WO2022193635A1 - 客服服务系统、方法、装置、电子设备及存储介质 - Google Patents

客服服务系统、方法、装置、电子设备及存储介质 Download PDF

Info

Publication number
WO2022193635A1
WO2022193635A1 PCT/CN2021/123865 CN2021123865W WO2022193635A1 WO 2022193635 A1 WO2022193635 A1 WO 2022193635A1 CN 2021123865 W CN2021123865 W CN 2021123865W WO 2022193635 A1 WO2022193635 A1 WO 2022193635A1
Authority
WO
WIPO (PCT)
Prior art keywords
information
feedback
digital human
user
image
Prior art date
Application number
PCT/CN2021/123865
Other languages
English (en)
French (fr)
Inventor
刘致远
黄刚
杨国基
刘炫鹏
常向月
刘云峰
Original Assignee
深圳追一科技有限公司
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 深圳追一科技有限公司 filed Critical 深圳追一科技有限公司
Publication of WO2022193635A1 publication Critical patent/WO2022193635A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/40—Information retrieval; Database structures therefor; File system structures therefor of multimedia data, e.g. slideshows comprising image and additional audio data
    • G06F16/43—Querying
    • G06F16/438—Presentation of query results
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44—Arrangements for executing specific programs
    • G06F9/451—Execution arrangements for user interfaces
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION 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
    • G06Q30/00—Commerce
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T13/00—Animation
    • G06T13/20—3D [Three Dimensional] animation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T13/00—Animation
    • G06T13/20—3D [Three Dimensional] animation
    • G06T13/40—3D [Three Dimensional] animation of characters, e.g. humans, animals or virtual beings

Definitions

  • the present application relates to the technical field of human-computer interaction, and more particularly, to a customer service system, method, apparatus, electronic device and storage medium.
  • the embodiments of the present application provide a customer service system, method, apparatus, electronic device, and storage medium.
  • an embodiment of the present application provides a customer service system, where the customer service system includes a central control module, an information processing module, a business mode control module, and an image generation module.
  • the central control module is used to establish a call connection with the intelligent terminal.
  • the information processing module is used to obtain the call data between the central control module and the intelligent terminal, and determine the feedback information fed back to the intelligent terminal according to the call data.
  • the service mode control module is used for acquiring user information from the call data, and determining the feedback service mode according to the user information.
  • the image generation module is used to obtain the digital human image including the digital human based on the feedback information if the feedback business mode is the digital human online mode, and send the digital human image to the central control module as the interactive feedback result;
  • the human form corresponds to the feedback information, and the central control module is also used to send the interactive feedback result to the intelligent terminal.
  • an embodiment of the present application provides a customer service method, which is performed by an electronic device.
  • the customer service method includes: establishing a call connection with an intelligent terminal; acquiring call data with the intelligent terminal, and determining a destination according to the call data. Feedback information fed back by the intelligent terminal; user information is obtained from the call data, and the feedback service mode is determined according to the user information; if the feedback service mode is the digital human online mode, the digital human image including the digital human is obtained based on the feedback information, and the digital human is obtained.
  • the human image is used as the interactive feedback result; wherein, the shape of the digital human in the digital human image corresponds to the feedback information; and the interactive feedback result is sent to the intelligent terminal.
  • an embodiment of the present application provides a customer service device, which is set in an electronic device, and the customer service device includes a connection module, a feedback information determination module, a business mode determination module, a digital human drive module, and an interactive feedback result sending module .
  • the connection module is used to establish a call connection with the intelligent terminal.
  • the feedback information determination module is used for acquiring the call data with the intelligent terminal, and determining the feedback information fed back to the intelligent terminal according to the call data.
  • the service mode determination module is used for acquiring user information from the call data, and determining the feedback service mode according to the user information.
  • the digital human driving module is used to obtain the digital human image including the digital human based on the feedback information if the feedback business mode is the digital human online mode, and use the digital human image as the interactive feedback result; wherein, the digital human image in the digital human image is the same as that of the digital human.
  • Corresponding feedback information is used for sending the interactive feedback result to the intelligent terminal.
  • embodiments of the present application provide an electronic device, the electronic device comprising: one or more processors; a memory; one or more computer-readable instructions, wherein the one or more computer-readable instructions are stored in a In the memory and configured to be executed by the one or more processors, the one or more computer readable instructions are configured to perform the steps of the customer service method provided by the first aspect.
  • the embodiments of the present application provide one or more computer-readable storage media, where the computer-readable storage media stores computer-readable instructions, and the computer-readable instructions are executed by a processor to perform the steps of the customer service method provided in the second aspect .
  • FIG. 1 shows a schematic diagram of an application environment of a customer service system provided by an embodiment of the present application.
  • FIG. 2 shows a schematic diagram of a module of a customer service system provided by an embodiment of the present application.
  • FIG. 3 is a schematic diagram of another module of the customer service system provided by the embodiment of the present application.
  • FIG. 4 shows a schematic flowchart of a customer service method provided by an embodiment of the present application.
  • FIG. 5 shows a schematic flowchart of establishing a call connection in an embodiment.
  • FIG. 6 shows another schematic flowchart of establishing a call connection in one embodiment.
  • FIG. 7 shows a schematic flowchart of obtaining feedback information in one embodiment.
  • FIG. 8 shows a schematic flowchart of determining the online mode of a digital human in one embodiment.
  • FIG. 9 shows a schematic flow chart of acquiring a digital human image in one embodiment.
  • FIG. 10 shows a schematic flowchart of sending interactive feedback results in one embodiment.
  • FIG. 11 shows a block diagram of functional modules of a customer service apparatus provided by an embodiment of the present application.
  • FIG. 12 shows a block diagram of functional modules of an electronic device proposed by an embodiment of the present application.
  • the traditional PSTN call method has been gradually replaced by the high-definition call method, such as the VOLTE (Voice over Long-Term Evolution) call method.
  • the high-definition call method Compared with the PSTN call method, the high-definition call method.
  • the voice and video calls of the call mode are clearer and the drop rate is lower.
  • the data service can be used at the same time.
  • the video call can be realized without the use of programs and software installed on the smart terminal, so that the call experience of both parties can be improved. better.
  • customer service personnel provide video customer service services to users through video interaction.
  • the intelligent terminal held by the customer service personnel can send the audio and video information of the customer service personnel to the users held by the user.
  • the user can feel more cordial and improve the user's consulting experience with the customer service personnel.
  • this method can effectively improve the user experience, it is easy to have a shortage of customer service personnel during the peak consultation period, which leads to a long waiting time for the user and a poor user experience.
  • the inventor of the present application invests in research and development, and the inventor finds that the video can be recorded in advance.
  • the pre-recorded video will be sent to the user's holding based on the demand.
  • the smart terminal can realize that during the peak consultation period, the screen displayed by the smart terminal held by the user is always the video of the customer service staff.
  • this method will cause the connection between the recorded videos to be unsmooth, and the displayed content is relatively simple. , resulting in a more blunt picture of the customer service staff that the user sees.
  • the inventor of the present application continues to invest in research and development.
  • the customer service system includes a central control module, an information processing module, a business mode control module and an image generation module.
  • the central control module is used to establish a call connection with the intelligent terminal.
  • the information processing module is used to obtain the call data between the central control module and the intelligent terminal, and determine the feedback information fed back to the intelligent terminal according to the call data.
  • the business mode control module is used to obtain user information from the call data, and determine the feedback business mode according to the user information.
