructed based on historical data, and includes a large number of historical conflict vectors constructed based on historical data, as well as actual conflict severity in historical data. The actual conflict severity can be evaluated based on the damage of the conflict to surrounding objects and the impact on other guests.

[0150] Step 620, based on the severity of the conflict, determine the alarm level and collaborative monitoring parameters.

[0151] The alarm level is a numerical value that represents the intensity of the alarm. For example, the alarm level can be divided into levels one to five, the lower the level, the milder the prompt, and the higher the level, the more urgent the prompt.

[0152] The alarm level can be displayed on the clerk terminal through different alarm combinations. As an example, the first-level alarm is to send a slight vibration through the clerk terminal and send an alarm reminder through a soft and slow voice. The second-level alarm is to send an alarm reminder through vibration and a slightly faster voice, and play a "beep-beep" mechanical prompt sound at the same time. The third-level alarm can add a flashing prompt light on the basis of the second-level alarm.

[0153] In some embodiments, the processor 120 may select an alarm level according to the severity of the conflict, and the higher the severity, the higher the alarm level. For example, when the severity of the conflict is level 1, an alarm reminder is issued through the first alarm level; when the severity of the conflict is level 2, an alarm reminder is issued through the second alarm level.

[0154] Collaborative monitoring parameters refer to parameters for enhanced monitoring of conflicting locations. Such as monitoring enhancement parameters, etc. Monitoring enhancement parameters refer to parameters for enhancing the monitoring level of environmental monitoring devices. In some embodiments, monitoring enhancement parameters include at least one of audio enhancement parameters and image enhancement parameters. Audio enhancement parameters refer to parameters for increasing the monitoring level of audio monitoring equipment, such as enabling additional audio monitoring devices for collaborative monitoring. Image enhancement parameters refer to parameters for increasing the monitoring level of image acquisition devices, such as enabling additional image acquisition devices for collaborative monitoring, expanding the monitoring range of image acquisition devices (such as changing to a wide-angle lens for monitoring), etc.

[0155] In some embodiments, when the severity of the conflict is greater than a preset conflict threshold, the processor 120 may generate collaborative monitoring parameters, such as turning on other audio monitoring devices and image acquisition devices around the conflict location, or expanding the monitoring range of the image acquisition device.

[0156] In some embodiments, the processor 120 may also update the conflict severity based on the sound information characteristics and determine the updated conflict severity; and determine the alarm level and collaborative monitoring parameters based on the updated conflict severity.

[0157] In some embodiments, the environmental audio feature may include sound feature information. Sound feature information refers to feature information related to sound. In some embodiments, the sound feature information includes content features, intensity features, and sound abnormality features.

[0158] Content features refer to features related to the language content in the sound, such as the presence of words such as threats and insults. In some embodiments, the processor 120 can identify the content features in the sound through a language recognition model.

[0159] The intensity feature is a feature that characterizes the change in the intensity of the sound. For example, the volume of the sound exceeds a preset first volume threshold, or the volume of the sound is lower than a preset second volume threshold, etc., wherein the second volume threshold is lower than the first volume threshold. For another example, the pitch of the sound exceeds a preset first pitch threshold, or the pitch of the sound is lower than a preset second pitch threshold, etc., wherein the second pitch threshold is lower than the first pitch threshold.

[0160] The abnormal sound feature refers to a feature indicating the presence of abnormal sounds in the sound. For example, abnormal sounds such as glass breaking, objects falling, and screaming may occur. In some embodiments, the processor 120 may determine the abnormal sound feature by comparing the frequency in the sound with the frequencies of different abnormal sounds. For example, the abnormal sound feature may be determined by comparing the frequency in the sound with the frequency of the sound of glass breaking and objects falling.

[0161] In some embodiments, in response to at least one of the content feature, intensity feature, and abnormal sound feature in the sound feature information satisfying the conflict escalation condition, the processor 120 increases the conflict severity level to obtain an updated severity level. The conflict escalation condition may include at least one of the following: the presence of threatening or insulting words in the content feature, the volume of the sound in the intensity feature exceeding the first volume threshold (the tone of the sound exceeding the first tone threshold), and the presence of an abnormal sound in the abnormal sound feature. The more the conflict escalation conditions are met, the more the conflict severity is increased.

[0162] In some embodiments, the processor 120 may determine the alarm level and the collaborative monitoring parameters according to the aforementioned method based on the updated update severity.

