Artificial intelligence-based risk identification method and device, computer device and medium

Abstract

The application belongs to the technical field of artificial intelligence, and relates to a risk identification method and device based on artificial intelligence, a computer device and a storage medium, which comprises the following steps: acquiring geomagnetic data of a target environment in the process that a user conducts face-to-face signing through a mobile terminal; performing physical position verification on the geomagnetic data based on a geomagnetic reference library; if the geographical position verification is passed, collecting a sound signal of the target environment; performing environment verification on the sound signal based on an acoustic feature model; if the environment verification is passed, performing index extraction based on the geomagnetic data to obtain geomagnetic related indexes, and performing index extraction based on the sound signal to obtain acoustic related indexes; performing scoring processing on the geomagnetic related indexes and the acoustic related indexes to obtain an environment consistency score; and performing data analysis on the environment consistency score to generate a fraud risk identification result. The application can be applied to a fraud risk identification scene in the field of financial technology, and improves the identification reliability and accuracy of fraud behaviors in remote face-to-face signing.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology and can be applied to the financial technology field, particularly to risk identification methods, devices, computer equipment and storage media based on artificial intelligence. Background Technology

[0002] In traditional lending models, institutions rely primarily on limited remote verification methods to confirm a borrower's actual business location. However, existing remote verification methods have significant security flaws. Specifically, during video interviews, users can use green screens or high-definition posters to fabricate a business background, rendering environmental authenticity verification ineffective and making it difficult to identify fraudulent activities. This exposes lending operations to high security risks. This inaccurate verification method may lead borrowers to make false statements about their business operations, increasing the institution's default risk and financial losses.

[0003] For example, in financing-related credit guarantee insurance within the financial insurance sector, if the borrower's true place of business cannot be accurately verified, the insurance company may incorrectly underwrite policies when the borrower fabricates a business location to fraudulently obtain loans and insurance coverage. If the borrower encounters business problems or maliciously defaults, the insurance company will face substantial payouts, severely impacting its operational stability and market reputation.

[0004] Therefore, there is an urgent need to provide a reliable method for verifying fraudulent activities through remote face-to-face interviews in order to improve the security of credit business and related financial and insurance business. Summary of the Invention

[0005] The purpose of this application is to propose a risk identification method, device, computer equipment, and storage medium based on artificial intelligence, so as to solve the technical problem that existing remote verification methods for the true business location of borrowers are difficult to identify environmental fraud.

[0006] Firstly, an artificial intelligence-based risk identification method is provided, including: During the face-to-face verification process conducted by the user via mobile device, the geomagnetic data of the target environment in which the user is currently located is obtained; The geomagnetic data is physically located based on a pre-set geomagnetic reference library. If the geographical location verification is successful, then the sound signal of the target environment is collected; The sound signal is subjected to environmental verification based on a preset acoustic feature model; If the environmental verification is successful, the corresponding geomagnetic related indicators are extracted based on the geomagnetic data, and the corresponding acoustic related indicators are extracted based on the sound signal. The environmental consistency score is obtained by scoring the geomagnetic and acoustic related indicators. Data analysis is performed on the environmental consistency score to generate fraud risk identification results corresponding to the face-to-face signing.

[0007] Secondly, an artificial intelligence-based risk identification device is provided, including: The acquisition module is used to acquire geomagnetic data of the target environment where the user is currently located during the face-to-face verification process via a mobile device. The first verification module is used to verify the physical location of the geomagnetic data based on a preset geomagnetic reference library. The acquisition module is used to acquire sound signals from the target environment if the geographical location verification is successful. The second verification module is used to perform environmental verification on the sound signal based on a preset acoustic feature model; The extraction module is used to extract corresponding geomagnetic related indicators based on the geomagnetic data if the environmental verification is passed, and to extract corresponding acoustic related indicators based on the sound signal. The scoring module is used to perform scoring processing based on the geomagnetic related indicators and the acoustic related indicators to obtain the corresponding environmental consistency score. The generation module is used to perform data analysis on the environmental consistency score and generate fraud risk identification results corresponding to the face-to-face signing.

[0008] Thirdly, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described artificial intelligence-based risk identification method.

[0009] Fourthly, a computer-readable storage medium is provided, which stores a computer program that, when executed by a processor, implements the steps of the aforementioned risk identification method based on artificial intelligence.

[0010] In the aforementioned scheme implemented by the AI-based risk identification method, device, computer equipment, and storage medium, during the user's face-to-face verification process via a mobile terminal, geomagnetic data of the target environment in which the user is currently located is acquired; then, the geomagnetic data is physically verified based on a preset geomagnetic reference library; if the geographical location verification is successful, sound signals from the target environment are collected; subsequently, the sound signals are environmentally verified based on a preset acoustic feature model; if the environmental verification is successful, corresponding geomagnetic related indicators are extracted based on the geomagnetic data, and corresponding acoustic related indicators are extracted based on the sound signals; subsequently, scoring processing is performed based on the geomagnetic related indicators and the acoustic related indicators to obtain a corresponding environmental consistency score; finally, data analysis is performed on the environmental consistency score to generate a fraud risk identification result corresponding to the face-to-face verification. Based on the above automated processing flow, this application verifies the physical location of the user's current target environment by using geomagnetic data based on a geomagnetic reference library, and verifies the environment by using acoustic feature models based on sound signals. Subsequently, geomagnetic indicators are extracted from the geomagnetic data, and acoustic indicators are extracted from the sound signals. An environmental consistency score is obtained by scoring the geomagnetic and acoustic indicators, and then data analysis is performed on the environmental consistency score to generate fraud risk identification results corresponding to the face-to-face verification. Thus, this application performs preliminary environmental verification from two different dimensions: location and spatial characteristics. Based on this, it comprehensively analyzes multi-source environmental data containing geomagnetic and acoustic indicators to form a complete and rigorous remote face-to-face verification authenticity verification system. This system can effectively identify and prevent various forms of fraud, improve the reliability of fraud identification in remote face-to-face verification, and ensure the accuracy of the generated fraud risk identification results. Attached Figure Description

[0011] To more clearly illustrate the solutions in this application, the accompanying drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is an exemplary system architecture diagram to which this application can be applied; Figure 2 This is a flowchart of an embodiment of the AI-based risk identification method according to this application; Figure 3 This is a schematic diagram of a structure of an embodiment of the AI-based risk identification device according to this application; Figure 4 This is a schematic diagram of the structure of one embodiment of the computer device according to this application. Detailed Implementation

[0013] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein in the specification of the application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application; the terms "comprising" and "having," and any variations thereof, in the specification, claims, and foregoing drawings of this application, are intended to cover non-exclusive inclusion. The terms "first," "second," etc., in the specification, claims, or foregoing drawings of this application are used to distinguish different objects, not to describe a particular order.

[0014] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0015] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0016] like Figure 1 As shown, system architecture 100 may include terminal device 101, network 102, and server 103. Terminal device 101 may be a laptop 1011, tablet 1012, or mobile phone 1013. Network 102 is used as a medium to provide a communication link between terminal device 101 and server 103. Network 102 may include various connection types, such as wired, wireless communication links, or fiber optic cables, etc.

