WO2021139053A1 - 基于远程光体积描述术的风险评估方法、装置、设备及存储介质 - Google Patents

基于远程光体积描述术的风险评估方法、装置、设备及存储介质 Download PDF

Info

Publication number
WO2021139053A1
WO2021139053A1 PCT/CN2020/087661 CN2020087661W WO2021139053A1 WO 2021139053 A1 WO2021139053 A1 WO 2021139053A1 CN 2020087661 W CN2020087661 W CN 2020087661W WO 2021139053 A1 WO2021139053 A1 WO 2021139053A1
Authority
WO
WIPO (PCT)
Prior art keywords
heart rate
target person
risk assessment
data
respiration
Prior art date
Application number
PCT/CN2020/087661
Other languages
English (en)
French (fr)
Inventor
王家桢
屠宁
赵之砚
Original Assignee
深圳壹账通智能科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳壹账通智能科技有限公司 filed Critical 深圳壹账通智能科技有限公司
Publication of WO2021139053A1 publication Critical patent/WO2021139053A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/03Credit; Loans; Processing thereof
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00Scenes; Scene-specific elements
    • G06V20/40Scenes; Scene-specific elements in video content
    • G06V20/41Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/40Spoof detection, e.g. liveness detection
    • G06V40/45Detection of the body part being alive
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/15Biometric patterns based on physiological signals, e.g. heartbeat, blood flow