[0163] In some embodiments of the present specification, by comprehensively analyzing the characteristics of sound information, the severity of the conflict can be assessed more accurately, thereby ensuring that the adjustment of alarm levels and monitoring parameters is more reasonable, so as to effectively reduce false alarms caused by misidentification and missed alarms caused by failure to detect escalation signals in a timely manner, and timely identify threats, insults and other abnormal situations, so that measures can be taken as early as possible to prevent conflicts from escalating into more serious incidents.

[0164] In some embodiments, the processor 120 may also determine a mediation indicator based on the severity of the conflict; determine a list of mediation clerks based on the mediation indicator, the conflict location, and the clerk information; and send additional mediation assistance instructions to the mediation clerks on the mediation clerk list.

[0165] Mediation indicators refer to the configuration indicators required to mediate conflicts. For example, the required clerk level, number of clerks, gender of clerks, etc. It is understandable that the higher the clerk level, the richer the mediation experience; the more clerks there are, the stronger the mediation efforts; and clerks of different genders play different roles in mediation. For example, male clerks can more effectively prevent conflicts, while female clerks' persuasion is more acceptable.

[0166] In some embodiments, the higher the severity of the conflict, the higher the level of the clerk, the more clerks, and the higher the proportion of male clerks required (conflicts can be forcibly prevented).

[0167] The mediation clerk list refers to a list of clerks who can mediate conflicts. The mediation clerk list includes at least one mediation clerk.

[0168] In some embodiments, the processor 120 may select the store clerk closest to the conflicting location as the mediation store clerk based on the store clerk information and the mediation index. For example, if the mediation index requires two ordinary male store clerks and one ordinary female store clerk, the processor 120 may select the two ordinary male store clerks and one ordinary female store clerk closest to the conflicting location as the mediation store clerks, thereby determining the mediation store clerk list. For more information about store clerk information, see Figure 1 , Figure 4 and its related description.

[0169] The assisting mediation instruction is a computer instruction to notify the store clerk to mediate the conflict. In some embodiments, the processor 120 may send the conflict location and conflict judgment information to each mediation clerk in the mediation clerk list so that the mediation clerk goes to the conflict location to mediate the conflict.

[0170] In some embodiments of this specification, the system can reasonably allocate appropriate mediators based on the severity of the conflict, thereby avoiding waste of resources; by accurately matching mediators, it avoids the situation where all store employees must participate in handling the conflict, thereby improving the efficiency of handling store conflicts.

[0171] In some embodiments, in response to the conflict trend being escalated and the closest distance between the mediation clerk and the conflict location being greater than a distance threshold, the alarm level is raised.

[0172] The conflict trend being escalation indicates that the severity of the conflict is gradually increasing. In some embodiments, the processor 120 may determine whether the conflict trend is escalation in a variety of ways. For example, the processor 120 may determine whether the conflict trend is escalation based on the conflict trend data in the conflict determination information. For details about the conflict trend data, see Figure 4 and related instructions.

[0173] The shortest distance between the mediation clerk and the conflicting position refers to the minimum value of the distance between the current position of at least one mediation clerk and the conflicting position.

[0174] It is understandable that when the conflict is escalating and the closest distance between the mediation clerk and the conflict location is greater than the distance threshold, it will take a long time for the mediation clerk to reach the conflict location. When the mediation clerk arrives, the severity of the conflict may be further increased. Therefore, it is necessary to increase the alarm level in advance to increase the intensity of reminding the clerk, prompting the mediation clerk to take action faster and mediate the conflict more effectively. The distance threshold is preset by the user according to actual needs.

[0175] In some embodiments of the present specification, dynamically adjusting the alarm level based on the real-time development of the conflict helps ensure that the conflict is taken seriously and handled promptly when it escalates, prompting relevant personnel to take quicker action; while ensuring that when the conflict escalates, there are sufficient personnel and resources to handle the conflict and protect the safety of customers and employees.

[0176] Step 630: Send the conflict judgment information to the clerk terminal and present it on the clerk terminal in an alarm manner corresponding to the alarm level. It can be understood that the conflict judgment information can be preferentially sent to one or more clerk terminals closest to the conflict location.

[0177] In some embodiments, the processor 120 can select the presentation method of the conflict judgment information according to the alarm level. For example, when the alarm level is level one, the processor 120 displays the conflict judgment information on the clerk terminal, and emits a slight vibration through the clerk terminal and uses a soft and slow voice to give an alarm reminder; when the alarm level is level two, the processor 120 displays the conflict judgment information on the clerk terminal, and vibrates and emits a slightly faster voice, and plays a "beep-beep" mechanical prompt sound for alarm reminders. When the alarm level is level three, the processor 120 displays the conflict judgment information on the clerk terminal, and vibrates and emits a slightly faster voice, and plays a "beep-beep" mechanical prompt sound, and flashes the prompt light at the same time.