  • the feedback business mode is used to represent the business mode of sending feedback information to the intelligent terminal.
  • the online mode is used to represent the mode in which the picture containing the digital human is presented on the smart terminal.
  • the image generation module is used to obtain the digital human image including the digital human based on the feedback information if the feedback business mode is the digital human online mode, and send the digital human image to the central control module as the interactive feedback result;
  • the human form corresponds to the feedback information
  • the central control module is also used to send the interactive feedback result to the intelligent terminal.
  • the feedback information obtained from the call data obtains the digital human image, so that the digital human is displayed on the intelligent terminal.
  • the customer service system When the customer service system is applied to the customer service service, it can effectively improve the intimacy of the user, and at the same time can replace the manual service, when in the peak consultation period At the same time, it can quickly respond to the user's consultation needs, and at the same time, considering the user's needs, select the corresponding feedback business mode to feed back information to the intelligent terminal, and improve the diversity of customer service forms to meet the needs of different users.
  • FIG. 1 shows a schematic diagram of an application environment suitable for the customer service system provided by the embodiment of the present application.
  • the customer service system, method, apparatus, electronic device, and storage medium provided by the embodiments of the present application can be applied to the interactive system 300 shown in FIG. 1 .
  • the interactive system 300 includes an intelligent terminal 301 and a server 302 , and the server 302 is connected to the intelligent terminal 301 in communication.
  • the server 302 may be implemented by an independent server or a server cluster composed of multiple servers.
  • the server may be a cloud server or a traditional computer room server, which is not specifically limited here.
  • the smart terminal 301 may be various electronic devices having a display screen and supporting data input, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, wearable electronic devices, and the like.
  • the data input can be based on the voice input module configured by the smart terminal 301 inputting voice, the character input module inputting characters, the image input module inputting images, the video input module inputting video, etc., or based on the gesture recognition configured by the smart terminal 301.
  • the module allows users to implement gesture input and other interactive methods.
  • the smart terminal 301 can be installed with a client application, and the user can communicate with the server 302 based on the client application (for example, an APP such as WeChat, a WeChat applet, etc.).
  • the server 302 has a corresponding Server application, the user can register a user account in the server 302 based on the client application, and communicate with the server 302 based on the user account, for example, the user logs in the user account in the client application, and based on the user account
  • the client application program can input text information, voice information, image information or video information, etc. After the client application program receives the information input by the user, it can send the information to the server 302 so that the server 302 can receive the information.
  • the information is processed and stored, and the server 302 can also receive the information and return the corresponding output information to the smart terminal 301 according to the information.
  • the smart terminal 301 can receive the dialing call request from the server 302 , and can also send the dialing call request to the server 302 .
  • the server 302 can be configured to receive the information input by the user, generate a picture of the simulated digital human according to the information, and send the picture to the smart terminal 301, so as to provide customer service to the smart terminal 301, and communicate with the user. Communicate with customer service.
  • the smart terminal 301 can receive the information input by the user, and present the screen of the simulated digital human sent by the server 302 to the smart terminal 301 .
  • the simulated digital human is a software program based on visual graphics, and after the software program is executed, it can present a robot form that simulates biological behavior or thought to the user.
  • the simulated digital human can be a simulated human-like simulated digital human, such as a human-like simulated digital human created based on the user's own or other natural person's form, or a simulated digital human with an animation effect, such as an animal-shaped or cartoon-shaped human-like digital human. Simulate a digital human.
  • the server 302 can receive the dial-up call request from the smart terminal 301, and can also send the dial-up call request to the smart terminal.
  • the smart terminal 301 may display the reply corresponding to the reply on the display screen of the smart terminal 301 or other image output devices connected thereto.
  • a simulated digital human image for information.
  • the audio corresponding to the digital human image can be played through the speaker of the smart terminal 301 or other audio output devices connected to it, and the display screen of the smart terminal 301 can also display the audio corresponding to the digital human image.
  • the text or graphics corresponding to the reply information realizes the polymorphic interaction with the user in terms of image, voice, text and other aspects.
  • an embodiment of the present application provides a customer service system 400 .
  • the customer service system 400 can be deployed on the server 302 of the interactive system 300 shown in FIG. 1 .
  • the customer service system 400 includes a central control module 410 , Information processing module 420 , business mode control module 430 and image generation module 440 .
  • the central control module 410 is used to establish a call connection with the intelligent terminal.
  • the information processing module 420 is configured to acquire call data between the central control module 410 and the smart terminal, and determine feedback information fed back to the smart terminal according to the call data.
  • the service mode control module 430 is configured to acquire user information from the call data, and determine a feedback service mode according to the user information.
  • the image generation module 440 is configured to obtain the digital human image including the digital human based on the feedback information if the feedback service mode is the digital human online mode, and send the digital human image to the central control module 410 as the interactive feedback result; wherein, the digital human image
  • the shape of the digital human corresponds to the feedback information, and the central control module 410 is also used for sending the interactive feedback result to the intelligent terminal.
  • the feedback service mode may be used to represent the service mode in which the customer service system 400 sends feedback information
  • the feedback service mode includes a digital human online mode
  • the digital human online mode is used to represent a situation where a picture including a digital human is presented on the smart terminal.
  • the digital human online mode may include a mode of feeding back interactive feedback results including digital human images to the intelligent terminal.
  • the intelligent terminal establishes a call connection with the customer service system 400 provided in this embodiment, and the customer service system 400 can feed back an interactive feedback result to the intelligent terminal, and the interactive feedback result can include continuity Therefore, the intelligent terminal can present the multiple digital human images, so as to achieve the effect of the digital human providing customer service for the user, without the need to specially arrange customer service staff to provide customer service for the customer, and to avoid the labor tension of the customer service staff. Realize rapid response to user's consulting needs.
  • the central control module 410 can establish a call connection with the smart terminal.
  • the communication method between the central control module 410 and the smart terminal may be a high-definition call, that is, the server 302 deployed with the customer service system 400 can make a high-definition call with the smart terminal, and the high-definition call process does not need to rely on the smart terminal.
  • An application installed on the device for example, the application may include software with video calling capabilities).
  • high-definition calls include VOLTE (Voice over Long-Term Evolution, long-term evolution voice bearer), smooth calls, and the like.
  • the central control module 410 may directly establish a call connection with the smart terminal, or may establish a call connection with the smart terminal indirectly, and there is no specific limitation on the manner in which the central control module 410 establishes a call connection with the smart terminal.
  • the central control module 410 when the central control module 410 directly establishes a call connection with the smart terminal, the central control module 410 may have a calling function and/or an answering function. Specifically, the central control module 410 can send a call request to the smart terminal through the SIM card number. After the smart terminal accepts the call request, the central control module 410 establishes a high-definition call connection with the smart terminal. Obtain user information, and send the digital human image to the smart terminal; the smart terminal can send a call request to the central control module 410, and after the central control module 410 accepts the call request, the central control module 410 establishes a high-definition call connection with the smart terminal.
  • a call platform and a SIP (Session Initiation Protocol) server may be set between the communication link formed by the central control module 410 and the smart terminal .
  • the calling platform can send a call request to the smart terminal through the SIM card number. After the smart terminal accepts the call request, the calling platform establishes a high-definition call connection with the smart terminal.
  • the -time Transport Protocol establishes a call connection with the SIP server
  • the SIP server establishes a data connection with the central control module 410
  • the call platform and the SIP server can be used as the transmission medium between the intelligent terminal and the central control module 410
  • the RTP protocol is used as data transmission.