Definitions

  • This application relates to the field of fraud detection technology, and in particular to a risk assessment method, device, terminal device, and computer-readable storage medium based on remote optical volume description.
  • Risk assessment is an important task of the risk control department and an important control method related to the security of various businesses.
  • the method requires the lender to handle the interview offline, which is very cumbersome and consumes a lot of manpower and material resources for the bank.
  • offline sign-offs have gradually developed into online remote face-to-face review, which can eliminate the need for lenders to go out to reach designated outlets and eliminate the need for each branch to be equipped with face-to-face personnel. Online remote face-to-face review personnel can greatly reduce the bank's personnel consumption.
  • facial recognition technology can be used in the loan process to confirm the identity of the lender; micro-expression technology can be used in the loan process to recognize the small facial expressions of the lender.
  • micro-expression technology can be used in the loan process to recognize the small facial expressions of the lender.
  • both expression and face recognition are based on the single-frame image analysis of the video, the utilization rate of the video is low, and it is difficult to analyze the continuous emotional changes of the lender during the entire face-to-face review process, which leads to fraud assessment Large workload and low accuracy.
  • This application provides a risk assessment method, electronic device, terminal equipment, and computer-readable storage medium based on remote light volume description. Its main purpose is to obtain the target person’s heart rate and breathing abnormal time point and question time point through the evaluation video. Relationship, more intuitive and effective to obtain the target person’s fraud risk score.
  • this application provides a risk assessment method based on remote light volume description, which includes the following steps:
  • a fraud risk assessment method based on light volume description is applied to an electronic device, wherein the method includes:
  • the heart rate and respiration data are fed back to the risk assessment system in real time through the system interface for the risk assessment system to perform fraud risk assessment based on the heart rate and respiration data.
  • the present application also provides an electronic device, the electronic device comprising: a memory and a processor, the memory includes a risk assessment program based on remote light volume description technology, the risk based on remote light volume description technology The following steps are implemented when the evaluation program is executed by the processor:
  • the risk assessment system performs fraud risk assessment based on the heart rate and breathing data.
  • this application also provides a terminal device, wherein the device includes a memory, a processor, and a computer program that is stored in the memory and can run on the processor, and the processor executes The computer program implements the steps of the risk assessment method based on remote light volume description as described above.
  • the present application also provides a computer-readable storage medium, the computer-readable storage medium includes a risk assessment program based on remote photovolume description, and the risk assessment based on remote photovolume description
  • the program is executed by the processor, the steps of the risk assessment method based on remote light volume description as described above are realized.
  • the risk assessment method, electronic device, terminal equipment and computer readable storage medium based on remote light volume description proposed in this application use remote light volume description to monitor the heart rate and respiration of the target person during the risk assessment process.
  • FIG. 1 is a schematic diagram of an application environment of a preferred embodiment of a risk assessment method based on remote light volume description in this application;
  • FIG. 2 is a schematic diagram of modules of a preferred embodiment of a risk assessment program based on remote light volume description in FIG. 1;
  • FIG. 3 is a flowchart of a preferred embodiment of a risk assessment method based on remote light volume description in this application;
  • Figure 4 is a heart rate curve diagram of a specific embodiment of the application.
  • Fig. 5 is a respiration curve diagram of a specific embodiment of the application.
  • This application provides a risk assessment method based on remote light volume description, which is applied to an electronic device 1.
  • FIG. 1 it is a schematic diagram of the application environment of the preferred embodiment of the risk assessment method based on remote light volume description of this application.
  • the electronic device 1 may be a terminal device with a computing function such as a server, a smart phone, a tablet computer, a portable computer, a desktop computer, and the like.
  • the electronic device 1 includes a processor 12, a memory 11, a camera device 13, a network interface 14, and a communication bus 15.
  • the memory 11 includes at least one type of readable storage medium.
  • the at least one type of readable storage medium may be a non-volatile storage medium such as flash memory, hard disk, multimedia card, card-type memory, and the like.
  • the readable storage medium may be an internal storage unit of the electronic device 1, such as a hard disk of the electronic device 1.
  • the readable storage medium may also be an external memory of the electronic device 1, such as a plug-in hard disk or a smart memory card (Smart Media Card, SMC) equipped on the electronic device 1. Secure Digital (SD) card, Flash Card, etc.
  • SD Secure Digital
  • the readable storage medium of the memory 11 is usually used to store the risk assessment program 10 based on remote volumetric description, the face image sample library, and the pre-trained AU installed in the electronic device 1. Classifiers, sentiment classifiers, etc.
  • the memory 11 can also be used to temporarily store data that has been output or will be output.
  • the processor 12 may be a central processing unit (CPU), a microprocessor or other data processing chip, which is used to run the program code or process data stored in the memory 11, for example to execute remote optical processing units.
  • CPU central processing unit
  • microprocessor or other data processing chip, which is used to run the program code or process data stored in the memory 11, for example to execute remote optical processing units.
  • the network interface 14 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface), and is generally used to establish a communication connection between the electronic device 1 and other electronic devices.
  • the communication bus 15 is used to realize the connection and communication between these components.
  • FIG. 1 only shows the electronic device 1 with the components 11-15, but it should be understood that it is not required to implement all the illustrated components, and more or fewer components may be implemented instead.
  • the electronic device 1 may also include a user interface.
  • the user interface may include an input unit such as a keyboard (Keyboard), a voice input device such as a microphone (microphone) and other devices with voice recognition functions, and a voice output device such as audio, earphones, etc.
  • the user interface may also include a standard wired interface and a wireless interface.
  • the electronic device 1 may also include a display, and the display may also be referred to as a display screen or a display unit.
  • the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an organic light-emitting diode (OLED) touch device, and the like.
  • the display is used for displaying information processed in the electronic device 1 and for displaying a visualized user interface.
  • the electronic device 1 further includes a touch sensor.
  • the area provided by the touch sensor for the user to perform a touch operation is called a touch area.
  • the touch sensor described here may be a resistive touch sensor, a capacitive touch sensor, or the like.
  • the touch sensor includes not only a contact type touch sensor, but also a proximity type touch sensor and the like.
  • the touch sensor may be a single sensor, or may be, for example, a plurality of sensors arranged in an array.
  • the area of the display of the electronic device 1 may be the same as or different from the area of the touch sensor.
  • the display and the touch sensor are stacked to form a touch display screen. The device detects the touch operation triggered by the user based on the touch screen.
  • the electronic device 1 may also include a radio frequency (RF) circuit, a sensor, an audio circuit, etc., which will not be repeated here.
  • RF radio frequency
  • the memory 11 as a computer storage medium may include an operating system and a risk assessment program 10 based on remote photovolume description; the processor 12 executes the remote based storage in the memory 11
  • the risk assessment procedure of photovolume description technique is as follows:
  • the risk assessment system performs fraud risk assessment based on the heart rate and breathing data.
  • the step of the risk assessment system performing fraud risk assessment based on the heart rate and respiration data includes:
  • the step of the risk assessment system performing fraud risk assessment based on the heart rate and respiration data further includes:
  • the voiceprint information of the questioner and the target person in the evaluation video is collected in advance, and the speaking time of the questioner and the target person is cut based on the voiceprint recognition technology to obtain that the target person starts to speak At the same time, automatically calculate whether the heart rate and breathing data of the target person at the time point are abnormal, and evaluate and score the fraud risk of the evaluation video through the artificial intelligence system;
  • the facial and lip movements of the target person in the evaluation video are monitored, and based on the speech-to-text or voiceprint recognition technology, the time point when the target person starts to speak is acquired, and the target person at the time point is automatically calculated Whether there is any abnormality in the heart rate and respiration data, and evaluate and score the fraud risk of the evaluation video through the artificial intelligence system.
  • the step of assessing and scoring the fraud risk of the assessed video through the artificial intelligence system includes:
  • a fraud risk score is obtained based on the abnormal ratio.
  • the formula for obtaining the abnormal ratio of the heart rate or respiration data is:
  • a is a heart rate value point or respiration value point in the heart rate or respiration data
  • n is an abnormal point of the heart rate or respiration data
  • m is an average heart rate value or respiration value in a preset time period before the abnormal point n.
  • the formula for obtaining a fraud risk score based on the abnormal ratio is:
  • P is the fraud risk score of the evaluation video
  • N is the heart rate or the sum of the respiration value points in the heart rate or respiration data of the evaluation video
  • ⁇ a is the heart rate or the abnormality of the respiration data of the evaluation video The sum of the proportions.
  • the risk assessment program 10 based on remote volumetric description technology may also be divided into one or more modules, and the one or more modules are stored in the memory 11 and executed by the processor 12 to complete this Application.
  • the module referred to in this application refers to a series of computer program instruction segments that can complete specific functions.
  • FIG. 2 it is a program module diagram of a preferred embodiment of the risk assessment program 10 based on remote light volume description in FIG. 1.
  • the risk assessment program 10 based on remote light volume description can be divided into:
  • the model creation unit 110 is configured to create a remote optical volume description technology model, and establish a system interface for real-time video streaming and feedback of the remote optical volume description technology model;
  • the evaluation data collection unit 120 is configured to transmit the video stream to be detected to the remote optical volume description model in real time through the system interface, and collect the information of the target person in the video stream based on the remote optical volume description model.
  • Heart rate and breathing data
  • the data feedback unit 130 is configured to feed back the heart rate and respiration data to the risk assessment system in real time through the system interface;