[0178] Step 640: Send the collaborative monitoring parameters to the environmental monitoring device.

[0179] In some embodiments, the processor 120 may send the collaborative control parameters to the environment monitoring device, so that the environment monitoring device performs collaborative monitoring according to the collaborative control parameters. For the description of collaborative monitoring of the environment monitoring device, please refer to the above text.

[0180] In some embodiments of the present specification, the severity of the conflict is determined by conflict judgment information and conflict location characteristics, and then the alarm level is determined to remind the store staff and generate collaborative monitoring parameters for collaborative monitoring, which can improve the timeliness and efficiency of conflict handling, avoid further expansion of the conflict, and ensure the normal operation of the store.

[0181] Some embodiments of the present specification also provide an intelligent store conflict identification device, which includes a processor and at least one memory; the at least one memory stores computer instructions; the processor executes at least part of the computer instructions to implement any of the methods described in the above embodiments.

[0182] Some embodiments of the present specification also provide a computer-readable storage medium, which stores computer instructions. When a computer reads the computer instructions in the storage medium, the computer executes any of the methods in the above embodiments.

[0183] The basic concepts have been described above. Obviously, for those skilled in the art, the above detailed disclosure is only for example and does not constitute a limitation of this specification. Although not explicitly stated here, those skilled in the art may make various modifications, improvements and corrections to this specification. Such modifications, improvements and corrections are suggested in this specification, so such modifications, improvements and corrections still belong to the spirit and scope of the exemplary embodiments of this specification.

[0184] At the same time, this specification uses specific words to describe the embodiments of this specification. For example, "one embodiment", "an embodiment", and / or "some embodiments" refer to a certain feature, structure or characteristic related to at least one embodiment of this specification. Therefore, it should be emphasized and noted that "one embodiment" or "an embodiment" or "an alternative embodiment" mentioned twice or more in different positions in this specification does not necessarily refer to the same embodiment. In addition, certain features, structures or characteristics in one or more embodiments of this specification can be appropriately combined.

[0185] In addition, unless explicitly stated in the claims, the order of the processing elements and sequences described in this specification, the use of alphanumeric characters, or the use of other names are not intended to limit the order of the processes and methods of this specification. Although the above disclosure discusses some invention embodiments that are currently considered useful through various examples, it should be understood that such details are only for illustrative purposes, and the attached claims are not limited to the disclosed embodiments. On the contrary, the claims are intended to cover all modifications and equivalent combinations that are consistent with the essence and scope of the embodiments of this specification. For example, although the system components described above can be implemented by hardware devices, they can also be implemented only by software solutions, such as installing the described system on an existing server or mobile device.

[0186] Similarly, it should be noted that in order to simplify the description disclosed in this specification and thus help understand one or more embodiments of the invention, in the above description of the embodiments of this specification, multiple features are sometimes combined into one embodiment, figure or description thereof. However, this disclosure method does not mean that the features required by the subject matter of this specification are more than the features mentioned in the claims. In fact, the features of the embodiments are less than all the features of the single embodiment disclosed above.

[0187] In some embodiments, numbers describing the number of components and attributes are used. It should be understood that such numbers used in the description of the embodiments are modified by the modifiers "about", "approximately" or "substantially" in some examples. Unless otherwise specified, "about", "approximately" or "substantially" indicate that the numbers are allowed to vary by ±20%. Accordingly, in some embodiments, the numerical parameters used in the specification and claims are approximate values, which may change according to the required features of individual embodiments. In some embodiments, the numerical parameters should take into account the specified significant digits and adopt the general method of retaining digits. Although the numerical domains and parameters used to confirm the breadth of their range in some embodiments of this specification are approximate values, in specific embodiments, the setting of such numerical values ​​is as accurate as possible within the feasible range.

[0188] Each patent, patent application, patent application publication, and other materials, such as articles, books, specifications, publications, documents, etc., cited in this specification are hereby incorporated by reference in their entirety. Except for application history documents that are inconsistent with or conflicting with the contents of this specification, documents that limit the broadest scope of the claims of this specification (currently or later attached to this specification) are also excluded. It should be noted that if the descriptions, definitions, and / or use of terms in the materials attached to this specification are inconsistent or conflicting with the contents described in this specification, the descriptions, definitions, and / or use of terms in this specification shall prevail.