[0017] Users can use terminal device 101 to interact with server 103 via network 102 to receive or send messages, etc. Various communication client applications can be installed on terminal device 101, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social media platform software, etc.

[0018] Terminal device 101 can be various electronic devices with a display screen and support web browsing. In addition to laptops 1011, tablets 1012, or mobile phones 1013, terminal device 101 can also be an e-book reader, an MP3 player (Moving Picture Experts Group Audio Layer III), an MP4 player (Moving Picture Experts Group Audio Layer IV), a laptop computer, and a desktop computer, etc.

[0019] Server 103 can be a server that provides various services, such as a backend server that provides support for the pages displayed on terminal device 101.

[0020] It should be noted that the AI-based risk identification method provided in this application is generally executed by a server / terminal device, and correspondingly, the AI-based risk identification device is generally installed in the server / terminal device.

[0021] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0022] Continue to refer to Figure 2 This document illustrates a flowchart of an embodiment of the AI-based risk identification method according to this application. The order of steps in the flowchart can be changed, and some steps can be omitted, depending on different requirements. The AI-based risk identification method provided in this application can be applied to any scenario requiring fraud risk identification, and thus can be applied to products in these scenarios, such as fraud risk identification products in the financial insurance field. The AI-based risk identification method includes the following steps: Step S201: During the user's face-to-face verification process via mobile device, obtain the geomagnetic data of the target environment where the user is currently located.

[0023] In this embodiment, the AI-based risk identification method operates on electronic devices (e.g., Figure 1The server / terminal device shown can acquire geomagnetic data of the user's current target environment via wired or wireless connection. It should be noted that the aforementioned wireless connection methods may include, but are not limited to, 3G / 4G / 5G connections, WiFi connections, Bluetooth connections, WiMAX connections, Zigbee connections, UWB (ultra-wideband) connections, and other currently known or future wireless connection methods. The implementing entity of this application is specifically a risk identification system, which can be simply referred to as the system.

[0024] The specific implementation process of obtaining the geomagnetic data of the target environment where the user is currently located will be described in more detail in subsequent specific embodiments of this application, and will not be elaborated on here.

[0025] Step S202: Verify the physical location of the geomagnetic data based on a preset geomagnetic reference library.

[0026] In this embodiment, the aforementioned geomagnetic reference database includes a geomagnetic fingerprint database or a reference database for the same area. The implementation process of the aforementioned physical location verification includes: performing a comprehensive and detailed comparison of the geomagnetic data with the geomagnetic fingerprint database or the reference database for the same area used in the user's historical application. If the GPS shows that the user's current location is in an office building, but the comparison reveals that the geomagnetic fingerprint exhibits extremely regular "open area" characteristics, this is clearly inconsistent with the complex and unique geomagnetic fingerprint characteristics that an office building should have due to the presence of a large amount of steel reinforcement and numerous electrical appliances. In this case, the system will determine that it is a software simulation or a location spoofing based on this obvious difference. Because under normal circumstances, the geomagnetic field inside an office building will exhibit complex characteristics due to interference from steel reinforcement and electrical appliances, while the geomagnetic characteristics of an "open area" are significantly different. This geomagnetic fingerprint data that does not conform to reality is likely generated by software simulation or spoofing of the location, thus determining that the location verification has failed. Through this logical judgment, the system can promptly identify potential fraud risks and ensure the security and authenticity of the face-to-face interview process.

[0027] Step S203: If the geographical location verification is successful, then collect the sound signal of the target environment.

[0028] In this embodiment, in the scenario of remote face-to-face interviews, the system flexibly employs two methods to acquire sound as a sound source. On the one hand, it cleverly utilizes the environmental noise naturally generated during the interview, such as the conversation between the interviewee and the interviewee, which naturally and authentically exists in the indoor space. On the other hand, the system also proactively emits "prompt tones" of specific frequencies. These prompt tones are carefully designed and selected, possessing specific frequency characteristics to better meet the needs of subsequent analysis. The microphone installed on the device continuously collects information on the propagation of sound within the room. It can not only accurately record the direct sound propagating directly from the sound source to the microphone—this sound propagation path is the shortest, with relatively little energy loss—but also keenly capture reflected waves after being reflected by indoor walls, furniture, and other objects. These reflected waves, having undergone different propagation paths and reflection processes, carry rich information about the indoor space. By comprehensively acquiring information on the propagation of sound within the room, a solid foundation is laid for subsequent analysis of the characteristics of the indoor space.

[0029] Step S204: Perform environmental verification on the sound signal based on a preset acoustic feature model.

[0030] In this embodiment, the specific implementation process of environmental verification of the sound signal based on the preset acoustic feature model will be further described in detail in subsequent specific embodiments of this application, and will not be elaborated on here.

[0031] Step S205: If the environmental verification is successful, then the corresponding geomagnetic related indicators are extracted based on the geomagnetic data, and the corresponding acoustic related indicators are extracted based on the sound signal.

[0032] In this embodiment, the aforementioned geomagnetic related indicators include geomagnetic wave frequency sequences and geomagnetic fingerprints. Geomagnetic data can be collected using devices such as magnetometers, and the time-domain geomagnetic signal can be converted into a frequency-domain signal using signal processing techniques (such as Fourier transform) to obtain the geomagnetic wave frequency sequence. This sequence reflects the frequency characteristics of the geomagnetic field changing over time. Different physical spaces will exhibit different geomagnetic wave frequencies due to variations in their internal structures (such as the distribution of reinforcing steel bars, electrical equipment, etc.). In addition to frequency characteristics, comprehensive information such as the intensity and direction of the geomagnetic field can also constitute a geomagnetic fingerprint. It is a unique identifier of the geomagnetic environment at a specific location, similar to a human fingerprint, and is unique. By performing multi-dimensional analysis and processing of geomagnetic data, a geomagnetic fingerprint that represents the geomagnetic characteristics of that environment can be extracted.

[0033] The aforementioned acoustic-related indicators include acoustic reflection frequency sequences and indoor impulse response indicators. The acoustic reflection frequency sequence is obtained by collecting sound signals from the environment using devices such as microphones and analyzing the frequency characteristics of sound reflected back from obstacles during propagation. Different spatial structures (such as room size, shape, and the placement of objects inside) will have different effects on sound reflection, resulting in different acoustic reflection frequency sequences. Additionally, a known acoustic impulse signal is emitted into the environment, and the received response signal is recorded. By analyzing the amplitude, phase, delay, and other characteristics of the response signal, the indoor impulse response indicator, or acoustic impulse response indicator, can be obtained. These indicators reflect the propagation and reflection characteristics of sound in the environment, further describing the acoustic characteristics of the environment.

[0034] Step S206: Based on the geomagnetic correlation index and the acoustic correlation index, a scoring process is performed to obtain the corresponding environmental consistency score.