  • the risk assessment unit 140 is used for the risk assessment system to perform fraud risk assessment based on the heart rate and respiration data.
  • this application also provides a risk assessment method based on remote light volume description.
  • FIG. 3 it is a flowchart of a preferred embodiment of a risk assessment method based on remote light volume description in this application.
  • the method can be executed by a device, and the device can be implemented by software and/or hardware.
  • the risk assessment method based on the remote light volume description technique includes:
  • S110 Create a remote optical volume description technology model, and establish a system interface for real-time video streaming and feedback of the remote optical volume description technology model.
  • the input parameters of the system interface for real-time video streaming and feedback include the video stream and the frequency of the output heart rate. Assuming that the frame rate of the input video stream is 15fps, the output heart rate or respiration frequency is at most 15 per second. Value, assuming that the frame rate of the input video stream is 60fps, the output heart rate or respiration frequency can be up to 60 values per second. In other words, the higher the frame rate of the input video stream, the output heart rate or respiration frequency The higher the value, the finer the heart rate or respiration curve is drawn, the easier it is to find the heart rate or respiration changes that occur at a certain moment, and the corresponding remote photovolume description model will take longer to collect data.
  • the frame rate of the video stream (or The frequency of the frame rate value of the video stream) outputs the heart rate or respiration.
  • the output parameters of the real-time video streaming and feedback system interface include: the time point and the heart rate or respiration corresponding to the time point.
  • S120 Transmit the video stream to be detected to the remote light volume description model in real time through the system interface, and collect heart rate and breathing data of the target person in the video stream based on the remote light volume description model.
  • S130 Real-time feedback of the heart rate and respiration data to the risk assessment system through the system interface, for the risk assessment system to perform fraud risk assessment based on the heart rate and respiration data.
  • the risk assessment system may be a variety of application systems or programs.
  • the risk assessment system may be a corresponding loan system or a remote face-to-face examination subsystem in the loan system.
  • the risk assessment system performs fraud risk assessment based on the heart rate and breathing data, including the following four situations:
  • the second case Convert the assessment video dialogue content of the target person in the risk assessment process into text data in real time, and then automatically determine the time point when the target person answers the question based on the artificial intelligence system, and automatically calculate the time point at the same time Whether there is an abnormality in the heart rate and breathing data of the target person, and evaluate and score the fraud risk of the evaluation video through the artificial intelligence system;
  • the third case pre-acquire the voiceprint information of the questioner and the target person in the evaluation video, and cut the speaking time of the questioner and the target person based on the voiceprint recognition technology to obtain the target The time point when the character starts to speak, and at the same time automatically calculates whether the heart rate and breathing data of the target character at the time point are abnormal, and evaluates and scores the fraud risk of the evaluation video through the artificial intelligence system;
  • the fourth case monitor the facial and lip movements of the target person in the evaluation video, and obtain the time point when the target person starts speaking based on the speech-to-text or voiceprint recognition technology, and automatically calculate the time point at the same time Whether there is an abnormality in the heart rate and breathing data of the target person, and evaluate and score the fraud risk of the evaluation video through the artificial intelligence system.
  • the step of assessing and scoring the fraud risk of the assessed video through the artificial intelligence system includes:
  • the formula for obtaining the abnormal ratio of the heart rate or respiration data is:
  • a is a heart rate value point or respiration value point in the heart rate or respiration data
  • n is an abnormal point of the heart rate or respiration data
  • m is an average heart rate value or respiration value in a preset time period before the abnormal point n.
  • the formula for obtaining a fraud risk score based on the abnormal ratio is:
  • P is the fraud risk score of the evaluation video
  • N is the heart rate or the sum of the respiration value points in the heart rate or respiration data of the evaluation video
  • ⁇ a is the heart rate or the abnormality of the respiration data of the evaluation video The sum of the proportions.
  • the steps for scoring fraud risk through AI artificial intelligence include:
  • the points with larger residual values (the points whose residual value exceeds the set residual value range, such as the value exceeding 3 times the standard deviation in the following example), high leverage points and strong
  • the influence points are set as preliminary abnormal points; here, the 3 ⁇ principle is used, assuming that the heart rate or breathing data obeys a normal distribution under normal conditions.
  • the abnormal value is defined as: a set of measured values and The deviation of the average value exceeds 3 times the standard deviation value.
  • the probability of occurrence of a value other than the average value 3 ⁇ is: P(
  • >3 ⁇ ) ⁇ 0.003.
  • the initial abnormality is confirmed as the confirmed abnormality, otherwise it is the normal point.
  • the abnormal ratio of the heart rate or breathing data at any point
  • the fraud risk score of the entire video ( ⁇ abnormal ratio of each point)/the total number of points on the heart rate and respiration curve of the entire video.
  • the above process of scoring fraud risk through AI artificial intelligence can also be achieved by training a fraud scoring model, that is, training a neural network model by collecting data to form a fraud scoring model, and then inputting the collected heart rate and breathing data into the laboratory.
  • a fraud scoring model that is, training a neural network model by collecting data to form a fraud scoring model, and then inputting the collected heart rate and breathing data into the laboratory.
  • Fraud scoring in the fraud scoring model described above using the method of training the scoring model, although it can improve the accuracy of fraud risk assessment, but its operation is complicated and the cost is high. In the specific application process, you can choose according to the application scenarios and needs. There is no restriction on the way of risk assessment.
  • FIG. 4 and FIG. 5 respectively show schematic structures of a heart rate curve and a breathing curve of a target person in the remote face-to-face audit blacklist according to a specific embodiment of the present application.
  • the lender had a brief heart rate rise and held his breath in the video.
  • the moment his heart rate rose and held his breath coincided with the time when the lender answered the interviewer’s question, and the interviewer only asked questions about his mobile phone number, and the lender stammered slightly when answering.
  • the above-mentioned risk assessment method based on remote light volume description technology uses remote light volume description technology to monitor the heart rate and respiration of the target person, and uses the heart rate and respiration to judge the mental state of the target person.
  • remote light volume description technology uses remote light volume description technology to monitor the heart rate and respiration of the target person, and uses the heart rate and respiration to judge the mental state of the target person.
  • AI artificial intelligence technology the relationship between the abnormal time point of heart rate and breathing during the entire face-to-face interview process and the time point of the target person’s answer to the question can be obtained, and fraud risk assessment can be made based on this, which can give a more intuitive and understandable result. Convincing judgment result.
  • an embodiment of the present application also proposes a terminal device, which includes a memory, a processor, and a computer program stored in the memory and capable of running on the processor.
  • a terminal device which includes a memory, a processor, and a computer program stored in the memory and capable of running on the processor.
  • the processor executes the computer program, The steps of implementing the risk assessment method based on remote light volume description.
  • the specific implementation of the terminal device of the present application is substantially the same as the specific implementation of the risk assessment method based on the remote light volume description technique, and will not be repeated here.
  • the embodiment of the present application also proposes a computer-readable storage medium.
  • the computer-readable storage medium may be non-volatile or volatile.
  • the computer-readable storage medium includes a risk assessment program based on remote photovolume description, and when the risk assessment program based on remote photovolume description is executed by a processor, the following operations are implemented:
  • a fraud risk assessment method based on light volume description, applied to an electronic device wherein the method includes:
  • the heart rate and respiration data are fed back to the risk assessment system in real time through the system interface for the risk assessment system to perform fraud risk assessment based on the heart rate and respiration data.
  • the step of the risk assessment system performing fraud risk assessment based on the heart rate and respiration data includes:
  • the step of the risk assessment system performing fraud risk assessment based on the heart rate and respiration data further includes:
  • the voiceprint information of the questioner and the target person in the evaluation video is collected in advance, and the speaking time of the questioner and the target person is cut based on the voiceprint recognition technology to obtain that the target person starts to speak At the same time, automatically calculate whether the heart rate and breathing data of the target person at the time point are abnormal, and evaluate and score the fraud risk of the evaluation video through the artificial intelligence system;
  • the facial and lip movements of the target person in the evaluation video are monitored, and based on the speech-to-text or voiceprint recognition technology, the time point when the target person starts to speak is acquired, and the target person at the time point is automatically calculated Whether there is any abnormality in the heart rate and respiration data, and evaluate and score the fraud risk of the evaluation video through the artificial intelligence system.
  • the step of assessing and scoring the fraud risk of the assessed video through the artificial intelligence system includes:
  • the formula for obtaining the abnormal ratio of the heart rate or respiration data is:
  • a is a heart rate value point or respiration value point in the heart rate or respiration data
  • n is an abnormal point of the heart rate or respiration data
  • m is an average heart rate value or respiration value in a preset time period before the abnormal point n.
  • the formula for obtaining a fraud risk score based on the abnormal ratio is:
  • P is the fraud risk score of the evaluation video
  • N is the heart rate or the sum of the respiration value points in the heart rate or respiration data of the evaluation video
  • ⁇ a is the heart rate or the abnormality of the respiration data of the evaluation video The sum of the proportions.
  • the specific implementation of the computer-readable storage medium of the present application is substantially the same as the specific implementation of the risk assessment method and the electronic device based on the remote light volume description technique, and will not be repeated here.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Multimedia (AREA)
  • Development Economics (AREA)
  • Technology Law (AREA)
  • General Business, Economics & Management (AREA)
  • Strategic Management (AREA)
  • Marketing (AREA)
  • Computational Linguistics (AREA)
  • Software Systems (AREA)
  • Economics (AREA)
  • Human Computer Interaction (AREA)
  • Image Analysis (AREA)
  • Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)