[0189] Finally, it should be understood that the embodiments described in this specification are only used to illustrate the principles of the embodiments of this specification. Other variations may also fall within the scope of this specification. Therefore, as an example and not a limitation, alternative configurations of the embodiments of this specification may be considered consistent with the teachings of this specification. Accordingly, the embodiments of this specification are not limited to the embodiments explicitly introduced and described in this specification.

Claims

1. An intelligent store conflict identification system, characterized in that: include: An environment collection module, wherein the environment collection module is configured to collect environment data through an environment monitoring device deployed in the store; The environment collection module includes an audio collection unit, and the audio collection unit is configured to collect first environment data through the environment monitoring device; as well as A processor, the processor being configured to: Determining conflict judgment information based on the first environment data, and judging whether a conflict occurs at a target position based on the conflict judgment information; In response to a conflict occurring at the target position, the target position is determined as a conflict position, and the conflict determination information is sent to a clerk terminal.

2. The system according to claim 1, characterized in that The determining, based on the first environment data, conflict judgment information, and judging, based on the conflict judgment information, whether a conflict occurs at a target position, includes: Preprocessing the first environmental data to obtain second environmental data; Extracting environmental audio features based on the second environmental data; In response to the environmental audio feature satisfying a preset judgment condition, determining the conflict judgment information based on the environmental audio feature and the store space feature; Based on the conflict judgment information, it is judged whether a conflict occurs at the target position.

3. The system according to claim 2, characterized in that The preprocessing of the first environmental data to obtain the second environmental data includes: processing the first environmental data and store environmental features based on a preprocessing layer of a feature acquisition model to obtain the second environmental data and a processing effect score; The extracting the environmental audio feature based on the second environmental data includes: processing the second environmental data and the processing effect score based on the extraction layer of the feature acquisition model to obtain the environmental audio feature; Wherein, the feature acquisition model is a machine learning model.

4. The system according to claim 1, characterized in that In response to a conflict occurring at the target location, the processor is further configured to: Determining the severity of the conflict based on the conflict judgment information and the conflict position characteristics; Determining an alert level and collaborative monitoring parameters based on the severity of the conflict; sending the conflict determination information to the clerk terminal and presenting it on the clerk terminal in an alarm manner corresponding to the alarm level; The collaborative monitoring parameters are sent to the environmental monitoring device.

5. A smart store conflict identification method, characterized in that: Applied to a smart store conflict identification system, the system comprises an environment collection module, the environment collection module is configured to collect environment data through an environment monitoring device deployed in the store; the environment collection module comprises an audio collection unit, the audio collection unit is configured to collect first environment data through the environment monitoring device; The method comprises: Determining conflict judgment information based on the first environment data, and judging whether a conflict occurs at a target position based on the conflict judgment information; In response to a conflict occurring at the target position, the target position is determined as a conflict position, and the conflict determination information is sent to a clerk terminal.

6. The method according to claim 5, characterized in that The determining, based on the first environment data, conflict judgment information, and judging, based on the conflict judgment information, whether a conflict occurs at a target position, includes: Preprocessing the first environmental data to obtain second environmental data; Extracting environmental audio features based on the second environmental data; In response to the environmental audio feature satisfying a preset judgment condition, determining the conflict judgment information based on the environmental audio feature and the store space feature; Based on the conflict judgment information, it is judged whether a conflict occurs at the target position.

7. The method according to claim 6, characterized in that The preprocessing of the first environmental data to obtain the second environmental data includes: processing the first environmental data and store environmental features based on a preprocessing layer of a feature acquisition model to obtain the second environmental data and a processing effect score; The extracting the environmental audio feature based on the second environmental data includes: processing the second environmental data and the processing effect score based on the extraction layer of the feature acquisition model to obtain the environmental audio feature; Wherein, the feature acquisition model is a machine learning model.

8. The method according to claim 5, characterized in that In response to a conflict occurring at the target location, the method further comprises: Determining the severity of the conflict based on the conflict judgment information and the conflict position characteristics; Determining an alert level and collaborative monitoring parameters based on the severity of the conflict; sending the conflict determination information to the clerk terminal and presenting it on the clerk terminal in an alarm manner corresponding to the alarm level; The collaborative monitoring parameters are sent to the environmental monitoring device.

9. An intelligent store conflict identification device, characterized in that: comprising a processor and at least one memory; the at least one memory storing computer instructions; The processor executes at least part of the computer instructions to implement the method according to any one of claims 5-8.

10. A computer-readable storage medium, characterized in that: The storage medium stores computer instructions. When a computer reads the computer instructions in the storage medium, the computer executes the method according to any one of claims 5 to 8.

Images