[0035] In this embodiment, the specific implementation process of scoring based on the geomagnetic and acoustic related indicators to obtain the corresponding environmental consistency score will be further described in detail in subsequent specific embodiments of this application, and will not be elaborated on here.

[0036] Step S207: Perform data analysis on the environmental consistency score to generate fraud risk identification results corresponding to the face-to-face signing.

[0037] In this embodiment, the specific implementation process of performing data analysis on the environmental consistency score to generate fraud risk identification results corresponding to the face-to-face signing will be further described in detail in subsequent specific embodiments of this application, and will not be elaborated on here.

[0038] This application acquires geomagnetic data of the target environment where the user is currently located during a face-to-face interview via a mobile device; then, it verifies the physical location of the geomagnetic data based on a preset geomagnetic benchmark library; if the geographical location verification is successful, it collects sound signals from the target environment; subsequently, it verifies the environment of the sound signals based on a preset acoustic feature model; if the environment verification is successful, it extracts corresponding geomagnetic related indicators based on the geomagnetic data and corresponding acoustic related indicators based on the sound signals; subsequently, it performs scoring processing based on the geomagnetic related indicators and the acoustic related indicators to obtain a corresponding environmental consistency score; finally, it performs data analysis on the environmental consistency score to generate a fraud risk identification result corresponding to the face-to-face interview. Based on the above automated processing flow, this application verifies the physical location of the user's current target environment by using geomagnetic data based on a geomagnetic reference library, and verifies the environment by using acoustic feature models based on sound signals. Subsequently, geomagnetic indicators are extracted from the geomagnetic data, and acoustic indicators are extracted from the sound signals. An environmental consistency score is obtained by scoring the geomagnetic and acoustic indicators, and then data analysis is performed on the environmental consistency score to generate fraud risk identification results corresponding to the face-to-face verification. Thus, this application performs preliminary environmental verification from two different dimensions: location and spatial characteristics. Based on this, it comprehensively analyzes multi-source environmental data containing geomagnetic and acoustic indicators to form a complete and rigorous remote face-to-face verification authenticity verification system. This system can effectively identify and prevent various forms of fraud, improve the reliability of fraud identification in remote face-to-face verification, and ensure the accuracy of the generated fraud risk identification results.

[0039] In some optional implementations, the step S201 of obtaining the geomagnetic data of the target environment where the user is currently located includes the following steps: Obtain a geomagnetic profile sequence corresponding to the target environment based on a preset magnetometer.

[0040] In this embodiment, the moment the face-to-face interview process officially begins, the system silently activates the magnetometer built into the user's mobile device (such as a smartphone) in the background. As a high-precision sensor specifically designed to measure magnetic field strength, the magnetometer continuously senses changes in the geomagnetic field strength in the surrounding environment. This sensed data is gradually aggregated chronologically, ultimately forming a complete geomagnetic profile sequence M. The entire data acquisition process occurs entirely in the background; the user is completely unaware of this operation and will not be distracted by any additional operational requirements. This ensures that the user can fully concentrate on the normal interview communication without any interference with the smoothness and naturalness of the interview process.

[0041] Based on a preset algorithm model, feature extraction is performed on the geomagnetic profile sequence to obtain the corresponding geomagnetic profile sequence features.

[0042] In this embodiment, once the system successfully acquires the collected geomagnetic profile sequence, it employs a specific algorithm model designed for the characteristics of geomagnetic data to conduct a comprehensive and in-depth analysis. Inside a building, the crisscrossing steel reinforcement structure acts like a massive and complex magnetic framework, exerting a specific influence on the surrounding geomagnetic field. Simultaneously, the current flowing within large electrical appliances generates electromagnetic fields, which interact with the surrounding geomagnetic field, creating unique disturbances to the local geomagnetic field. These disturbances, generated by the steel reinforcement structure and large electrical appliances, are like unique fingerprints—a "geomagnetic fingerprint" specific to the indoor environment, possessing extremely high uniqueness. The algorithm model can accurately identify these unique disturbance features through meticulous analysis and comparison of the data in the geomagnetic profile sequence. These extracted disturbance features (i.e., geomagnetic profile sequence features) are carefully stored by the system as identification information of the indoor environment's geomagnetic fingerprint, providing crucial and indispensable evidence for subsequent physical location verification.

[0043] Obtain the magnetic field change gradient information calculated based on the magnetometer.

[0044] In this embodiment, to ensure that the collected geomagnetic data more accurately reflects the user's actual location, the system guides the user to make slight movements of the phone using clear and explicit prompts. For example, the system prompts the user to slowly circle around and photograph the surrounding environment. During this process, the user does not need to make large movements; simply following the prompts to move the phone slightly is sufficient. While the user moves the phone, the magnetometer remains active, continuously collecting geomagnetic data. Based on this newly collected geomagnetic data, the system uses mathematical methods to calculate the magnetic field gradient. Magnetic field gradient. The calculation formula is: ,in, , , These represent the rates of change of geomagnetic intensity in the x, y, and z directions, respectively. This is analogous to measuring the rate of change of the geomagnetic field in three-dimensional space along three different directions, reflecting the dynamic changes of the geomagnetic field in different spatial directions. i, j, and k are unit vectors in the three directions, defining the direction of the rates of change in these directions, much like determining north, south, east, and west on a map. By calculating the gradient of magnetic field changes, the system can describe the dynamic changes of the geomagnetic field during mobile phone movement in a more precise and detailed way, like creating a detailed and dynamic map of the geomagnetic field's changes, providing richer data support for subsequent logical judgments.

[0045] The geomagnetic profile sequence features and the magnetic field change gradient information are integrated to obtain corresponding integrated data.

[0046] In this embodiment, the geomagnetic profile sequence features and magnetic field change gradient information are integrated and processed, and the resulting integrated data is used as the corresponding geomagnetic data.

[0047] The integrated data is used as the geomagnetic data.

[0048] This application acquires a geomagnetic profile sequence corresponding to the target environment based on a preset magnetometer; then, it extracts features from the geomagnetic profile sequence using a preset algorithm model to obtain corresponding geomagnetic profile sequence features; subsequently, it acquires magnetic field change gradient information calculated based on the magnetometer; next, it integrates the geomagnetic profile sequence features and the magnetic field change gradient information to obtain corresponding integrated data; finally, it uses the integrated data as the geomagnetic data. Based on the above processing flow, this application, by collecting and analyzing geomagnetic data, extracts unique indoor geomagnetic fingerprints and combines this with the calculation of magnetic field change gradients during mobile phone movement. This facilitates comparison of the acquired geomagnetic data with a geomagnetic reference database, thereby determining the authenticity of the user's location, effectively preventing fraudulent activities through software-simulated location, and providing a physical basis for the authenticity verification of remote face-to-face verification.

[0049] In some optional implementations of this embodiment, step S204 includes the following steps: The corresponding indoor impulse response index is obtained by performing cepstral analysis on the sound signal.