  • the call platform can receive the call request sent by the smart terminal through the SIM card number. After the call request sent by the smart terminal is allowed, the call platform can establish a high-definition call connection with the smart terminal.
  • the call platform can establish a call connection with the SIP server through the RTP protocol.
  • the SIP server establishes a data connection with the central control module 410
  • the call platform and the SIP server can be used as the transmission medium between the intelligent terminal and the central control module 410
  • the RTP protocol is used as the data transmission rule.
  • the call data may include information sent by the intelligent terminal received by the central control module 410 and information sent to the intelligent terminal.
  • the information may be the previous call records between the intelligent terminal and the central control module 410, or may be information sent by the intelligent terminal to the central control module 410 in real time.
  • the server can pre-store the previous call record between the smart terminal and the central control module 410, and when the call data needs to be acquired, the call record can be extracted from the server.
  • the feedback information may include information determined to be fed back to the user based on the user's intention.
  • the user's intention can be determined based on the call data including the call record and the information sent by the smart terminal received in real time, and then the information fed back to the user can be determined.
  • the call record and the data sent by the smart terminal in real time can be combined, and the user's intention can be determined based on the combined information. For example, based on the call records, it is determined that the packages the user wants to apply for are package A and package B, and the packages that he does not want to apply include package C and package D.
  • the real-time data sent by the smart terminal pole includes package processing, it can be determined that the user wants to apply for the package. A or Package B, and take that as the user's intent.
  • the service mode control module 430 may acquire user information from the call data.
  • User information may include information related to the user.
  • the user information may be the information entered by the user on the smart terminal 301, the user's voice, the user's image, etc.
  • the type of the user information is not specifically limited here.
  • the information when the user information is the information entered by the user in the smart terminal 301, the information can be matched with the pre-stored information, and when the matching is successful, the service mode corresponding to the pre-stored information can be used as the feedback service mode.
  • the information entered by the user in the smart terminal 301 may be symbols, numbers, and the like.
  • the voiceprint feature of the voice can be extracted, and the voiceprint feature can be matched with the pre-stored voiceprint feature of the user.
  • the business model corresponding to the pattern feature is used as the feedback business model.
  • the user information is the user's image
  • the feature of the image can be extracted, and the feature can be matched with the pre-stored user's feature.
  • the business model corresponding to the pre-stored user's feature can be used as the feedback service. model.
  • the business model required by the user may be stored in advance, and the business model may be associated with the user.
  • the business model associated with the user may be used as the feedback business model.
  • the feedback business mode may include a digital human online mode.
  • the digital human image can be sent to the intelligent terminal, and the intelligent terminal can play a continuous number of digital human images, so as to achieve the digital human The effect of customer service provided by users.
  • the feedback information may include driving information
  • the driving information may be used to drive the digital human.
  • driving information for driving the digital human can be acquired based on the feedback information, the digital human is driven by the driving information, and an image of the digital human is acquired.
  • the form of the driving information may include text, voice, semantics, etc., and the type of the driving information is not specifically limited here.
  • the form of the digital human may include the action state of the digital human.
  • the shape of the digital human may be the shape of the digital human's eyeball, mouth shape, expression, arms, and the like.
  • the driving information may be a continuous parameter with a time attribute
  • the movement of the digital human is controlled according to the time sequence
  • the driving information may be associated with the action state of the digital human.
  • the driving information may include controlling the 3D digital human head to rise 15 degrees at time t1, controlling the 3D digital human head to swing left and right at time t2, and controlling the 3D digital human to smile at time t3.
  • the digital human image can be understood as an image obtained by photographing the digital human at a certain angle, and the driving information of the digital human in the digital human image corresponds to the action state.
  • the digital human in this embodiment can be obtained by 3D modeling, or can be generated by using a deep learning model to generate a picture quality of each frame is close to a realistic image captured by a camera, and the digital human is like a real person captured by a camera.
  • a video digital human can be generated from coherent photorealistic images.
  • the method for obtaining a digital human by 3D modeling may include: acquiring a plurality of sample images including the target model; acquiring the shape of the target model according to the plurality of sample images; acquiring the original digital human and modeling information, where the modeling information includes The original key points of the original digital human; the target key points of the target model are obtained according to the shape, and the original key points are corresponding to the target key points to generate a digital human.
  • the target mannequins may include mannequins associated with digital people.
  • the target model can be the broadcast host C, or a person with a similar face, skeleton and body to the broadcast host C, or a face similar to the broadcast host C. , skeletons, dummies of similar stature (eg wax figures).
  • the sample images may include multiple images of the target model from different angles.
  • the sample images may be images of the target model under various actions, pronunciations, expressions, etc., from various angles.
  • the cameras for capturing images of the target model can be arranged in a spherical shape around the target model, wherein, relative to the same orientation of the target model, cameras with different focal lengths can be arranged.
  • each camera device can be used to collect images including the target model at the same time, so as to obtain a sample image.
  • the shape of the target model may include information related to the body change of the target model.
  • the shape can be a drooping corner of the mouth, a right eye deviation, a raised head, a raised right hand, and so on.
  • each part of the target model can be obtained from the sample image through the target detection algorithm, and then the shape of the part can be determined based on the change state of the same part in multiple consecutive sample images, so as to obtain each part of the target model Shape.
  • the target detection algorithm may be a sliding window target detection algorithm, a two-stage target detection algorithm, a one-stage target detection algorithm, and other algorithms.
  • the original digital human may include an already constructed digital human model.
  • the original digital human may be an average face model in a certain region, or may be a 3D animation model in an industrial animation, and the type of the original digital human is not specifically limited here.
  • the modeling information may include parameter information for constructing the original digital human, through which the original digital human can be restored, so that the original digital human can be presented.
  • the shape of the target model can be combined with the modeling information, so that the shape features of the target model are added to the original digital human, so as to obtain a digital human whose shape is approximately the same as that of the target model.
  • the original key points of the original digital human may include the positions used to identify, locate and control various parts of the original digital human.
  • the original key points may be the left eye corner position, right eye corner position, mouth corner position, facial contour position, eyebrow position, nose wing position, thumb position, shoulder position, etc. of the original digital human. It should be noted that the denser the original key points in the modeling information are in various parts of the original digital human, the more accurate the final digital human is.
  • the target key points may include positions for identifying and locating various parts of the target model.
  • the target key points may be the left eye corner position, right eye corner position, mouth corner position, facial contour position, eyebrow position, nose wing position, thumb position, shoulder position, etc. of the target model. It should be noted that the denser the target key points are in various parts of the target model, the more accurate the digital human being finally constructed.
  • the original key points of each part of the original digital person can be in a one-to-one correspondence with the target key points of the same part of the target model.
  • the original key points include various positions of the facial contour of the original digital person
  • the target key points include various positions of the facial contour of the target model.
  • the positions of the lower court of the face correspond one-to-one
  • the other original key points of the original digital human face correspond to other target key points of the target model's face one-to-one in the same way, and will not be repeated here.
  • target keypoints in motion can be mapped to original keypoints.
  • the target key points of the target model can be obtained, the face of the target model in the continuous sample images can be marked, and the same target key points of the face of the target model can be associated according to the time sequence of the continuous sample images, so as to obtain the face of the target model.
  • the dynamic change trajectories of each target key point of The amplitude difference between the amplitude of the dynamic change trajectory of the original key point and the dynamic change trajectory of the target key point. If the amplitude difference is greater than the preset amplitude threshold, it is determined that the original key point of the original digital human needs to be corrected.