Abstract

一种基于远程光体积描述术的风险评估方法、装置、设备及存储介质,涉及欺诈检测技术领域。其中的方法包括创建远程光体积描述术模型,并建立所述远程光体积描述术模型的用于实时视频流传输和反馈的***接口(S110);将待检测的视频流通过所述***接口实时传输至所述远程光体积描述术模型,基于所述远程光体积描述术模型采集所述视频流中的目标人物的心率和呼吸数据(S120);通过所述***接口实时反馈所述心率和呼吸数据至风险评估***,用于所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估(S130)。通过评估视频获取目标人物的心率和呼吸异常时间点与提问时间点的关系,更加直观有效的获取目标人物的欺诈风险评分。

Description

基于远程光体积描述术的风险评估方法、装置、设备及存储介质
本申请要求于2020年1月6日提交中国专利局、申请号为202010009999.1,申请名称为“基于远程光体积描述术的风险评估方法、装置及计算机可读存储介质”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本申请涉及欺诈检测技术领域,尤其涉及一种基于远程光体积描述术的风险评估方法、装置、终端设备及计算机可读存储介质。
背景技术
风险评估是风控部门的重要任务,也是关系到各类业务安全的重要控制手段,而随着互联网时代的发展,大量业务从线下搬到了线上,例如对于贷款业务来说,传统的贷款方式需要贷款人在线下办理面签,流程非常繁琐,对银行来说也消耗较多的人力物力。目前,线下面签逐渐发展成了线上远程面审,这样可以免除贷款者需要外出到达指定网点,以及无需每个网点均配备面签人员,在线远程面审人员可以大大减少银行的人员消耗。
针对线上远程面审,业内将诸多AI人工智能产品嵌入其中,试图通过视频对贷款者进行分析,从而建立反欺诈风控机制。举例,人脸识别技术在贷款过程中使用,可以对贷款者的身份进行确认;微表情技术在贷款过程中使用,可以对贷款者的面部细小的表情进行识别。但是,发明人发现表情和人脸识别都是基于视频的单帧图像进行分析,对视频的利用率较低,较难分析出贷款者在整个面审过程中连续的情绪变化,导致欺诈评估的工作量大、准确性低。
发明内容
本申请提供一种基于远程光体积描述术的风险评估方法、电子装置、终端设备及计算机可读存储介质,其主要目的在于通过评估视频获取目标人物的心率和呼吸异常时间点与提问时间点的关系,更加直观有效的获取目标人物的欺诈风险评分。
为实现上述目的,本申请提供一种基于远程光体积描述术的风险评估方法,包括以下步骤:
基于光体积描述术的欺诈风险评估方法,应用于电子装置,其中,所述 方法包括:
创建远程光体积描述术模型,并建立所述远程光体积描述术模型的用于实时视频流传输和反馈的***接口;
将待检测的视频流通过所述***接口实时传输至所述远程光体积描述术模型,基于所述远程光体积描述术模型采集所述视频流中的目标人物的心率和呼吸数据;
通过所述***接口实时反馈所述心率和呼吸数据至风险评估***,用于所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估。
为实现上述目的,本申请还提供一种电子装置,该电子装置包括:存储器及处理器,所述存储器中包括基于远程光体积描述术的风险评估程序,所述基于远程光体积描述术的风险评估程序被所述处理器执行时实现如下步骤:
创建远程光体积描述术模型,并建立所述远程光体积描述术模型的用于实时视频流传输和反馈的***接口;
将待检测的视频流通过所述***接口实时传输至所述远程光体积描述术模型,基于所述远程光体积描述术模型采集所述视频流中的目标人物的心率和呼吸数据;
通过所述***接口实时反馈所述心率和呼吸数据至风险评估***,用于
所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估。
此外,为实现上述目的,本申请还提供一种终端设备,其中,该设备包括存储器、处理器以及存储在所述存储器中并可以在所述处理器上运行的计算机程序,所述处理器执行所述计算机程序时实现如上所述的基于远程光体积描述术的风险评估方法的步骤。
此外,为实现上述目的,本申请还提供一种计算机可读存储介质,所述计算机可读存储介质中包括基于远程光体积描述术的风险评估程序,所述基于远程光体积描述术的风险评估程序被处理器执行时,实现如上所述的基于远程光体积描述术的风险评估方法的步骤。
本申请提出的基于远程光体积描述术的风险评估方法、电子装置、终端设备及计算机可读存储介质,利用远程光体积描述术在风险评估过程中对目标人物进行心率和呼吸的监测,同时,结合AI人工智能技术,获取整个评估过程中的心率和呼吸的异常时间点与目标人物回答问题的时间点的关系,以 此为依据进行欺诈风险评估,能够给出更加直观可理解和信服的判断结果。
附图说明
图1为本申请基于远程光体积描述术的风险评估方法较佳实施例的应用环境示意图;
图2为图1中基于远程光体积描述术的风险评估程序较佳实施例的模块示意图;
图3为本申请基于远程光体积描述术的风险评估方法较佳实施例的流程图;
图4为本申请具体实施例的心率曲线图;
图5为本申请具体实施例的呼吸曲线图。
本申请目的的实现、功能特点及优点将结合实施例,参照附图做进一步说明。
具体实施方式
应当理解,此处所描述的具体实施例仅仅用以解释本申请,并不用于限定本申请。
本申请提供一种基于远程光体积描述术的风险评估方法,应用于一种电子装置1。参照图1所示,为本申请基于远程光体积描述术的风险评估方法较佳实施例的应用环境示意图。
在本实施例中,电子装置1可以是服务器、智能手机、平板电脑、便携计算机、桌上型计算机等具有运算功能的终端设备。
该电子装置1包括:处理器12、存储器11、摄像装置13、网络接口14及通信总线15。
存储器11包括至少一种类型的可读存储介质。所述至少一种类型的可读存储介质可为如闪存、硬盘、多媒体卡、卡型存储器等的非易失性存储介质。在一些实施例中,所述可读存储介质可以是所述电子装置1的内部存储单元,例如该电子装置1的硬盘。在另一些实施例中,所述可读存储介质也可以是所述电子装置1的外部存储器,例如所述电子装置1上配备的插接式硬盘,智能存储卡(Smart Media Card,SMC),安全数字(Secure Digital,SD)卡,闪存卡(Flash Card)等。
在本实施例中,所述存储器11的可读存储介质通常用于存储安装于所述 电子装置1的基于远程光体积描述术的风险评估程序10、人脸图像样本库及预先训练好的AU分类器、情绪分类器等。所述存储器11还可以用于暂时地存储已经输出或者将要输出的数据。
处理器12在一些实施例中可以是一中央处理器(Central Processing Unit,CPU),微处理器或其他数据处理芯片,用于运行存储器11中存储的程序代码或处理数据,例如执行基于远程光体积描述术的风险评估程序10等。
网络接口14可选地可以包括标准的有线接口、无线接口(如WI-FI接口),通常用于在该电子装置1与其他电子设备之间建立通信连接。
通信总线15用于实现这些组件之间的连接通信。
图1仅示出了具有组件11-15的电子装置1,但是应理解的是,并不要求实施所有示出的组件,可以替代的实施更多或者更少的组件。
可选地,该电子装置1还可以包括用户接口,用户接口可以包括输入单元比如键盘(Keyboard)、语音输入装置比如麦克风(microphone)等具有语音识别功能的设备、语音输出装置比如音响、耳机等,可选地用户接口还可以包括标准的有线接口、无线接口。
可选地,该电子装置1还可以包括显示器,显示器也可以称为显示屏或显示单元。在一些实施例中可以是LED显示器、液晶显示器、触控式液晶显示器以及有机发光二极管(Organic Light-Emitting Diode,OLED)触摸器等。显示器用于显示在电子装置1中处理的信息以及用于显示可视化的用户界面。
可选地,该电子装置1还包括触摸传感器。所述触摸传感器所提供的供用户进行触摸操作的区域称为触控区域。此外,这里所述的触摸传感器可以为电阻式触摸传感器、电容式触摸传感器等。而且,所述触摸传感器不仅包括接触式的触摸传感器,也可包括接近式的触摸传感器等。此外,所述触摸传感器可以为单个传感器,也可以为例如阵列布置的多个传感器。
此外,该电子装置1的显示器的面积可以与所述触摸传感器的面积相同,也可以不同。可选地,将显示器与所述触摸传感器层叠设置,以形成触摸显示屏。该装置基于触摸显示屏侦测用户触发的触控操作。
可选地,该电子装置1还可以包括射频(Radio Frequency,RF)电路,传感器、音频电路等等,在此不再赘述。