[0050] In this embodiment, after acquiring the collected sound signal, the system uses cepstral analysis (CDE) and related signal processing techniques for in-depth analysis. The CDE process is complex. First, it performs a logarithmic transformation on the sound signal's spectrum. The spectrum is the frequency domain representation of the sound signal, showing the distribution of different frequency components. Performing a logarithmic transformation on the spectrum is like re-encoding the frequency components of the sound; this transformation highlights some important features of the signal, making subsequent analysis more accurate and effective. After the logarithmic transformation, the system performs an inverse Fourier transform. The Fourier transform is an important tool for converting signals from the time domain to the frequency domain, while the inverse Fourier transform is its reverse process, converting the signal back from the frequency domain to the time domain. Through this series of complex operations, the system can extract the room impulse response (IRT) index from the sound signal. The IRT index accurately reflects the time history and energy changes of sound as it travels from the sound source, undergoes multiple reflections indoors, and reaches the microphone. For example, it can record the time points and energy intensity of each stage of sound reflection, such as the first and second reflections, providing crucial data support for accurately determining the characteristics of indoor spaces.

[0051] Call the pre-established acoustic feature model.

[0052] In this embodiment, different physical spaces possess unique acoustic characteristics due to variations in their structure, size, and internal objects. Taking several common spaces as examples, open warehouse spaces are typically large, resulting in fewer sound reflections during propagation and allowing energy to travel further, thus exhibiting a longer reverberation time (RT60). Reverberation time (RT60) refers to the time required for sound intensity to decay to one millionth of its original intensity (i.e., a 60dB decrease), serving as a crucial indicator of the duration of sound reverberation within an indoor space. In contrast, narrow office cubicles are relatively small, causing frequent reflections of sound from walls, partitions, and other objects during propagation, leading to faster energy loss and a shorter reverberation time. The acoustic characteristics of residential buildings differ from the previous two types, being influenced by various factors such as room layout and interior materials.

[0053] The system pre-establishes acoustic characteristic models for various physical spaces. These models are based on a large number of actual measurements and data analysis and can accurately describe the acoustic characteristics of different spaces.

[0054] The indoor impulse response index is compared with the acoustic feature model to obtain the corresponding comparison results.

[0055] In this embodiment, after acquiring the indoor impulse response indicators of the current environment, the system compares these indicators with a pre-established acoustic feature model to match the actually measured "acoustic features" with a standard "acoustic template." Through comparative analysis, the system can determine whether the current environment matches the environment claimed by the user. For example, if a user claims to be in a factory workshop, but the acoustic fingerprint displays features consistent with a small, enclosed space, the system can accurately identify this semantic conflict in the environment, generating a comparison result indicating a failed comparison. This helps detect potential fraud and ensures the authenticity and security of remote face-to-face verification.

[0056] If the comparison result is a pass, then the environmental verification is deemed successful.

[0057] In this embodiment, the comparison result may include "comparison passed" or "comparison failed". If the comparison result is detected as "comparison passed", the environmental verification is determined to be passed; otherwise, the environmental verification is determined to be failed.

[0058] If the comparison result is that the comparison fails, then the environmental verification is deemed to have failed.

[0059] This application obtains the corresponding indoor impulse response index by performing cepstral analysis on the sound signal; then, it calls a pre-established acoustic feature model; subsequently, it compares the indoor impulse response index with the acoustic feature model to obtain the corresponding comparison result; if the comparison result is successful, the environmental verification is deemed successful; otherwise, if the comparison result is unsuccessful, the environmental verification is deemed unsuccessful. Based on the above processing flow, this application, by collecting sound signals, extracting indoor impulse response indices using cepstral analysis, and then comparing these indices with a pre-established acoustic feature model, can accurately determine the characteristics of indoor spaces. It can effectively identify whether the environment claimed by the user matches the actual environment, thus accurately completing environmental verification and preventing fraudulent activities through falsified background environments, further enhancing the reliability of remote face-to-face verification.

[0060] In some alternative implementations, step S206 includes the following steps: The cross-correlation coefficients are calculated based on the geomagnetic correlation index and the acoustic correlation index.

[0061] In this embodiment, the specific implementation process of calculating the corresponding cross-correlation coefficients based on the geomagnetic correlation index and the acoustic correlation index will be further described in detail in subsequent specific embodiments of this application, and will not be elaborated on here.

[0062] The geomagnetic correlation index, the acoustic correlation index, and the cross-correlation coefficient are scored based on a preset scoring strategy to obtain the corresponding scoring results.

[0063] In this embodiment, the specific implementation process of scoring the geomagnetic correlation index, the acoustic correlation index, and the cross-correlation coefficient based on the preset scoring strategy to obtain the corresponding scoring results will be further described in detail in subsequent specific embodiments of this application, and will not be elaborated on here.

[0064] The scoring result is used as the environmental consistency score.

[0065] This application calculates the corresponding cross-correlation coefficients based on the geomagnetic and acoustic correlation indicators; then, it scores the geomagnetic, acoustic, and cross-correlation indicators using a preset scoring strategy to obtain corresponding scoring results; subsequently, these scoring results are used as the environmental consistency score. Based on this processing flow, this application calculates the corresponding cross-correlation coefficients based on the geomagnetic and acoustic correlation indicators, then scores the geomagnetic, acoustic, and cross-correlation indicators using a scoring strategy, and uses the obtained scoring results as the corresponding environmental consistency score. This allows for efficient and accurate calculation of the environmental consistency score, ensuring its accuracy.

[0066] In some optional implementations, the geomagnetic correlation index includes at least the geomagnetic fluctuation frequency, and the acoustic correlation index includes at least the acoustic reflection frequency; the calculation of the corresponding cross-correlation coefficient based on the geomagnetic correlation index and the acoustic correlation index includes the following steps: The geomagnetic fluctuation frequency is obtained from the geomagnetic related indicators, and the acoustic reflection frequency is obtained from the acoustic related indicators.

[0067] In this embodiment, the aforementioned geomagnetic related indicators include: Geomagnetic wave frequency sequence: Geomagnetic data is collected using devices such as magnetometers, and the time-domain geomagnetic signal is converted into a frequency-domain signal using signal processing techniques (such as Fourier transform) to obtain the geomagnetic wave frequency sequence. This sequence reflects the frequency characteristics of the geomagnetic field changing over time. Different physical spaces will exhibit different geomagnetic wave frequencies due to differences in their internal structures (such as the distribution of reinforcing steel bars, electrical equipment, etc.). Geomagnetic fingerprint: In addition to frequency characteristics, the intensity, direction, and other comprehensive information of the geomagnetic field can also constitute a geomagnetic fingerprint. It is a unique identifier of the geomagnetic environment at a specific location, similar to a human fingerprint, and is unique. Through multi-dimensional analysis and processing of geomagnetic data, a geomagnetic fingerprint that represents the geomagnetic characteristics of that environment is extracted.