  • the target key points of the target model can be obtained, the face of the target model in the continuous sample images can be marked, and the same target key points of the face of the target model can be associated according to the time sequence of the continuous sample images, so as to obtain the face of the target model.
  • the dynamic change trajectory of each target key point, and the dynamic change trajectory of each target key point corresponds to the original key point of the original digital human face, so as to obtain the dynamic change trajectory of the original key point of the original digital human face, based on the original key point
  • the dynamic change trajectory of the point is used to obtain the facial change amplitude of the original digital person at each moment, and the facial change amplitude of the target model at each moment is obtained based on the dynamic change trajectory of the target key point, and the facial change amplitude of the original digital person at each moment is compared. and the face change range of the target model at each moment.
  • the difference between the face change range of the original digital person and the target model’s face change range at the same time is greater than the preset range threshold, it is determined that the original key points of the original digital person need to be checked. Correction; if the difference between the face change amplitude of the original digital person and the target model's face change amplitude at the same moment is less than or equal to the preset amplitude threshold, it is determined that the face construction of the current digital person meets expectations.
  • a simulated digital human model can also be constructed in advance, and each frame of a realistic image with a picture quality close to that taken by a camera can be generated by simulating the digital human.
  • the construction of the simulated digital human generation model can include: acquiring a plurality of sample images including the target model and camera parameters corresponding to each sample image; obtain sample image configuration parameters corresponding to camera parameters; obtain the angle information of the target model according to the sample image, and associate the angle information with the sample image configuration parameters; configure parameters and angle information according to the sample image
  • a simulated digital human model is constructed, and a preset simulated digital human model is obtained.
  • the camera parameters may include parameters used by the photographing device for photographing the sample image when photographing the target model.
  • camera parameters can be focal length, aperture size, and the like.
  • the sample image configuration parameters may include parameters of the sample image generated by the photographing device for photographing the sample image by photographing the target model.
  • the sample image configuration parameters can be pixel size, image exposure, the proportion of the target model in the image, the position where the target model contacts the ground, and so on.
  • the angle information may include the angle at which the target model appears in the sample image. For example, when the included angle between the face orientation of the target model and the preset axis direction in the sample image is 15 degrees, 15 degrees may be used as the angle information.
  • the sample image may be identified to obtain the angle of the target model.
  • each part of the target model can be obtained from the sample image through the target detection algorithm, and then the angle of the part can be determined based on the change state of the same part in multiple consecutive sample images, so as to obtain the angle of each part of the target model , and use the angle of each part as the angle information.
  • the sample image can be regarded as consisting of multiple regions and multiple points, and the state of multiple regions and multiple points of the target model at various angles is obtained based on the configuration parameters of the sample image and the angle information. Areas and points at various angles are combined to construct a simulated digital human model, so that the simulated digital human model can output images including the target model at different angles.
  • the process of constructing a simulated digital human generation model does not require 3D modeling, and the obtained simulated digital human is also closer to a real model, and the effect is more realistic. It may be necessary to model different real mannequins to obtain a simulated digital human condition.
  • the central control module 410 can send the interactive feedback result to the intelligent terminal through the communication link between it and the intelligent terminal, and the intelligent terminal can display the digital human image in the interactive feedback result.
  • the digital human image includes continuous image frames
  • the effect of digital human video can be presented on the smart terminal.
  • the central control module 410 can form a video call with the intelligent terminal, and from the call
  • the user information is obtained from the data, and the feedback service mode is determined based on the user information. If the feedback service mode is the online digital human mode, the digital human image is obtained through the feedback information obtained based on the call data, so that the digital human is displayed on the smart terminal.
  • the customer service system 400 is applied to customer service, it can effectively improve the intimacy of users, and at the same time, it can replace manual services. When it is in the peak period of consultation, it can quickly respond to the consultation needs of users. Feedback business model Feedback information to intelligent terminals, improve the diversity of customer service forms, so as to meet the needs of different users.
  • the information processing module 420 may include a speech recognition unit 421 and a text feedback unit 422 .
  • the voice recognition unit 421 is configured to acquire the user's voice from the call data, and recognize the user's voice to obtain text input information.
  • the text feedback unit 422 is configured to determine the user's intention according to the text input information, and obtain text output information based on the user's intention; the text output information is used to determine feedback information.
  • the feedback information may include at least one of text output information, feedback voice information, and the like.
  • textual output information may be directly used as feedback information.
  • Text output information is feedback information in text format.
  • speech synthesis may be performed based on the text output information to generate feedback speech information, and the feedback speech information may be used as the feedback information.
  • the feedback voice information refers to feedback information in a voice format.
  • the text output information and the feedback voice information may also be used as the feedback information.
  • the speech recognition unit 421 and the semantic unit 422 may form a communication connection with the central control module 410 respectively, and the speech recognition unit 421 may convert the speech into text input information through ASR (Automatic Speech Recognition) technology , and send the text input information to the central control module 410, the central control module 410 can receive the text input information, and send the text input information to the semantic unit 422, and the semantic unit 422 determines the user's intention according to the text input information, and Get text output information for feedback to the user.
  • ASR Automatic Speech Recognition
  • the speech recognition unit 421 can recognize the user's speech information by means of linguistics and acoustics, stochastic model, artificial neural network, probabilistic grammar analysis, etc.
  • the method of user's speech information is not specifically limited here.
  • an intention recognition model and a feedback platform can be set in the semantic unit 422, and the intention recognition model can perform intention recognition on the text input information to obtain the user's intention, and then transmit the intention to the feedback platform to obtain the text output information fed back to the user.
  • an intent identification model may be used to identify the user information, and the intent identification model may use a machine learning model such as an RNN model, a CNN model, a VAE model, a BERT, and a support vector machine, which is not limited herein.
  • the intent recognition model may also be a variant or combination of the above-mentioned machine learning model.
  • the intent recognition model can be deployed on a smart terminal or a server. The deployment carrier of the intent recognition model is not specifically limited here, and is specifically determined by the actual scenario of obtaining user intent in this embodiment.
  • the feedback platform may include a question and answer library, a customer service vocabulary library, etc. constructed based on the needs of the customer service business type, and the construction of the feedback platform is not specifically limited here.
  • the shape of the digital human can correspond to the text one-to-one.
  • the digital human can be driven by the text in the text output information, so that the shape of the digital human in the output digital human image can be matched with the text output. information corresponds.
  • the user's speech can be recognized, text input information can be obtained, and text output for feedback to the user can be obtained based on the text input information.
  • information and drive the digital human through the text output information, and present the content fed back to the user in the form of a digital human, so as to simulate the scene of face-to-face communication between the user and the customer service staff, and improve the user experience.
  • the feedback information includes feedback speech information; as shown in FIG. 3 , the information processing module 420 may include a speech recognition unit 421 , a text feedback unit 422 and a speech synthesis unit 423 .
  • the voice recognition unit 421 is configured to acquire the user's voice from the call data, and recognize the user's voice to obtain text input information.
  • the text feedback unit 422 is configured to determine the user's intention according to the text input information, and obtain text output information based on the user's intention.
  • the speech synthesis unit 423 is configured to perform speech synthesis processing on the text output information to obtain feedback speech information.
  • the speech synthesis unit 423 can form a communication connection with the central control module 410, and the speech recognition unit 421 can convert the speech into text input information through the ASR technology, and send the text input information to the central control module 410.