在图1所示的装置实施例中,作为一种计算机存储介质的存储器11中可 以包括操作***、以及基于远程光体积描述术的风险评估程序10;处理器12执行存储器11中存储的基于远程光体积描述术的风险评估程序10时实现如下步骤:
创建远程光体积描述术模型,并建立所述远程光体积描述术模型的用于实时视频流传输和反馈的***接口;
将待检测的视频流通过所述***接口实时传输至所述远程光体积描述术模型,基于所述远程光体积描述术模型采集所述视频流中的目标人物的心率和呼吸数据;
通过所述***接口实时反馈所述心率和呼吸数据至风险评估***,用于
所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估。
优选地,所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估的步骤包括:
在所述风险评估***的显示界面中实时显示所述目标人物的心率和呼吸数据;
当所述心率或呼吸数据在预设时间内升高幅度超过设定值时,进一步判定所述心率或呼吸数据升高提示的时间点与所述风险评估***提出问题的时间点是否在预设邻近范围内;
当所述心率或呼吸数据升高提示的时间点与所述风险评估***提出问题的时间点在预设邻近范围内时,判定所述目标人物存在欺诈风险。
优选地,所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估的步骤还包括:
将所述目标人物在风险评估过程中的评估视频对话内容实时转换成文本数据,然后基于人工智能***自动判断所述目标人物回答问题的时间点,同时自动测算所述时间点的目标人物的心率和呼吸数据是否存在异常,并通过所述人工智能***对所述评估视频的欺诈风险进行评估打分;
或者,预先采集所述评估视频中提问人员和所述目标人物的声纹信息,并基于声纹识别技术对所述提问人员和所述目标人物的说话时间进行切割,获取所述目标人物开始说话的时间点,同时自动测算所述时间点的目标人物的心率和呼吸数据是否存在异常,并通过所述人工智能***对所述评估视频的欺诈风险进行评估打分;
或者,对所述评估视频中的目标人物的面部嘴唇动作进行监测,并基于语音转文字或声纹识别技术,获取所述目标人物开始说话的时间点,同时自动测算所述时间点的目标人物的心率和呼吸数据是否存在异常,并通过所述人工智能***对所述评估视频的欺诈风险进行评估打分。
优选地,所述通过所述人工智能***对所述评估视频的欺诈风险进行评估打分的步骤包括:
设定所述心率或呼吸数据中残值较大的点、高杠杆点和强影响点作为初步异常点;
判断所述初步异常点与所述目标人物开始说话的时间点之间的时间差是否在设定范围内,若所述时间差在所述设定范围内,则确认所述初步异常点为确认异常点;
获取所述心率或呼吸数据的异常比例,并
基于所述异常比例获取欺诈风险评分。
优选地,所述心率或呼吸数据的异常比例获取公式为:
Figure PCTCN2020087661-appb-000001
其中,a为心率或呼吸数据中的心率值点或呼吸值点,n为心率或呼吸数据的异常点,m为所述异常点n之前预设时间段内的平均心率值或呼吸值。
优选地,所述基于所述异常比例获取欺诈风险评分的公式为:
Figure PCTCN2020087661-appb-000002
其中,P为所述评估视频的欺诈风险评分,N为所述评估视频的心率或呼吸数据中的心率值点或呼吸值点的总和,∑a为所述评估视频的心率或呼吸数据的异常比例总和。
在其他实施例中,基于远程光体积描述术的风险评估程序10还可以被分割为一个或者多个模块,一个或者多个模块被存储于存储器11中,并由处理器12执行,以完成本申请。本申请所称的模块是指能够完成特定功能的一系列计算机程序指令段。
参照图2所示,为图1中基于远程光体积描述术的风险评估程序10较佳实施例的程序模块图。所述基于远程光体积描述术的风险评估程序10可以被分割为:
模型创建单元110,用于创建远程光体积描述术模型,并建立所述远程光 体积描述术模型的用于实时视频流传输和反馈的***接口;
评估数据采集单元120,用于将待检测的视频流通过所述***接口实时传输至所述远程光体积描述术模型,基于所述远程光体积描述术模型采集所述视频流中的目标人物的心率和呼吸数据;
数据反馈单元130,用于将所述心率和呼吸数据通过所述***接口实时反馈至风险评估***;
风险评估单元140,用于所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估。
此外,本申请还提供一种基于远程光体积描述术的风险评估方法。参照图3所示,为本申请基于远程光体积描述术的风险评估方法较佳实施例的流程图。该方法可以由一个装置执行,该装置可以由软件和/或硬件实现。
在本实施例中,基于远程光体积描述术的风险评估方法包括:
S110:创建远程光体积描述术模型,并建立所述远程光体积描述术模型的用于实时视频流传输和反馈的***接口。
其中,实时视频流传输和反馈的***接口的输入参数包括视频流和输出心率的频度,假设输入的视频流的帧率为15fps,则输出的心率或呼吸的频度最多每秒钟15个值,假设输入的视频流的帧率为60fps,则输出的心率或呼吸的频度最多每秒钟60个值,换言之,输入的视频流帧率越高,则可输出的心率或呼吸频度也越高,进而绘制出的心率或呼吸曲线也越精细,在某个瞬间发生的心率或呼吸变化越容易被发现,对应的远程光体积描述术模型采集数据时的耗时也会较长。
另外,对于不同的业务要求,可以自行进行输出心率或呼吸的频度设置,如果输入参数中设置的心率或呼吸频度高于输入视频流本身的帧率,则以视频流的帧率(或视频流的帧率值的频度)输出心率或呼吸。
此外,实时视频流传输和反馈***接口的输出参数包括:时间点和与时间点对应的心率或呼吸。
S120:将待检测的视频流通过所述***接口实时传输至所述远程光体积描述术模型,基于所述远程光体积描述术模型采集所述视频流中的目标人物的心率和呼吸数据。
S130:通过所述***接口实时反馈所述心率和呼吸数据至风险评估***, 用于所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估。
其中,所述风险评估***可以为多种应用***或者程序,当风险评估对象为贷款人时,该风险评估***可以为对应的贷款***或者贷款***中的远程面审子***。
具体地,在贷款人面审过程中:1面签开始话术、2信息核实/提问、3联网核查、4核对银行***、5拍照合影及上传、6签合同/贷款信息告知、7合同简单告知、资金成本、还款等信息告知、8客户条款朗读、9贷后信息告知及确认;以上这些环节均由面审人员和贷款人通过远程视频的方式完成,整个过程中会产生两段视频,一段是对面审人员的录制,另一段是对贷款人的录制,两段视频长度相同,在归档时,会合并到一个画面中,便于后续视频的观看时可以两个画面同步播放。在归档前画面未做合并时,将贷款人的视频传输给远程光体积描述术模型的实时视频流传输和反馈***接口,能够实时获取贷款人的心率和呼吸数据。
此外,风险评估***根据所述心率和呼吸数据进行欺诈风险评估包括以下四种情况:
第一种情况:
1、在所述风险评估***的显示界面中实时显示所述目标人物的心率和呼吸数据;
2、当所述心率或呼吸数据在预设时间内升高幅度超过设定值时,进一步判定所述心率或呼吸数据升高提示的时间点与所述风险评估***提出问题的时间点是否在预设邻近范围内;
3、当所述心率或呼吸数据升高提示的时间点与所述风险评估***提出问题的时间点在预设邻近范围内时,判定所述目标人物存在欺诈风险。
第二种情况:将所述目标人物在风险评估过程中的评估视频对话内容实时转换成文本数据,然后基于人工智能***自动判断所述目标人物回答问题的时间点,同时自动测算所述时间点的目标人物的心率和呼吸数据是否存在异常,并通过所述人工智能***对所述评估视频的欺诈风险进行评估打分;
第三种情况:预先采集所述评估视频中提问人员和所述目标人物的声纹信息,并基于声纹识别技术对所述提问人员和所述目标人物的说话时间进行切割,获取所述目标人物开始说话的时间点,同时自动测算所述时间点的目 标人物的心率和呼吸数据是否存在异常,并通过所述人工智能***对所述评估视频的欺诈风险进行评估打分;
第四种情况:对所述评估视频中的目标人物的面部嘴唇动作进行监测,并基于语音转文字或声纹识别技术,获取所述目标人物开始说话的时间点,同时自动测算所述时间点的目标人物的心率和呼吸数据是否存在异常,并通过所述人工智能***对所述评估视频的欺诈风险进行评估打分。
在本申请的一个具体实施方式中,通过所述人工智能***对所述评估视频的欺诈风险进行评估打分的步骤包括:
1、设定所述心率或呼吸数据中残值较大的点、高杠杆点和强影响点作为初步异常点;