[0068] The aforementioned acoustic-related indicators include: Acoustic reflection frequency sequence: This involves collecting sound signals from the environment using devices such as microphones, analyzing the frequency characteristics of sound reflected back from obstacles during propagation, and obtaining the acoustic reflection frequency sequence. Different spatial structures (such as room size, shape, and the placement of objects inside) will have different effects on sound reflection, resulting in different acoustic reflection frequency sequences. Indoor impulse response index (or acoustic impulse response index): A known acoustic impulse signal is emitted into the environment, and the received response signal is recorded. By analyzing the amplitude, phase, delay, and other characteristics of the response signal, the acoustic impulse response index can be obtained. These indicators can reflect the propagation and reflection characteristics of sound in the environment, further describing the acoustic characteristics of the environment.

[0069] Call the preset cross-correlation coefficient formula.

[0070] In this embodiment, in a remote face-to-face verification scenario, the system performs cross-indicator correlation calculations to assess the consistency of environmental data, focusing on verifying the correlation between geomagnetic wave frequency and acoustic reflection frequency. Specifically, the system first collects geomagnetic wave frequency sequences and acoustic reflection frequency sequences, which record the frequencies of geomagnetic waves and acoustic reflections at different times. Next, the system calculates the covariance Cov(X,Y) of these two sequences. Covariance is a statistic used to measure the degree of linear correlation between two variables; it reflects the trend of the co-change of the geomagnetic wave frequency sequence X and the acoustic reflection frequency sequence Y. If both sequences show a significant increasing or decreasing trend, the covariance value will be large; conversely, if their trends are not significantly correlated, the covariance value will be small. Simultaneously, the system also calculates the standard deviations of sequences X and Y separately. and Standard deviation is a measure of the dispersion of a set of data; it reflects how dispersed the data points are relative to the mean. A larger standard deviation indicates more dispersed data; a smaller standard deviation indicates more concentrated data.

[0071] After obtaining the covariance and the standard deviations of the two sequences, the system will apply the cross-correlation coefficient formula. The system calculates the cross-correlation coefficient between geomagnetic wave frequencies and acoustic reflection frequencies. This cross-correlation coefficient ranges from -1 to 1. A coefficient close to 1 indicates a high positive correlation between geomagnetic waves and acoustic reflections, meaning that an increase in one variable is likely to increase the other. A coefficient close to -1 indicates a high negative correlation, meaning an increase in one variable is likely to decrease the other. A coefficient close to 0 indicates a weak correlation between geomagnetic waves and acoustic reflections, with no significant correlation between their changes. In this way, the system can accurately measure the degree of correlation between geomagnetic waves and acoustic reflections, thereby determining the consistency of environmental data and providing crucial information for subsequent comprehensive assessments.

[0072] The geomagnetic wave frequency and the acoustic reflection frequency are calculated based on the cross-correlation coefficient formula to obtain the corresponding first calculation result.

[0073] In this embodiment, the geomagnetic wave frequency and acoustic reflection frequency can be calculated based on the above cross-correlation coefficient formula, and the first calculation result obtained can be used as the corresponding cross-correlation coefficient.

[0074] The first calculation result is used as the cross-correlation coefficient.

[0075] This application obtains the geomagnetic wave frequency from the geomagnetic related indicators and the acoustic reflection frequency from the acoustic related indicators; then it calls a preset cross-correlation coefficient formula; subsequently, it calculates the geomagnetic wave frequency and the acoustic reflection frequency based on the cross-correlation coefficient formula to obtain a corresponding first calculation result; and then uses the first calculation result as the cross-correlation coefficient. Based on the above processing flow, this application calculates the geomagnetic wave frequency and the acoustic reflection frequency using the cross-correlation coefficient formula, and uses the obtained first calculation result as the corresponding cross-correlation coefficient. This can accurately measure the degree of correlation between geomagnetic waves and acoustic reflections, thereby judging the consistency of environmental data and providing key basis for subsequent comprehensive assessment.

[0076] In some optional implementations of this embodiment, the scoring process of the geomagnetic correlation index, the acoustic correlation index, and the cross-correlation coefficient based on a preset scoring strategy to obtain the corresponding scoring results includes the following steps: Weighted data corresponding to the geomagnetic correlation index, the acoustic correlation index, and the cross-correlation coefficient are generated based on a preset weight allocation strategy.

[0077] In this embodiment, the weight allocation strategy includes assigning corresponding weights to each indicator based on the degree of influence of different indicators on environmental consistency. For example, if geomagnetic fingerprinting has higher accuracy and stability in distinguishing different environments, then it can be assigned a larger weight; while the weight of the acoustic impulse response indicator may be relatively smaller. The weight allocation can be determined through expert experience, data analysis, or machine learning algorithms.

[0078] Furthermore, based on the above weight allocation strategy, weight data corresponding to the above geomagnetic related indicators, acoustic related indicators, and cross-correlation coefficients can be generated.

[0079] Call the preset scoring model.

[0080] In this embodiment, the construction process of the above-mentioned scoring model includes: constructing an environmental consistency scoring model using methods such as weighted average, fuzzy comprehensive evaluation, and neural network models. Taking the weighted average method as an example, assuming that n environmental feature indicators I1, I2, ... In, the corresponding weights are w1 and w2. If wn, then the formula for calculating the Environmental Consistency Score (ECS) is: The scoring model combines the values ​​of various indicators to calculate the final environmental consistency score.

[0081] Based on the scoring model, the geomagnetic correlation index, the acoustic correlation index, the cross-correlation coefficient, and the weight data are calculated and processed to obtain the corresponding second calculation result.

[0082] In this embodiment, the above-mentioned geomagnetic related indicators, acoustic related indicators, cross-correlation coefficients and weight data can be input into the above-mentioned scoring model for calculation and processing, and the second calculation result obtained can be used as the corresponding scoring result.

[0083] The second calculation result is used as the scoring result.

[0084] In this embodiment, historical environmental data and known cases of fraud or normal behavior can be further used to calibrate the calculated environmental consistency score (score result). By analyzing the distribution of fraud and normal cases within different score intervals, the parameters or weights of the scoring model can be adjusted to make the score more accurately reflect the true state of the environment.

[0085] Furthermore, in practical applications, the system continuously collects new environmental data and user feedback. Based on this real-time data, the environmental consistency scoring model is continuously optimized and updated to improve the accuracy and reliability of the scoring. For example, if it is found that the environmental characteristics of a certain area have changed, rendering the original scoring model inapplicable, the model needs to be adjusted in a timely manner to adapt to the new environmental conditions.

[0086] This application generates weight data corresponding to the geomagnetic related indicators, acoustic related indicators, and cross-correlation coefficients based on a preset weight allocation strategy; then, it calls a preset scoring model; subsequently, it calculates and processes the geomagnetic related indicators, acoustic related indicators, cross-correlation coefficients, and weight data based on the scoring model to obtain a corresponding second calculation result; and finally, it uses the second calculation result as the scoring result. Based on the above processing flow, this application generates weight data corresponding to the geomagnetic related indicators, acoustic related indicators, and cross-correlation coefficients based on a weight allocation strategy, and then calculates and processes the geomagnetic related indicators, acoustic related indicators, cross-correlation coefficients, and weight data based on the use of a scoring model, and uses the calculated second calculation result as the corresponding scoring result, thereby automatically and accurately completing the scoring processing of geomagnetic related indicators, acoustic related indicators, and cross-correlation coefficients, ensuring the accuracy of the obtained scoring results.