  • the control module 410 can receive the text input information, and send the text input information to the semantic unit 422.
  • the semantic unit 422 determines the user's intention according to the text input information, obtains the text output information fed back to the user, and sends the text output information to The central control module 410, the central control module 410 sends the text output information to the speech synthesis unit 423, and the speech synthesis unit 423 generates feedback speech information from the text output information through the TTS (text to speech, Text To Speech) technology.
  • TTS text to speech, Text To Speech
  • the speech synthesis unit 423 can synthesize the text output information by a parameter synthesis method, a waveform synthesis method, a rule synthesis method, etc. to obtain feedback speech information.
  • the shape of the digital human can correspond to the voice one by one.
  • the digital human can be driven by the voice in the voice feedback information, so that the shape of the digital human in the output digital human image and the voice feedback information corresponds. That is, the feedback voice information is used to drive the digital human in the digital human image, so that the form of the digital human in the digital human image corresponds to the feedback voice information.
  • the mouth shape parameters of the digital human can be controlled by voice.
  • the changes of the key points of the mouth of the digital human can be obtained when the digital human emits the sound corresponding to the feedback voice information, so as to obtain the mouth shape parameters used to represent the changes of the key points of the mouth.
  • the key points of the mouth may include the positions for identifying, locating and controlling the various parts of the mouth of the digital human.
  • the key points of the mouth may include the left corner of the mouth, the right corner of the mouth, the chin-labial fold, the bottom of the nose, and so on.
  • the feedback voice information can have a time attribute, and the mouth shape changes of the digital human can be controlled at each time node corresponding to the voice in the feedback voice information, so that the intelligent terminal can accurately Present the process of changing the mouth shape of the digital human.
  • the user's speech can be recognized to obtain text input information, and based on the text input information,
  • the text output information fed back to the user, and through the text output information, the text output information is converted into feedback voice information, and the digital human is driven by the feedback voice information, so as to simulate the scene of face-to-face communication between the user and the customer service personnel, and improve the user experience.
  • an electronic card can be generated while generating the digital human image.
  • the image generating module 440 can include a card unit 441, and the card unit 441 is used to obtain prompts according to the feedback information information, generate an electronic card including prompt information, and send the electronic card to the central control module 410, so that the central control module 410 sends the electronic card as an interactive feedback result to the smart terminal.
  • the electronic card can be used to display prompt content (such as service packages, package fees, etc.).
  • prompt content such as service packages, package fees, etc.
  • the electronic card when the driving information is that the packages that the user can handle include package A, package B, and package C, the prompt content can be package A, package B, and package C arranged in a preset format, and the prompt content can be The content is presented on the electronic card.
  • the card type of the electronic card may include static type and dynamic type.
  • the card unit 441 can generate an electronic card and make the smart terminal present the electronic card, so as to obtain the effect that the electronic card is in a static state;
  • the electronic card is of the dynamic type, the card unit 441 can Multiple electronic cards are generated, and the multiple electronic cards can be continuous electronic cards.
  • the smart terminal presents multiple electronic cards according to the sequence of the electronic cards, the effect of changing the video of the electronic cards can be presented.
  • the electronic card can be combined with the digital human image and sent to the smart terminal as an interactive feedback result, or can be independently sent to the smart terminal as a feedback result.
  • the electronic card and the digital human image can be overlapped, so that the electronic card covers the non-core area of the digital human image, and the overlapping form is formed.
  • the image is sent to the smart terminal as the interactive feedback result.
  • the non-core area may be an area other than the digital human in the digital human image, and may also be an area other than the digital human head in the digital human image.
  • the electronic card and the digital human image are fused to form an image including the electronic card and the digital human, in the process of transmitting the image to the smart terminal in the form of a data packet, it is not necessary to transmit the electronic card and the digital human image separately.
  • the smart terminal presents interactive feedback results, it can directly display images including electronic cards and digital people.
  • the electronic card and the digital human image can be respectively transmitted to the smart terminal in the form of data packets, and when the smart terminal receives the electronic card and the digital human image, it can The electronic card is used as the first layer, and the digital human image is used as the second layer.
  • the smart terminal needs to display the first layer and the second layer at the same time, the first layer can be placed on the top layer, and the second layer can be placed on the top layer. placed on the bottom layer.
  • the electronic card can be generated by the electronic card generating unit 131, and the customer service system 400 can send the digital human image to the intelligent terminal and generate the electronic card at the same time, so that the electronic card can be generated.
  • the user can view the prompt content of the electronic card at the same time, helping the user to better understand and handle the business.
  • the feedback service mode may further include a digital human offline mode, and the digital human offline mode is used to represent a mode in which a picture that does not include a digital human is presented on the intelligent terminal;
  • the image generation module 440 is further configured to generate a recommendation image from the feedback information according to a preset information arrangement rule if the feedback service mode is the digital human offline mode, and send the recommendation image to the central control module 410; wherein the central control module 410 It is also used to send the recommended image to the intelligent terminal as the interactive feedback result.
  • the offline mode of the digital human may include a mode in which the interactive feedback result including the recommended image is fed back to the intelligent terminal.
  • the intelligent terminal establishes a call connection with the customer service system 400 provided in this embodiment, and the customer service system 400 can feed back an interactive feedback result to the intelligent terminal, and the interactive feedback result can include continuity
  • the intelligent terminal can present the multiple promotion images, and the screen presented by the intelligent terminal does not contain digital people.
  • the information in the promotion images is used to provide users with customer service services, so that the intelligent terminal can directly present the images.
  • the screen containing the feedback information allows the user to obtain the information fed back by the customer service system 400 more intuitively, which is convenient for the hearing impaired user to obtain the information, so as to provide various customer service services for the user.
  • the feedback service type involved in the feedback information may be obtained first, an arrangement template may be obtained according to the feedback service type, and the feedback information may be arranged in the arrangement template, thereby obtaining a recommendation image containing the feedback information.
  • the feedback service type may be used to represent the type of feedback content in the feedback information.
  • the feedback business type may include consumption flow, available packages, remaining usage of packages, preferential activities, etc.
  • the layout template may include templates for arranging feedback information of various feedback service types.
  • the feedback content in the feedback information is obtained
  • the feedback business types involved in obtaining the feedback content include consumption flow and preferential activities
  • an arrangement template capable of arranging the feedback business types including consumption flow and preferential activities is obtained, which will be combined with the consumption flow.
  • the feedback information corresponding to the running water is arranged in the first area in the layout template.
  • the first area is used to arrange the feedback information whose feedback business type is consumption running water, and arrange the feedback information corresponding to the preferential activities in the layout template.
  • the second area is used for arranging the feedback information that the feedback service type is preferential activity. That is to say, feedback information of various feedback service types can be arranged in the arrangement template.
  • the image generation module 440 can arrange the feedback information on the recommendation image when the feedback service mode is in the offline mode of the digital human, and feed back the recommendation image to the intelligent terminal, so that the recommendation image can be presented on the intelligent terminal, It is convenient for users to directly view the feedback information in the featured image.
  • an embodiment of the present application provides a customer service method, which can be applied to the interactive system 300 shown in FIG. 1 .
  • the customer service method may include the following steps S11 to S15.
  • Step S11 Establish a call connection with the smart terminal.
  • step S11 may be similar to the corresponding operations provided in the foregoing embodiments. Therefore, for step 11, reference may be made to the specific steps for establishing a call connection between the central control module 410 and the smart terminal, which will not be repeated here.