2、判断所述初步异常点与所述目标人物开始说话的时间点之间的时间差是否在设定范围内,若所述时间差在所述设定范围内,则确认所述初步异常点为确认异常点;
3、获取所述心率或呼吸数据的异常比例,并基于所述异常比例获取欺诈风险评分。
具体地,所述心率或呼吸数据的异常比例获取公式为:
Figure PCTCN2020087661-appb-000003
其中,a为心率或呼吸数据中的心率值点或呼吸值点,n为心率或呼吸数据的异常点,m为所述异常点n之前预设时间段内的平均心率值或呼吸值。
所述基于所述异常比例获取欺诈风险评分的公式为:
Figure PCTCN2020087661-appb-000004
其中,P为所述评估视频的欺诈风险评分,N为所述评估视频的心率或呼吸数据中的心率值点或呼吸值点的总和,∑a为所述评估视频的心率或呼吸数据的异常比例总和。
换言之,通过AI人工智能进行欺诈风险打分的步骤包括:
1、查找心率和呼吸数据的初步异常点,其中可将残值较大的点(残值超过设定残值范围的点,如下例中超过3倍标准差的值)、高杠杆点和强影响点等设定为初步异常点;此处,采用3σ原则的方法,假设心率或呼吸数据正常情况下服从正态分布,在3σ的原则下,异常值被定义为:一组测定值中与平均值的偏差超过3倍标准差的值。在正态分布的假设下,举例平均值3σ之外 的值出现概率为:P(|x-μ|>3σ)<=0.003。
2、目标人物开始说话的时间点与心率或呼吸初步异常点的时间点之间的时间差的绝对值在1s以内的,则确认该初步异常点为确认异常点,否则为正常点。
3、获取任一点的心率或呼吸数据的异常比例;某一点的异常比例=|1-异常点的心率或呼吸值/异常点前面3秒的平均心率或呼吸值|,某一点值的是心率或呼吸曲线上的心率值点或者呼吸值点。
4、基于所述异常比例获取欺诈风险评分。其中,整个视频的欺诈风险分数=(Σ每个点的异常比例)/整个视频的心率和呼吸曲线上的点的总数。
需要说明的是,上述通过AI人工智能进行欺诈风险打分的过程还可以通过训练欺诈打分模型实现,即通过采集数据训练神经网络模型,形成欺诈打分模型,然后将采集到的心率和呼吸数据输入所述欺诈打分模型中进行欺诈打分,采用训练打分模型的方式,虽然能够提高欺诈风险评估的准确性,但是其操作复杂,成本较高,在具体应用过程中,可根据应用场景及需求,自行选择风险评估的途径,再此不做限制。
作为具体示例,图4和图5分别示出了根据本申请具体实施例的远程面审黑名单目标人物的心率曲线和呼吸曲线示意结构。
根据图4和图5所示,贷款人在视频中有短暂的心率上升及屏住呼吸的情况。再次播放视频时发现,其心率上升和屏住呼吸的瞬间,与该贷款人回答面审人员问题的时间点相吻合,且面审人员仅仅是对其手机号码进行提问,贷款者回答时略微结巴停顿,同时从心率和呼吸情况可以看出贷款者的紧张情绪,进而可结合人工智能实现对目标人物的欺诈评估。
上述基于远程光体积描述术的风险评估方法所提出的利用远程光体积描述术在对目标人物进行心率和呼吸的监测,利用心率和呼吸对目标人物的心理状态进行判断。同时,结合AI人工智能技术,获取整个面审过程中的心率和呼吸的异常时间点与目标人物回答问题的时间点的关系,以此为依据进行欺诈风险评估,能够给出更加直观可理解和信服的判断结果。
此外,本申请实施例还提出一种终端设备,该设备包括存储器、处理器以及存储在所述存储器中并可以在所述处理器上运行的计算机程序,所述处理器执行所述计算机程序时实现所述基于远程光体积描述术的风险评估方法 的步骤。
本申请之终端设备的具体实施方式与上述基于远程光体积描述术的风险评估方法的具体实施方式大致相同,在此不再赘述。
此外,本申请实施例还提出一种计算机可读存储介质,所述计算机可读存储介质可以是非易失性,也可以是易失性。所述计算机可读存储介质中包括基于远程光体积描述术的风险评估程序,所述基于远程光体积描述术的风险评估程序被处理器执行时实现如下操作:
基于光体积描述术的欺诈风险评估方法,应用于电子装置,其中,所述方法包括:
创建远程光体积描述术模型,并建立所述远程光体积描述术模型的用于实时视频流传输和反馈的***接口;
将待检测的视频流通过所述***接口实时传输至所述远程光体积描述术模型,基于所述远程光体积描述术模型采集所述视频流中的目标人物的心率和呼吸数据;
通过所述***接口实时反馈所述心率和呼吸数据至风险评估***,用于所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估。
优选地,所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估的步骤包括:
在所述风险评估***的显示界面中实时显示所述目标人物的心率和呼吸数据;
当所述心率或呼吸数据在预设时间内升高幅度超过设定值时,进一步判定所述心率或呼吸数据升高提示的时间点与所述风险评估***提出问题的时间点是否在预设邻近范围内;
当所述心率或呼吸数据升高提示的时间点与所述风险评估***提出问题的时间点在预设邻近范围内时,判定所述目标人物存在欺诈风险。
优选地,所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估的步骤还包括:
将所述目标人物在风险评估过程中的评估视频对话内容实时转换成文本数据,然后基于人工智能***自动判断所述目标人物回答问题的时间点,同时自动测算所述时间点的目标人物的心率和呼吸数据是否存在异常,并通过 所述人工智能***对所述评估视频的欺诈风险进行评估打分;
或者,预先采集所述评估视频中提问人员和所述目标人物的声纹信息,并基于声纹识别技术对所述提问人员和所述目标人物的说话时间进行切割,获取所述目标人物开始说话的时间点,同时自动测算所述时间点的目标人物的心率和呼吸数据是否存在异常,并通过所述人工智能***对所述评估视频的欺诈风险进行评估打分;
或者,对所述评估视频中的目标人物的面部嘴唇动作进行监测,并基于语音转文字或声纹识别技术,获取所述目标人物开始说话的时间点,同时自动测算所述时间点的目标人物的心率和呼吸数据是否存在异常,并通过所述人工智能***对所述评估视频的欺诈风险进行评估打分。
优选地,所述通过所述人工智能***对所述评估视频的欺诈风险进行评估打分的步骤包括:
设定所述心率或呼吸数据中残值较大的点、高杠杆点和强影响点作为初步异常点;
判断所述初步异常点与所述目标人物开始说话的时间点之间的时间差是否在设定范围内,若所述时间差在所述设定范围内,则确认所述初步异常点为确认异常点;
获取所述心率或呼吸数据的异常比例,并基于所述异常比例获取欺诈风险评分。
优选地,所述心率或呼吸数据的异常比例获取公式为:
Figure PCTCN2020087661-appb-000005
其中,a为心率或呼吸数据中的心率值点或呼吸值点,n为心率或呼吸数据的异常点,m为所述异常点n之前预设时间段内的平均心率值或呼吸值。
优选地,所述基于所述异常比例获取欺诈风险评分的公式为:
Figure PCTCN2020087661-appb-000006
其中,P为所述评估视频的欺诈风险评分,N为所述评估视频的心率或呼吸数据中的心率值点或呼吸值点的总和,∑a为所述评估视频的心率或呼吸数据的异常比例总和。
本申请之计算机可读存储介质的具体实施方式与上述基于远程光体积描述术的风险评估方法、电子装置的具体实施方式大致相同,在此不再赘述。
需要说明的是,在本文中,术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含,从而使得包括一系列要素的过程、装置、物品或者方法不仅包括那些要素,而且还包括没有明确列出的其他要素,或者是还包括为这种过程、装置、物品或者方法所固有的要素。在没有更多限制的情况下,由语句“包括一个……”限定的要素,并不排除在包括该要素的过程、装置、物品或者方法中还存在另外的相同要素。
上述本申请实施例序号仅仅为了描述,不代表实施例的优劣。通过以上的实施方式的描述,本领域的技术人员可以清楚地了解到上述实施例方法可借助软件加必需的通用硬件平台的方式来实现,当然也可以通过硬件,但很多情况下前者是更佳的实施方式。基于这样的理解,本申请的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来,该计算机软件产品存储在如上所述的一个存储介质(如ROM/RAM、磁碟、光盘)中,包括若干指令用以使得一台终端设备(可以是手机,计算机,服务器,或者网络设备等)执行本申请各个实施例所述的方法。
以上仅为本申请的优选实施例,并非因此限制本申请的专利范围,凡是利用本申请说明书及附图内容所作的等效结构或等效流程变换,或直接或间接运用在其他相关的技术领域,均同理包括在本申请的专利保护范围内。