[0087] In some optional implementations of this embodiment, step S207 includes the following steps: Obtain the preset risk threshold.

[0088] In this embodiment, the selection of the above-mentioned risk threshold value is not specifically limited, and can be set according to actual business needs.

[0089] Determine whether the environmental consistency score is greater than or equal to the risk threshold.

[0090] In this embodiment, a numerical comparison result can be obtained by comparing the calculated environmental consistency score with a risk threshold. The numerical comparison result includes whether the environmental consistency score is greater than or equal to the risk threshold, or whether the environmental consistency score is less than the risk threshold.

[0091] If the environmental consistency score is greater than or equal to the risk threshold, a first fraud identification result is generated indicating that there is no fraud risk in the face-to-face interview.

[0092] In this embodiment, if the environmental consistency score is detected to be greater than or equal to the aforementioned risk threshold, it is determined that there is no environmental fraud risk, and a first fraud identification result indicating that the face-to-face signing does not have fraud risk is generated.

[0093] If the environmental consistency score is less than the risk threshold, a second fraud identification result is generated indicating that the face-to-face interview has a fraud risk.

[0094] In this embodiment, if the environmental consistency score is detected to be less than the above-mentioned risk threshold, it is determined that there is an environmental fraud risk, and then a second fraud identification result is generated indicating that the face-to-face signing has a fraud risk.

[0095] This application obtains a preset risk threshold; then determines whether the environmental consistency score is greater than or equal to the risk threshold; if the environmental consistency score is greater than or equal to the risk threshold, a first fraud identification result is generated indicating that the face-to-face interview does not pose a fraud risk; and if the environmental consistency score is less than the risk threshold, a second fraud identification result is generated indicating that the face-to-face interview poses a fraud risk. Based on the above processing flow, this application automatically and accurately generates corresponding fraud identification results by comparing the calculated environmental consistency score with the risk threshold, thereby improving the accuracy of fraud identification result generation.

[0096] In some optional implementations, the system also has a blacklist clustering analysis function, the specific implementation process of which includes: After calculating the Environmental Consistency Score (ECS) fingerprint, the system records these ECS fingerprints into an "Environment Blacklist Database." This database acts like an archive specifically for recording suspicious environments, storing various environmental information that may pose a fraud risk. Next, the system performs cluster analysis on the ECS fingerprints of different accounts. During this clustering analysis, the system employs appropriate clustering algorithms, such as distance-based clustering. The core idea of ​​this algorithm is to determine whether ECS fingerprints belong to the same cluster based on their similarity. Specifically, the algorithm calculates the distance between different ECS fingerprints; the closer the distance, the more similar the two fingerprints; the farther the distance, the lower the similarity.

[0097] Then, the clustering algorithm groups similar ECS fingerprints together based on this distance information. Through this clustering operation, the system can clearly see the distribution of ECS fingerprints among different accounts. If the clustering results show a high degree of overlap in the ECS fingerprints of different accounts—meaning multiple accounts' ECS fingerprints are grouped into the same category—it indicates that these loan applications are likely originating from the same physical location, potentially a fraud ring. Once the system detects this situation, it immediately triggers the cluster risk control early warning mechanism, taking timely measures such as further investigation and verification, and restricting the operation of related accounts to prevent large-scale fraud and ensure the safe and stable operation of the remote interview system.

[0098] In some alternative implementations, the user information obtained is subject to user consent and complies with relevant laws and policies.

[0099] Furthermore, any software tools or components not belonging to our company that appear in the embodiments of this application are merely illustrative examples and do not represent actual use.

[0100] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0101] It should be emphasized that, in order to further ensure the privacy and security of the above fraud risk identification results, the above fraud risk identification results can also be stored in a blockchain node.

[0102] The blockchain referred to in this application is a novel application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanisms, and encryption algorithms. Essentially, a blockchain is a decentralized database, a chain of data blocks linked together using cryptographic methods. Each data block contains information about a batch of network transactions, used to verify the validity of the information (anti-counterfeiting) and generate the next block. A blockchain can include an underlying blockchain platform, a platform product service layer, and an application service layer.

[0103] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) refers to the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results. Foundational technologies for artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0104] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing related hardware through computer-readable instructions. These computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. The aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, optical disk, or read-only memory (ROM), or random access memory (RAM).

[0105] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0106] Further reference Figure 3 As a response to the above Figure 2 To implement the method shown, this application provides an embodiment of an artificial intelligence-based risk identification device, which is similar to... Figure 2 Corresponding to the method embodiments shown, this device can be specifically applied to various electronic devices.

[0107] like Figure 3 As shown, the AI-based risk identification device 300 described in this embodiment includes: an acquisition module 301, a first verification module 302, a collection module 303, a second verification module 304, an extraction module 305, a scoring module 306, and a generation module 307. Wherein: The acquisition module 301 is used to acquire geomagnetic data of the target environment where the user is currently located during the face-to-face signing process via a mobile terminal. The first verification module 302 is used to verify the physical location of the geomagnetic data based on a preset geomagnetic reference library. The acquisition module 303 is used to acquire sound signals from the target environment if the geographical location verification is successful. The second verification module 304 is used to perform environmental verification on the sound signal based on a preset acoustic feature model; The extraction module 305 is used to extract corresponding geomagnetic related indicators based on the geomagnetic data and to extract corresponding acoustic related indicators based on the sound signal if the environmental verification is passed. The scoring module 306 is used to perform scoring processing based on the geomagnetic related indicators and the acoustic related indicators to obtain the corresponding environmental consistency score. The generation module 307 is used to perform data analysis on the environmental consistency score and generate fraud risk identification results corresponding to the face-to-face signing.

[0108] In this embodiment, the operations performed by the above modules or units correspond one-to-one with the steps of the artificial intelligence-based risk identification method in the aforementioned implementation method, and will not be repeated here.

[0109] In some optional implementations of this embodiment, the first verification module 302 includes: The first acquisition submodule is used to acquire a geomagnetic profile sequence corresponding to the target environment based on a preset magnetometer. The extraction submodule is used to extract features from the geomagnetic profile sequence based on a preset algorithm model to obtain the corresponding geomagnetic profile sequence features. The second acquisition submodule is used to acquire magnetic field change gradient information calculated based on the magnetometer. The integration submodule is used to integrate the geomagnetic profile sequence features and the magnetic field change gradient information to obtain the corresponding integrated data; The first determining submodule is used to use the integrated data as the geomagnetic data.

[0110] In this embodiment, the operations performed by the above modules or units correspond one-to-one with the steps of the artificial intelligence-based risk identification method in the aforementioned implementation method, and will not be repeated here.