  • Step S12 Acquire call data with the smart terminal, and determine feedback information fed back to the smart terminal according to the call data.
  • step S12 may be similar to the corresponding operations provided in the above-mentioned embodiments. Therefore, in step S12, referring to the above-mentioned information processing module 420, the call data between the central control module 410 and the smart terminal can be obtained, and the call data can be obtained according to the call data. The specific steps of determining the feedback information fed back to the intelligent terminal from the data will not be repeated here.
  • Step S13 Obtain user information from the call data, and determine a feedback service mode according to the user information.
  • the feedback service mode is used to represent the service mode of sending feedback information to the smart terminal
  • the feedback service mode may include a digital human online mode
  • the digital human online mode is used to represent a mode in which a picture containing a digital human is presented on the intelligent terminal .
  • step S13 may be similar to the corresponding operations provided in the above-mentioned embodiments. Therefore, in step 13, referring to the above-mentioned service mode control module 430, the user information may be obtained from the call data, and the feedback service mode may be determined according to the user information. The specific steps are not repeated here.
  • Step S14 if the feedback business mode is the digital human online mode, obtain a digital human image including the digital human based on the feedback information, and use the digital human image as the interactive feedback result; wherein, the digital human image in the digital human image corresponds to the feedback information. .
  • step S14 may be similar to the corresponding operations provided in the foregoing embodiments. Therefore, in step 13, reference may be made to the above-mentioned image generation module 440 to obtain a digital human image including a digital human based on the feedback information, and convert the digital human image to the digital human image. As the specific steps of interactive feedback results, details are not repeated here.
  • Step S15 Send the interactive feedback result to the smart terminal.
  • step S15 may be similar to the corresponding operations provided in the above-mentioned embodiments. Therefore, in step 13, reference may be made to the above-mentioned specific steps of the central control module 410 sending the interactive feedback result to the intelligent terminal, which is not repeated here. Repeat.
  • a video call can be formed with the smart terminal, user information can be obtained from the call data, and a feedback service mode can be determined based on the user information, and if the feedback service mode is an online digital human mode , the digital human image is obtained through the feedback information obtained based on the call data, so that the digital human is displayed on the intelligent terminal.
  • this customer service method is applied to customer service, it can effectively improve the user's intimacy, and can also replace manual services. During the peak period of consultation, it can quickly respond to the user's consultation needs. At the same time, considering the user's needs, it can select the corresponding feedback business model to feed back information to the intelligent terminal, and improve the diversity of customer service forms to meet the needs of different users.
  • the embodiment of the present application further provides a customer service method, which can be applied to the interactive system 300 shown in FIG. 1 , and the customer service method can include the following steps S21 to S25.
  • the customer service method provided in this embodiment may include the same or similar steps as those in the above-mentioned embodiment. For the execution of the same or similar steps, reference may be made to the foregoing introduction, which will not be repeated in this specification.
  • Step S21 Establish a call connection with the smart terminal.
  • the server can send a call request to the smart terminal, and when the smart terminal allows the call request, the server can establish a call connection with the smart terminal; as shown in FIG. 5 , the above step S21 may include the following steps S211 to S213.
  • Step S211 Obtain the terminal identifier of the intelligent terminal.
  • the terminal identifier may be an identifier used to establish a high-definition call connection.
  • the terminal identifier may be a standard SIM card number, a mini SIM card number, a Micro SIM card number, and the like.
  • Step S212 Send a call request to the intelligent terminal through the terminal identifier.
  • dialing can be performed based on the terminal identification, thereby implementing a call request to the intelligent terminal.
  • Step S213 If it is determined that the smart terminal allows the call request, establish a call connection with the smart terminal.
  • an instruction may be generated on the smart terminal for enabling the smart terminal to establish a call connection with the server.
  • the intelligent terminal can also send a call request to the server, and after the server allows the call request, the intelligent terminal can establish a call connection with the server; as shown in FIG. 6 , the above step S21 can also include the following steps S214 to S214 to Step S215.
  • Step S214 Receive the call request sent by the intelligent terminal.
  • the customer service identifier may be called through the intelligent terminal, and then a call request may be sent to the server.
  • the customer service identifier can be set according to business requirements.
  • Step S215 If it is determined that the call request is allowed, a call connection is established with the smart terminal.
  • the server can either send a call request to the intelligent terminal to establish a call connection with the intelligent terminal, or receive a call request sent by the intelligent terminal to establish a call connection with the intelligent terminal.
  • Step S22 Acquire call data with the smart terminal, and determine feedback information fed back to the smart terminal according to the call data.
  • the characters that can be input by the user may be corresponding to the service feedback information; as shown in FIG. 7 , the above step S22 may include the following steps S221 to S224.
  • Step S221 Acquire a key-in command from the call data, where the key-in command is used to represent the command generated by the user operating on the interface of the intelligent terminal.
  • an instruction corresponding to the input character is generated, and the instruction can be used as a key-in instruction.
  • Step S222 Obtain the character corresponding to the input instruction.
  • Step S223 Obtain historical feedback information from the call data, where the historical feedback information includes interrelated service feedback information and preset characters.
  • the historical feedback information may include service feedback information and preset characters associated with the service feedback information.
  • the historical feedback information may include that the service feedback information is the available balance, the preset character is 1, the service feedback information is the remaining amount of the package, the preset character is 2, the balance is associated with the preset character "1", and the service feedback information is associated with the preset character "1".
  • the default character "2" is associated.
  • Step S224 If the character matches the preset character, the service feedback information associated with the preset character is used as the feedback information.
  • historical feedback information can be obtained from the call data, and then the association relationship between the service feedback information and preset characters can be obtained.
  • Associated business feedback information so as to achieve rapid response to user needs.
  • Step S23 Obtain user information from the call data, and determine a feedback service mode according to the user information.
  • the feedback service mode is used to represent the service mode for sending feedback information to the intelligent terminal.
  • the feedback service mode includes a digital human online mode, and the digital human online mode uses It is used to characterize the mode in which the picture containing the digital human is presented on the smart terminal.
  • the feedback service mode may be determined through the user image; as shown in FIG. 8 , the above step S23 may include the following steps S231 to S233.
  • Step S231 Obtain a user image including the user from the user information.
  • an image capture device for capturing images can be set on the smart terminal, and a user image including the user can be obtained through the image capture device; it can also be set on the smart terminal to receive information sent by other devices to obtain User image.
  • Step S232 Perform image recognition on the user image to obtain user characteristics.
  • the user image can be recognized by an image recognition algorithm to obtain the user characteristics.
  • the image recognition algorithm may include a maximum flow algorithm, a Hungarian algorithm, a breadth-first search algorithm, a depth-first search algorithm, and the like, which are not specifically limited here.
  • Step S233 If the user characteristics are consistent with the preset user characteristics, it is determined that the feedback service mode is the digital human online mode.
  • the feedback service mode can be determined to be the digital human online mode.
  • the function of the preset user feature is user identity verification, so as to prevent illegal users from using the intelligent terminal to use the customer service service.
  • the business mode associated with the user characteristics can be determined as the digital human online mode or the digital human offline mode model.
  • the role of the preset user feature is pattern matching.
  • the feedback service mode can be determined through the user image.
  • Step S24 if the feedback business mode is the digital human online mode, obtain a digital human image including the digital human based on the feedback information, and use the digital human image as the interactive feedback result; wherein, the digital human image in the digital human image corresponds to the feedback information. .