Claims (20)

  1. 一种基于远程光体积描述术的风险评估方法,应用于电子装置,所述方法包括:
    创建远程光体积描述术模型,并建立所述远程光体积描述术模型的用于实时视频流传输和反馈的***接口;
    将待检测的视频流通过所述***接口实时传输至所述远程光体积描述术模型,基于所述远程光体积描述术模型采集所述视频流中的目标人物的心率和呼吸数据;
    通过所述***接口实时反馈所述心率和呼吸数据至风险评估***,用于所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估。
  2. 根据权利要求1所述的基于远程光体积描述术的风险评估方法,其中,所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估的步骤包括:
    在所述风险评估***的显示界面中实时显示所述目标人物的心率和呼吸数据;
    当所述心率或呼吸数据在预设时间内升高幅度超过设定值时,进一步判定所述心率或呼吸数据升高提示的时间点与所述风险评估***提出问题的时间点是否在预设邻近范围内;
    当所述心率或呼吸数据升高提示的时间点与所述风险评估***提出问题的时间点在预设邻近范围内时,判定所述目标人物存在欺诈风险。
  3. 根据权利要求1所述的基于远程光体积描述术的风险评估方法,其中,所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估的步骤还包括:
    将所述目标人物在风险评估过程中的评估视频对话内容实时转换成文本数据,然后基于人工智能***自动判断所述目标人物回答问题的时间点,同时自动测算所述时间点的目标人物的心率和呼吸数据是否存在异常,并通过所述人工智能***对所述评估视频的欺诈风险进行评估打分;
    或者,预先采集所述评估视频中提问人员和所述目标人物的声纹信息,并基于声纹识别技术对所述提问人员和所述目标人物的说话时间进行切割,获取所述目标人物开始说话的时间点,同时自动测算所述时间点的目标人物的心率和呼吸数据是否存在异常,并通过所述人工智能***对所述评估视频 的欺诈风险进行评估打分;
    或者,对所述评估视频中的目标人物的面部嘴唇动作进行监测,并基于语音转文字或声纹识别技术,获取所述目标人物开始说话的时间点,同时自动测算所述时间点的目标人物的心率和呼吸数据是否存在异常,并通过所述人工智能***对所述评估视频的欺诈风险进行评估打分。
  4. 根据权利要求3所述的基于远程光体积描述术的风险评估方法,其中,所述通过所述人工智能***对所述评估视频的欺诈风险进行评估打分的步骤包括:
    设定所述心率或呼吸数据中残值较大的点、高杠杆点和强影响点作为初步异常点;
    判断所述初步异常点与所述目标人物开始说话的时间点之间的时间差是否在设定范围内,若所述时间差在所述设定范围内,则确认所述初步异常点为确认异常点;
    获取所述心率或呼吸数据的异常比例,并基于所述异常比例获取欺诈风险评分。
  5. 根据权利要求4所述的基于远程光体积描述术的风险评估方法,其中,所述心率或呼吸数据的异常比例获取公式为:
    Figure PCTCN2020087661-appb-100001
    其中,a为心率或呼吸数据中的心率值点或呼吸值点,n为心率或呼吸数据的异常点,m为所述异常点n之前预设时间段内的平均心率值或呼吸值。
  6. 根据权利要求4所述的基于远程光体积描述术的风险评估方法,其中,所述基于所述异常比例获取欺诈风险评分的公式为:
    Figure PCTCN2020087661-appb-100002
    其中,P为所述评估视频的欺诈风险评分,N为所述评估视频的心率或呼吸数据中的心率值点或呼吸值点的总和,∑a为所述评估视频的心率或呼吸数据的异常比例总和。
  7. 一种电子装置,该电子装置包括:存储器及处理器,所述存储器中包括基于远程光体积描述术的风险评估程序,所述基于远程光体积描述术的风险评估程序被所述处理器执行时实现如下步骤:
    创建远程光体积描述术模型,并建立所述远程光体积描述术模型的用于 实时视频流传输和反馈的***接口;
    将待检测的视频流通过所述***接口实时传输至所述远程光体积描述术模型,基于所述远程光体积描述术模型采集所述视频流中的目标人物的心率和呼吸数据;
    通过所述***接口实时反馈所述心率和呼吸数据至风险评估***,用于所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估。
  8. 根据权利要求7所述的电子装置,其中,所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估的步骤包括:
    在所述风险评估***的显示界面中实时显示所述目标人物的心率和呼吸数据;
    当所述心率或呼吸数据在预设时间内升高幅度超过设定值时,进一步判定所述心率或呼吸数据升高提示的时间点与所述风险评估***提出问题的时间点是否在预设邻近范围内;
    当所述心率或呼吸数据升高提示的时间点与所述风险评估***提出问题的时间点在预设邻近范围内时,判定所述目标人物存在欺诈风险。
  9. 根据权利要求7所述的电子装置,其中,所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估的步骤还包括:
    将所述目标人物在风险评估过程中的评估视频对话内容实时转换成文本数据,然后基于人工智能***自动判断所述目标人物回答问题的时间点,同时自动测算所述时间点的目标人物的心率和呼吸数据是否存在异常,并通过所述人工智能***对所述评估视频的欺诈风险进行评估打分;
    或者,预先采集所述评估视频中提问人员和所述目标人物的声纹信息,并基于声纹识别技术对所述提问人员和所述目标人物的说话时间进行切割,获取所述目标人物开始说话的时间点,同时自动测算所述时间点的目标人物的心率和呼吸数据是否存在异常,并通过所述人工智能***对所述评估视频的欺诈风险进行评估打分;
    或者,对所述评估视频中的目标人物的面部嘴唇动作进行监测,并基于语音转文字或声纹识别技术,获取所述目标人物开始说话的时间点,同时自动测算所述时间点的目标人物的心率和呼吸数据是否存在异常,并通过所述人工智能***对所述评估视频的欺诈风险进行评估打分。
  10. 一种终端设备,该设备包括存储器、处理器以及存储在所述存储器中并在所述处理器上运行的计算机程序,所述处理器执行所述计算机程序时实现如下步骤:
    创建远程光体积描述术模型,并建立所述远程光体积描述术模型的用于实时视频流传输和反馈的***接口;
    将待检测的视频流通过所述***接口实时传输至所述远程光体积描述术模型,基于所述远程光体积描述术模型采集所述视频流中的目标人物的心率和呼吸数据;
    通过所述***接口实时反馈所述心率和呼吸数据至风险评估***,用于所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估。
  11. 根据权利要求10所述的终端设备,其中,所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估,包括:
    在所述风险评估***的显示界面中实时显示所述目标人物的心率和呼吸数据;
    当所述心率或呼吸数据在预设时间内升高幅度超过设定值时,进一步判定所述心率或呼吸数据升高提示的时间点与所述风险评估***提出问题的时间点是否在预设邻近范围内;
    当所述心率或呼吸数据升高提示的时间点与所述风险评估***提出问题的时间点在预设邻近范围内时,判定所述目标人物存在欺诈风险。
  12. 根据权利要求10所述的终端设备,其中,所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估,还包括:
    将所述目标人物在风险评估过程中的评估视频对话内容实时转换成文本数据,然后基于人工智能***自动判断所述目标人物回答问题的时间点,同时自动测算所述时间点的目标人物的心率和呼吸数据是否存在异常,并通过所述人工智能***对所述评估视频的欺诈风险进行评估打分;
    或者,预先采集所述评估视频中提问人员和所述目标人物的声纹信息,并基于声纹识别技术对所述提问人员和所述目标人物的说话时间进行切割,获取所述目标人物开始说话的时间点,同时自动测算所述时间点的目标人物的心率和呼吸数据是否存在异常,并通过所述人工智能***对所述评估视频的欺诈风险进行评估打分;
    或者,对所述评估视频中的目标人物的面部嘴唇动作进行监测,并基于语音转文字或声纹识别技术,获取所述目标人物开始说话的时间点,同时自动测算所述时间点的目标人物的心率和呼吸数据是否存在异常,并通过所述人工智能***对所述评估视频的欺诈风险进行评估打分。
  13. 根据权利要求12所述的终端设备,其中,所述通过所述人工智能***对所述评估视频的欺诈风险进行评估打分,包括:
    设定所述心率或呼吸数据中残值较大的点、高杠杆点和强影响点作为初步异常点;
    判断所述初步异常点与所述目标人物开始说话的时间点之间的时间差是否在设定范围内,若所述时间差在所述设定范围内,则确认所述初步异常点为确认异常点;
    获取所述心率或呼吸数据的异常比例,并基于所述异常比例获取欺诈风险评分。
  14. 根据权利要求13所述的终端设备,其中,所述心率或呼吸数据的异常比例获取公式为:
    Figure PCTCN2020087661-appb-100003
    其中,a为心率或呼吸数据中的心率值点或呼吸值点,n为心率或呼吸数据的异常点,m为所述异常点n之前预设时间段内的平均心率值或呼吸值。
  15. 一种计算机可读存储介质,所述计算机可读存储介质中包括基于远程光体积描述术的风险评估程序,所述基于远程光体积描述术的风险评估程序被处理器执行时,实现下步骤:
    创建远程光体积描述术模型,并建立所述远程光体积描述术模型的用于实时视频流传输和反馈的***接口;
    将待检测的视频流通过所述***接口实时传输至所述远程光体积描述术模型,基于所述远程光体积描述术模型采集所述视频流中的目标人物的心率和呼吸数据;
    通过所述***接口实时反馈所述心率和呼吸数据至风险评估***,用于所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估。
  16. 根据权利要求15所述的计算机可读存储介质,其中,所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估,包括:
    在所述风险评估***的显示界面中实时显示所述目标人物的心率和呼吸数据;
    当所述心率或呼吸数据在预设时间内升高幅度超过设定值时,进一步判定所述心率或呼吸数据升高提示的时间点与所述风险评估***提出问题的时间点是否在预设邻近范围内;
    当所述心率或呼吸数据升高提示的时间点与所述风险评估***提出问题的时间点在预设邻近范围内时,判定所述目标人物存在欺诈风险。
  17. 根据权利要求15所述的计算机可读存储介质,其中,所述风险评估***根据所述心率和呼吸数据进行欺诈风险评估,还包括:
    将所述目标人物在风险评估过程中的评估视频对话内容实时转换成文本数据,然后基于人工智能***自动判断所述目标人物回答问题的时间点,同时自动测算所述时间点的目标人物的心率和呼吸数据是否存在异常,并通过所述人工智能***对所述评估视频的欺诈风险进行评估打分;
    或者,预先采集所述评估视频中提问人员和所述目标人物的声纹信息,并基于声纹识别技术对所述提问人员和所述目标人物的说话时间进行切割,获取所述目标人物开始说话的时间点,同时自动测算所述时间点的目标人物的心率和呼吸数据是否存在异常,并通过所述人工智能***对所述评估视频的欺诈风险进行评估打分;
    或者,对所述评估视频中的目标人物的面部嘴唇动作进行监测,并基于语音转文字或声纹识别技术,获取所述目标人物开始说话的时间点,同时自动测算所述时间点的目标人物的心率和呼吸数据是否存在异常,并通过所述人工智能***对所述评估视频的欺诈风险进行评估打分。
  18. 根据权利要求17所述的计算机可读存储介质,其中,所述通过所述人工智能***对所述评估视频的欺诈风险进行评估打分,包括:
    设定所述心率或呼吸数据中残值较大的点、高杠杆点和强影响点作为初步异常点;
    判断所述初步异常点与所述目标人物开始说话的时间点之间的时间差是否在设定范围内,若所述时间差在所述设定范围内,则确认所述初步异常点为确认异常点;
    获取所述心率或呼吸数据的异常比例,并基于所述异常比例获取欺诈风 险评分。
  19. 根据权利要求18所述的计算机可读存储介质,其中,所述心率或呼吸数据的异常比例获取公式为:
    Figure PCTCN2020087661-appb-100004
    其中,a为心率或呼吸数据中的心率值点或呼吸值点,n为心率或呼吸数据的异常点,m为所述异常点n之前预设时间段内的平均心率值或呼吸值。
  20. 根据权利要求18所述的计算机可读存储介质,其中,所述基于所述异常比例获取欺诈风险评分的公式为:
    Figure PCTCN2020087661-appb-100005
    其中,P为所述评估视频的欺诈风险评分,N为所述评估视频的心率或呼吸数据中的心率值点或呼吸值点的总和,∑a为所述评估视频的心率或呼吸数据的异常比例总和。
PCT/CN2020/087661 2020-01-06 2020-04-29 基于远程光体积描述术的风险评估方法、装置、设备及存储介质 WO2021139053A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010009999.1 2020-01-06
CN202010009999.1A CN111192141A (zh) 2020-01-06 2020-01-06 基于远程光体积描述术的风险评估方法、装置及存储介质