[0111] In some optional implementations of this embodiment, the second verification module 304 includes: The analysis submodule is used to perform cepstral analysis on the sound signal to obtain the corresponding indoor impulse response index; Calling submodules is used to invoke pre-built acoustic feature models; The comparison submodule is used to compare the indoor impulse response index with the acoustic feature model to obtain the corresponding comparison results; The first determination submodule is used to determine that the environment verification is passed if the comparison result is a successful comparison. The second determination submodule is used to determine that the environment verification has failed if the comparison result is that the comparison fails.

[0112] In this embodiment, the operations performed by the above modules or units correspond one-to-one with the steps of the artificial intelligence-based risk identification method in the aforementioned implementation method, and will not be repeated here.

[0113] In some optional implementations of this embodiment, the scoring module 306 includes: The calculation submodule is used to calculate the corresponding cross-correlation coefficients based on the geomagnetic correlation index and the acoustic correlation index; The scoring submodule is used to score the geomagnetic correlation index, the acoustic correlation index and the cross-correlation coefficient based on a preset scoring strategy to obtain the corresponding scoring results. The second determining submodule is used to use the scoring result as the environmental consistency score.

[0114] In this embodiment, the operations performed by the above modules or units correspond one-to-one with the steps of the artificial intelligence-based risk identification method in the aforementioned implementation method, and will not be repeated here.

[0115] In some optional implementations of this embodiment, the geomagnetic correlation index includes at least the geomagnetic fluctuation frequency, and the acoustic correlation index includes at least the acoustic reflection frequency; the calculation submodule includes: An acquisition subunit is used to acquire the geomagnetic fluctuation frequency from the geomagnetic related indicators and the acoustic reflection frequency from the acoustic related indicators; The first calling subunit is used to call the preset cross-correlation coefficient formula; The first calculation subunit is used to calculate the geomagnetic wave frequency and the acoustic reflection frequency based on the cross-correlation coefficient formula to obtain the corresponding first calculation result; The first determining subunit is used to take the first calculation result as the cross-correlation coefficient.

[0116] In this embodiment, the operations performed by the above modules or units correspond one-to-one with the steps of the artificial intelligence-based risk identification method in the aforementioned implementation method, and will not be repeated here. In some optional implementations of this embodiment, the scoring submodule includes: A generation subunit is used to generate weight data corresponding to the geomagnetic related index, the acoustic related index, and the cross-correlation coefficient based on a preset weight allocation strategy. The second calling subunit is used to call the preset scoring model; The second calculation subunit is used to calculate and process the geomagnetic related index, the acoustic related index, the cross-correlation coefficient and the weight data based on the scoring model to obtain the corresponding second calculation result; The second determining subunit is used to take the second calculation result as the scoring result.

[0117] In this embodiment, the operations performed by the above modules or units correspond one-to-one with the steps of the artificial intelligence-based risk identification method in the aforementioned implementation method, and will not be repeated here.

[0118] In some optional implementations of this embodiment, the generation module 307 includes: The third acquisition submodule is used to acquire preset risk thresholds; The judgment submodule is used to determine whether the environmental consistency score is greater than or equal to the risk threshold; The third determination submodule is used to generate a first fraud identification result that the face-to-face signing does not have a fraud risk if the environmental consistency score is greater than or equal to the risk threshold. The third determination submodule is used to generate a second fraud identification result if the environmental consistency score is less than the risk threshold.

[0119] In this embodiment, the operations performed by the above modules or units correspond one-to-one with the steps of the artificial intelligence-based risk identification method in the aforementioned implementation method, and will not be repeated here. To address the aforementioned technical problems, embodiments of this application also provide a computer device. Please refer to [link / reference needed]. Figure 4 , Figure 4 This is a basic structural block diagram of the computer device in this embodiment.

[0120] The computer device 4 includes a memory 41, a processor 42, and a network interface 43 that are interconnected via a system bus. It should be noted that only the computer device 4 with components 41-43 is shown in the figure; however, it should be understood that it is not required to implement all the shown components, and more or fewer components can be implemented alternatively. Those skilled in the art will understand that the computer device described here is a device capable of automatically performing numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0121] The computer device can be a desktop computer, laptop, handheld computer, or cloud server, etc. The computer device can interact with the user via a keyboard, mouse, remote control, touchpad, or voice control.

[0122] The memory 41 includes at least one type of readable storage medium, including flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 41 may be an internal storage unit of the computer device 4, such as the hard disk or memory of the computer device 4. In other embodiments, the memory 41 may also be an external storage device of the computer device 4, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the computer device 4. Of course, the memory 41 may also include both the internal storage unit and its external storage device of the computer device 4. In this embodiment, the memory 41 is typically used to store the operating system and various application software installed on the computer device 4, such as computer-readable instructions for risk identification methods based on artificial intelligence. In addition, the memory 41 can also be used to temporarily store various types of data that have been output or will be output.

[0123] In some embodiments, the processor 42 may be a central processing unit (CPU), controller, microcontroller, microprocessor, or other data processing chip. The processor 42 is typically used to control the overall operation of the computer device 4. In this embodiment, the processor 42 is used to execute computer-readable instructions stored in the memory 41 or to process data, for example, to execute computer-readable instructions for the artificial intelligence-based risk identification method.

[0124] The network interface 43 may include a wireless network interface or a wired network interface, which is typically used to establish communication connections between the computer device 4 and other electronic devices.

[0125] Compared with the prior art, the embodiments of this application have the following beneficial effects: In this embodiment, during the face-to-face verification process via a mobile device, the application verifies the physical location of the user's current target environment using geomagnetic data based on a geomagnetic reference library, and verifies the environment using sound signals based on an acoustic feature model. Subsequently, geomagnetic-related indicators are extracted from the geomagnetic data, and acoustic-related indicators are extracted from the sound signals. An environmental consistency score is obtained by scoring the geomagnetic and acoustic-related indicators, and then data analysis is performed on the environmental consistency score to generate a fraud risk identification result corresponding to the face-to-face verification. Thus, this application performs preliminary environmental verification from two different dimensions: location and spatial characteristics. Based on this, it comprehensively analyzes multi-source environmental data containing geomagnetic and acoustic-related indicators, forming a complete and rigorous remote face-to-face verification authenticity verification system. This system can effectively identify and prevent various forms of fraud, improve the reliability of fraud identification in remote face-to-face verification, and ensure the accuracy of the generated fraud risk identification results.

[0126] This application also provides another embodiment, namely, providing a computer-readable storage medium storing computer-readable instructions that can be executed by at least one processor to cause the at least one processor to perform the steps of the artificial intelligence-based risk identification method described above.

[0127] Compared with the prior art, the embodiments of this application have the following main advantages: In this embodiment, during the face-to-face verification process via a mobile device, the application verifies the physical location of the user's current target environment using geomagnetic data based on a geomagnetic reference library, and verifies the environment using sound signals based on an acoustic feature model. Subsequently, geomagnetic-related indicators are extracted from the geomagnetic data, and acoustic-related indicators are extracted from the sound signals. An environmental consistency score is obtained by scoring the geomagnetic and acoustic-related indicators, and then data analysis is performed on the environmental consistency score to generate a fraud risk identification result corresponding to the face-to-face verification. Thus, this application performs preliminary environmental verification from two different dimensions: location and spatial characteristics. Based on this, it comprehensively analyzes multi-source environmental data containing geomagnetic and acoustic-related indicators, forming a complete and rigorous remote face-to-face verification authenticity verification system. This system can effectively identify and prevent various forms of fraud, improve the reliability of fraud identification in remote face-to-face verification, and ensure the accuracy of the generated fraud risk identification results.