  • the user may select a designated digital person, and send an image containing the form of the digital person to the smart terminal; as shown in FIG. 9 , the above step S24 may include the following Steps S241 to S243.
  • Step S241 Acquire a digital human selection instruction and digital human driving information according to the feedback information.
  • the instruction for selecting a digital person may include an instruction for selecting a target digital person from a plurality of digital persons.
  • the digital human driving information may include instructions for driving the target digital human.
  • the feedback information can be associated with the digital person, and after establishing a call connection with the smart terminal, the digital person associated with the feedback information can be determined, thereby generating a digital person selection instruction for selecting the digital person;
  • the digital person designated by the user may also be obtained from the feedback information, and a digital person selection instruction for selecting the digital person may be generated.
  • the expression form of the digital person selection instruction corresponds to the acquisition of the digital person selection instruction.
  • the digital person currently designated by the user can be determined in real time based on the received feedback information, and the designated digital person can be driven. People can be different digital people that change in real time.
  • Step S242 Acquire a target digital human corresponding to the digital human selection instruction according to a preset digital human selection rule.
  • corresponding preset digital human selection rules may be set for different representations of digital human selection instructions.
  • the digital human selection instruction may be the digital human feature described in the feedback information, and the digital human selection instruction is screened from multiple digital human beings to obtain the target digital human.
  • the digital human includes a simulated digital human that looks like a real person and an artificial digital human that looks like an animation effect. Each simulated digital human is pre-numbered.
  • the feedback information determines that the user chooses a simulated digital human that looks like a real person
  • the simulated digital human that looks like a real person is obtained.
  • the target digital human is determined from a plurality of real-life simulated digital human beings according to the sequence of numbers.
  • the digital human selection instruction may be information used in the feedback information to characterize the digital human designated by the user.
  • the relationship between the feedback information and the digital human can be preset on the server.
  • the SIM card number can be associated with the digital person, and after the smart terminal establishes a call connection with the server through the SIM card number, the digital person associated with the SIM card number is used as the target digital person.
  • Step S243 Drive the target digital human based on the digital human driving information, and acquire an image of the target digital human including the target digital human.
  • the method of acquiring the target digital human image based on the digital human driving information is substantially the same as the method of acquiring the digital human image based on the feedback information in step S14, and details are not repeated here.
  • a corresponding target digital human can be selected based on user requirements, and the target digital human can be driven, so as to present a digital human image including the target digital human on the smart terminal, thereby It can bring a richer experience to users.
  • Step S25 Send the interactive feedback result to the smart terminal.
  • the above step S25 may include the following steps S251 to 253 .
  • Step S251 Obtain the privacy information in the interactive feedback result.
  • the privacy information may include information related to user privacy.
  • the private information can be ID number, user balance, dial record and so on.
  • the private information in the customer service feedback result may be obtained in different ways based on the content type in the customer service feedback result.
  • the text output information contained in the electronic card may be acquired, and the information related to user privacy in the text output information may be used as privacy information.
  • the customer service feedback result includes voice feedback information
  • text output information corresponding to the voice feedback information may be acquired, and information related to user privacy in the text output information may be used as privacy information.
  • Step S252 Update the interactive feedback result according to the preset information protection mode, so that the private information in the updated interactive feedback result is hidden.
  • the preset information protection mode may be set based on an actual scene.
  • the private information is text in an electronic card
  • one or more characters in the text can be hidden.
  • the text in the electronic card is an ID number
  • the ID number is 110101200108150612
  • the ID number can be updated, and the text in the updated electronic card is 1101012001XXXXXXX.
  • the text output information corresponding to the feedback voice information can be obtained.
  • the text output information is the user's balance
  • the numbers in the user's balance can be corresponding to letters, and the corresponding letters can be used as the updated letters.
  • customer service feedback For example, the number “1” can be assigned to the letter “A”, the number “2” to the letter “B”, the number “3” to the letter “C”, and the number “4" to the letter “D” , the number "5" corresponds to the letter "E”, and the number "6” corresponds to the letter "F”.
  • the user balance is 12345
  • the user balance can be updated to "ABCDE", which will correspond to "ABCDE” , as the updated customer service feedback result.
  • the manner in which numbers and letters correspond to each other may be preset by the user.
  • Step S253 Send the updated interactive feedback result to the smart terminal.
  • the smart terminal can play the updated customer service feedback result, thereby protecting the privacy information in the customer service feedback result.
  • the embodiment of the present application further provides a customer service method, which can be applied to the interactive system 300 shown in FIG. 1 , and the customer service method can include the following steps (1) to (5).
  • the customer service method provided in this embodiment may include the same or similar steps as those in the above-mentioned embodiment. For the execution of the same or similar steps, reference may be made to the foregoing description, which will not be repeated in this specification.
  • the feedback service mode is used to represent the service mode of sending feedback information to the intelligent terminal.
  • the feedback service mode includes digital human online mode, digital human online mode It is used to characterize the mode in which the picture containing the digital human is presented on the smart terminal.
  • the feedback business mode is the digital human online mode, obtain the digital human image including the digital human based on the feedback information, and use the digital human image as the interactive feedback result; wherein, the digital human form in the digital human image and the feedback information correspond.
  • the feedback service mode may also include a digital human offline mode, and the digital human offline mode is used to represent a mode in which a picture that does not include a digital human is presented on the intelligent terminal; the customer service provided by this embodiment
  • the service method may further include: if the feedback business mode is the digital human offline mode, generating a recommendation image from the feedback information according to a preset information arrangement rule, and using the recommendation image as an interactive feedback result.
  • the offline mode of the digital human may include a mode in which the interactive feedback result including the recommended image is fed back to the intelligent terminal.
  • the intelligent terminal establishes a call connection with the customer service system 400 provided in this embodiment, and the customer service system 400 can feed back an interactive feedback result to the intelligent terminal, and the interactive feedback result can include continuity
  • the intelligent terminal can present the multiple promotion images, and the screen presented by the intelligent terminal does not contain digital people.
  • the information in the promotion images is used to provide users with customer service services, so that the intelligent terminal can directly present the images.
  • the screen containing the feedback information allows the user to obtain the information fed back by the customer service system 400 more intuitively, which is convenient for the hearing-impaired user to obtain the information, so as to provide various customer service services for the user.
  • the feedback service type involved in the feedback information may be obtained first, an arrangement template may be obtained according to the feedback service type, and the feedback information may be arranged in the arrangement template, thereby obtaining a recommendation image containing the feedback information.
  • the feedback service type may be used to represent the type of feedback content in the feedback information.
  • the feedback business type may include consumption flow, available packages, remaining usage of packages, preferential activities, etc.
  • the layout template may include templates for arranging feedback information of various feedback service types.
  • the feedback information may be arranged to a target template; and generating a recommendation image from the feedback information according to a preset information arrangement rule may include: obtaining from a plurality of preset templates according to the feedback information.
  • Target template arrange the feedback information to the target template to get the recommended image.
  • the target template may include a plurality of areas for arranging feedback information, wherein the areas may be in one-to-one correspondence with the feedback service types involved in the feedback information, and this area is used for arranging the types of feedback services in the feedback information. Feedback information corresponding to this area.
  • the feedback content in the feedback information is obtained
  • the feedback business types involved in obtaining the feedback content include consumption flow and preferential activities
  • an arrangement template capable of arranging the feedback business types including consumption flow and preferential activities is obtained, which will be combined with the consumption flow.
  • the feedback information corresponding to the running water is arranged in the first area in the layout template.