Publications (1)

Publication Number Publication Date
WO2021139053A1 true WO2021139053A1 (zh) 2021-07-15

Family

ID=70708744

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/087661 WO2021139053A1 (zh) 2020-01-06 2020-04-29 基于远程光体积描述术的风险评估方法、装置、设备及存储介质

Country Status (2)

Country Link
CN (1) CN111192141A (zh)
WO (1) WO2021139053A1 (zh)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112907348A (zh) * 2021-01-14 2021-06-04 中国工商银行股份有限公司 移动终端、线上面审数据处理方法、装置及***
CN117438091B (zh) * 2023-12-20 2024-03-01 山西工程科技职业大学 一种基于大数据的运动强度检测***及方法

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102473238A (zh) * 2009-08-20 2012-05-23 皇家飞利浦电子股份有限公司 用于图像分析的***和方法
CN107004074A (zh) * 2014-12-23 2017-08-01 英特尔公司 使用高级传感器监测来进行增强型用户认证的技术
CN109325851A (zh) * 2018-09-29 2019-02-12 平安科技(深圳)有限公司 欺诈概率的展示方法、***、计算机设备及可读存储介质
CN110427803A (zh) * 2019-06-18 2019-11-08 平安科技(深圳)有限公司 基于视频分析的测谎方法、装置、电子设备及存储介质

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
AU2003217564A1 (en) * 2002-02-22 2003-09-09 Datex-Ohmeda, Inc. Monitoring physiological parameters based on variations in a photoplethysmographic signal
US20150025917A1 (en) * 2013-07-15 2015-01-22 Advanced Insurance Products & Services, Inc. System and method for determining an underwriting risk, risk score, or price of insurance using cognitive information
CN107704834B (zh) * 2017-10-13 2021-03-30 深圳壹账通智能科技有限公司 微表情面审辅助方法、装置及存储介质
US10993630B2 (en) * 2017-10-19 2021-05-04 Hill-Rom Services Pte. Ltd. Respiration rate estimation from a photoplethysmography signal
CN108734591A (zh) * 2018-04-09 2018-11-02 中国平安人寿保险股份有限公司 欺诈案件的评估方法、装置、存储介质及终端
CN108537186A (zh) * 2018-04-13 2018-09-14 深圳壹账通智能科技有限公司 一体机欺诈审核的方法、装置、设备及可读存储介质
CN110222554A (zh) * 2019-04-16 2019-09-10 深圳壹账通智能科技有限公司 欺诈识别方法、装置、电子设备及存储介质

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102473238A (zh) * 2009-08-20 2012-05-23 皇家飞利浦电子股份有限公司 用于图像分析的***和方法
CN107004074A (zh) * 2014-12-23 2017-08-01 英特尔公司 使用高级传感器监测来进行增强型用户认证的技术
CN109325851A (zh) * 2018-09-29 2019-02-12 平安科技(深圳)有限公司 欺诈概率的展示方法、***、计算机设备及可读存储介质
CN110427803A (zh) * 2019-06-18 2019-11-08 平安科技(深圳)有限公司 基于视频分析的测谎方法、装置、电子设备及存储介质

Also Published As

Publication number Publication date
CN111192141A (zh) 2020-05-22

Similar Documents

Publication Publication Date Title
WO2020211388A1 (zh) 基于预测模型的行为预测方法、装置、设备及存储介质
Luo et al. ARBEE: Towards automated recognition of bodily expression of emotion in the wild
CN107704834B (zh) 微表情面审辅助方法、装置及存储介质
Hossain et al. Cloud-assisted speech and face recognition framework for health monitoring
US10521503B2 (en) Authenticating a respondent to an electronic survey
US10426378B1 (en) Systems and methods for automated body mass index calculation to determine value
US10572778B1 (en) Machine-learning-based systems and methods for quality detection of digital input
US11033216B2 (en) Augmenting questionnaires
US12001434B2 (en) Method for monitoring user behavior when interacting with content and a system for its implementation
US11151385B2 (en) System and method for detecting deception in an audio-video response of a user
US11546182B2 (en) Methods and systems for managing meeting notes
US11521149B2 (en) Generating sentiment metrics using emoji selections
US20140278506A1 (en) Automatically evaluating and providing feedback on verbal communications from a healthcare provider
WO2021139053A1 (zh) 基于远程光体积描述术的风险评估方法、装置、设备及存储介质
US11483574B2 (en) Systems and methods for emotion and perception based video compression and video perception enhancement
WO2021179719A1 (zh) 人脸活体检测方法、装置、介质及电子设备
CN116762125A (zh) 环境协作智能***和方法
Spencer et al. Risk markers for male perpetration of sexual assault on college campuses: A meta-analysis
CN115170049A (zh) 可信度评分方法、装置、电子设备、存储介质和程序产品
TWI691923B (zh) 金融交易詐騙偵測防範系統及其方法
WO2019000051A1 (en) DATA ANALYSIS METHOD AND LEARNING SYSTEM
CN112990301A (zh) 情绪数据标注方法、装置、计算机设备和存储介质
CN114220055B (zh) 辅助用户面谈的方法、装置、计算机设备及存储介质
CN116033259B (zh) 生成短视频方法、装置、计算机设备及存储介质
Vinh et al. Advertisement system based on facial expression recognition and convolutional neural network

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20912287

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 07.11.2022)

122 Ep: pct application non-entry in european phase

Ref document number: 20912287

Country of ref document: EP

Kind code of ref document: A1