[0128] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0129] Obviously, the embodiments described above are only some embodiments of this application, not all embodiments. The accompanying drawings show preferred embodiments of this application, but do not limit the patent scope of this application. This application can be implemented in many different forms; rather, the purpose of providing these embodiments is to provide a more thorough and comprehensive understanding of the disclosure of this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing specific embodiments, or make equivalent substitutions for some of the technical features. Any equivalent structures made using the content of this application's specification and drawings, directly or indirectly applied to other related technical fields, are similarly within the scope of patent protection of this application.

[0130] It should be noted that any AI models, software tools, or components not belonging to this company appearing in the embodiments of this application are merely illustrative examples and do not represent actual use. All user personal information involved in the embodiments of this application has been authorized (with the knowledge and consent) by the relevant parties or has been fully authorized by all parties, and the executing entity may obtain it through various legal and compliant means. The collection, storage, use, processing, transmission, provision, and disclosure of the information, data, and signals involved all comply with relevant laws and regulations and do not violate public order and good morals.

Claims

1. A risk identification method based on artificial intelligence, characterized in that, Includes the following steps: During the face-to-face verification process conducted by the user via mobile device, the geomagnetic data of the target environment in which the user is currently located is obtained; The geomagnetic data is physically located based on a pre-set geomagnetic reference library. If the geographical location verification is successful, then the sound signal of the target environment is collected; The sound signal is subjected to environmental verification based on a preset acoustic feature model; If the environmental verification is successful, the corresponding geomagnetic related indicators are extracted based on the geomagnetic data, and the corresponding acoustic related indicators are extracted based on the sound signal. The environmental consistency score is obtained by scoring the geomagnetic and acoustic related indicators. Data analysis is performed on the environmental consistency score to generate fraud risk identification results corresponding to the face-to-face signing.

2. The risk identification method based on artificial intelligence according to claim 1, characterized in that, The step of obtaining the geomagnetic data of the target environment where the user is currently located specifically includes: Obtain a geomagnetic profile sequence corresponding to the target environment based on a preset magnetometer; Based on a preset algorithm model, feature extraction is performed on the geomagnetic profile sequence to obtain the corresponding geomagnetic profile sequence features. Obtain the magnetic field change gradient information calculated based on the magnetometer; The geomagnetic profile sequence features and the magnetic field change gradient information are integrated to obtain corresponding integrated data. The integrated data is used as the geomagnetic data.

3. The risk identification method based on artificial intelligence according to claim 1, characterized in that, The step of performing environmental verification on the sound signal based on a preset acoustic feature model specifically includes: Perform cepstral analysis on the sound signal to obtain the corresponding indoor impulse response index; Call the pre-established acoustic feature model; The indoor impulse response index is compared with the acoustic feature model to obtain the corresponding comparison results; If the comparison result is a pass, then the environmental verification is deemed to have passed; If the comparison result is that the comparison fails, then the environmental verification is deemed to have failed.

4. The risk identification method based on artificial intelligence according to claim 1, characterized in that, The step of scoring based on the geomagnetic correlation index and the acoustic correlation index to obtain the corresponding environmental consistency score specifically includes: Calculate the corresponding cross-correlation coefficients based on the geomagnetic correlation index and the acoustic correlation index; The geomagnetic correlation index, the acoustic correlation index, and the cross-correlation coefficient are scored based on a preset scoring strategy to obtain the corresponding scoring results. The scoring result is used as the environmental consistency score.

5. The risk identification method based on artificial intelligence according to claim 4, characterized in that, The geomagnetic correlation index includes at least the geomagnetic fluctuation frequency, and the acoustic correlation index includes at least the acoustic reflection frequency; the step of calculating the corresponding cross-correlation coefficient based on the geomagnetic correlation index and the acoustic correlation index specifically includes: The geomagnetic fluctuation frequency is obtained from the geomagnetic related indicators, and the acoustic reflection frequency is obtained from the acoustic related indicators; Call the preset cross-correlation coefficient formula; The geomagnetic wave frequency and the acoustic reflection frequency are calculated based on the cross-correlation coefficient formula to obtain the corresponding first calculation result; The first calculation result is used as the cross-correlation coefficient.

6. The risk identification method based on artificial intelligence according to claim 4, characterized in that, The step of scoring the geomagnetic correlation index, the acoustic correlation index, and the cross-correlation coefficient based on a preset scoring strategy to obtain the corresponding scoring results specifically includes: Based on a preset weight allocation strategy, weight data corresponding to the geomagnetic related index, the acoustic related index, and the cross-correlation coefficient are generated respectively. Call the preset scoring model; Based on the scoring model, the geomagnetic correlation index, the acoustic correlation index, the cross-correlation coefficient, and the weight data are calculated and processed to obtain the corresponding second calculation result; The second calculation result is used as the scoring result.

7. The risk identification method based on artificial intelligence according to claim 1, characterized in that, The step of performing data analysis on the environmental consistency score to generate a fraud risk identification result corresponding to the face-to-face interview specifically includes: Obtain the preset risk threshold; Determine whether the environmental consistency score is greater than or equal to the risk threshold; If the environmental consistency score is greater than or equal to the risk threshold, a first fraud identification result is generated indicating that there is no fraud risk in the face-to-face interview. If the environmental consistency score is less than the risk threshold, a second fraud identification result is generated indicating that the face-to-face interview has a fraud risk.

8. A risk identification device based on artificial intelligence, characterized in that, include: The acquisition module is used to acquire geomagnetic data of the target environment where the user is currently located during the face-to-face verification process via a mobile device. The first verification module is used to verify the physical location of the geomagnetic data based on a preset geomagnetic reference library. The acquisition module is used to acquire sound signals from the target environment if the geographical location verification is successful. The second verification module is used to perform environmental verification on the sound signal based on a preset acoustic feature model; The extraction module is used to extract corresponding geomagnetic related indicators based on the geomagnetic data if the environmental verification is passed, and to extract corresponding acoustic related indicators based on the sound signal. The scoring module is used to perform scoring processing based on the geomagnetic related indicators and the acoustic related indicators to obtain the corresponding environmental consistency score. The generation module is used to perform data analysis on the environmental consistency score and generate fraud risk identification results corresponding to the face-to-face signing.

9. A computer device, characterized in that, It includes a memory and a processor, wherein the memory stores computer-readable instructions, and the processor executes the computer-readable instructions to implement the steps of the artificial intelligence-based risk identification method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the artificial intelligence-based risk identification method as described in any one of claims 1 to 7.

Images