  • the first area is used to arrange the feedback information whose feedback business type is consumption running water, and arrange the feedback information corresponding to the preferential activities in the layout template.
  • the second area is used for arranging the feedback information that the feedback service type is preferential activity. That is to say, feedback information of various feedback service types can be arranged in the arrangement template.
  • an electronic card may be generated while generating the digital human image; the customer service method provided in this embodiment may further include: obtaining prompt content for displaying on the smart terminal according to the feedback information; Generate an electronic card including prompt content, and use the electronic card as the interactive feedback result.
  • the electronic card can be used to display prompt content (such as service packages, package fees, etc.).
  • prompt content such as service packages, package fees, etc.
  • the electronic card when the driving information is that the packages that the user can handle include package A, package B, and package C, the prompt content can be package A, package B, and package C arranged in a preset format, and the prompt content can be The content is presented on the electronic card.
  • the card type of the electronic card may include static type and dynamic type.
  • the electronic card is a static type, an electronic card can be generated, and the smart terminal can present the electronic card, so as to obtain a static presentation effect of the electronic card;
  • the electronic card is a dynamic type, multiple electronic cards can be generated,
  • the plurality of electronic cards may be continuous electronic cards, and when the smart terminal presents the plurality of electronic cards according to the sequence of the electronic cards, the effect of changing and playing the video of the electronic cards may be presented.
  • the electronic card can be combined with the digital human image and sent to the smart terminal as an interactive feedback result, or can be independently sent to the smart terminal as a feedback result.
  • the electronic card and the digital human image can be overlapped, so that the electronic card covers the non-core area of the digital human image, and will be formed after overlapping.
  • the image is sent to the smart terminal as the interactive feedback result.
  • the non-core area may be an area other than the digital human in the digital human image, and may also be an area other than the digital human head in the digital human image.
  • the electronic card and the digital human image are fused to form an image including the electronic card and the digital human, in the process of transmitting the image to the intelligent terminal in the form of a data packet, it is not necessary to transmit the electronic card and the digital human image separately.
  • the smart terminal presents interactive feedback results, it can directly display images including electronic cards and digital people.
  • the electronic card and the digital human image can be respectively transmitted to the smart terminal in the form of data packets, and when the smart terminal receives the electronic card and the digital human image, it can The electronic card is used as the first layer, and the digital human image is used as the second layer.
  • the smart terminal needs to display the first layer and the second layer at the same time, the first layer can be placed on the top layer, and the second layer can be placed on the top layer. placed on the bottom layer.
  • the digital human image can be sent to the smart terminal and the electronic card can be generated at the same time, so that the user can view the prompt content of the electronic card at the same time during the high-definition call with the digital human customer service, helping the user to better understand and handle the business. .
  • the customer service device may include a connection module 61 , a feedback information determination module 62 , a business mode determination module 63 , and a digital human drive module 64 and interactive feedback result sending module 65 .
  • the connection module 61 is used to establish a call connection with the intelligent terminal.
  • the feedback information determining module 62 is configured to acquire call data with the smart terminal, and determine feedback information fed back to the smart terminal according to the call data.
  • the service mode determination module 63 is configured to acquire user information from the call data, and determine a feedback service mode according to the user information.
  • the digital human driving module 64 is configured to obtain a digital human image that can include a digital human based on the feedback information if the feedback business mode is the digital human online mode, and use the digital human image as the interactive feedback result; The form corresponds to the feedback information.
  • the interactive feedback result sending module 65 is configured to send the interactive feedback result to the smart terminal.
  • the connection module 61 may include a terminal identification acquiring unit, a calling unit and a first connection unit.
  • the terminal identification obtaining unit is used for obtaining the terminal identification of the intelligent terminal.
  • the calling unit is used to send a calling request to the intelligent terminal through the terminal identification.
  • the first connection unit is configured to establish a call connection with the intelligent terminal if it is determined that the intelligent terminal allows the call request.
  • connection module 61 may include a call receiving unit and a second connection unit.
  • the call receiving unit is used for receiving the call request sent by the intelligent terminal.
  • the second connection unit is configured to establish a call connection with the smart terminal if it is determined that the call request is allowed.
  • the digital human driving module 64 may include an instruction acquisition unit, a target digital human acquisition unit, and a target digital human image acquisition unit.
  • the instruction acquisition unit is used for acquiring the digital human selection instruction and the digital human driving information according to the feedback information.
  • the target digital person obtaining unit is configured to obtain the target digital person corresponding to the digital person selection instruction according to a preset digital person selection rule.
  • the target digital human image acquisition unit is configured to drive the target digital human based on the digital human driving information, and acquire an image of the target digital human that may include the target digital human.
  • the customer service apparatus may further include a digital human offline processing module.
  • the digital human offline processing module is configured to generate a recommendation image from the feedback information according to a preset information arrangement rule if the feedback business mode is the digital human offline mode, and use the recommendation image as an interactive feedback result.
  • the digital human offline processing module may include a target template acquisition unit and a recommendation image acquisition unit.
  • the target template obtaining unit is configured to obtain the target template from a plurality of preset templates according to the feedback information.
  • the recommended image acquisition unit is used for arranging the feedback information to the target template to obtain the recommended image.
  • the feedback information determination module 62 may include a key-in instruction acquisition unit, a character acquisition unit, a historical feedback information acquisition unit, and a character association unit.
  • the key-in instruction obtaining unit is used to obtain key-in instructions from the call data
  • the key-in instructions are used to represent the instructions generated by the user operating on the interface of the intelligent terminal.
  • the character acquisition unit is used to acquire the character corresponding to the key-in instruction.
  • the historical feedback information acquisition unit is used for acquiring historical feedback information from the call data, and the historical feedback information may include interrelated service feedback information and preset characters.
  • the character association unit is configured to use the service feedback information associated with the preset character as the feedback information if the character matches the preset character.
  • the service mode determination module 63 may include a user image acquisition unit, a user feature acquisition unit, and an online mode determination unit.
  • the user image acquisition unit is configured to acquire a user image including the user from the user information.
  • the user feature acquisition unit is used to perform image recognition on the user image to obtain the user feature.
  • the online mode determining unit is configured to determine that the feedback service mode is the digital human online mode if the user characteristics are consistent with the preset user characteristics.
  • the interactive feedback result sending module 65 may include a privacy information acquisition unit, an interactive feedback result updating unit, and an interactive feedback result sending unit.
  • the private information acquisition unit is used for acquiring the private information in the interactive feedback result.
  • the interactive feedback result updating unit is configured to update the interactive feedback result according to the preset information protection mode, so that the privacy information in the updated interactive feedback result is hidden.
  • the interactive feedback result sending unit is configured to send the updated interactive feedback result to the intelligent terminal.
  • the customer service apparatus may further include a prompt content acquisition module and an interactive feedback result generation module.
  • the prompt content acquisition module is configured to acquire prompt content for display on the smart terminal according to the feedback information.
  • the interactive feedback result generating module is used to generate an electronic card that can include prompt content, and use the electronic card as the interactive feedback result.
  • the electronic device may include a processor 810, a communication module 820, a memory 830, and a bus.
  • the bus can be an ISA bus, a PCI bus, an EISA bus, a CAN bus, or the like.
  • the bus can be divided into address bus, data bus, control bus and so on. in:
  • Memory 830 for storing computer readable instructions may be used to store software computer readable instructions as well as various data.
  • the memory 830 may mainly include an area for storing computer-readable instructions and an area for storing data, wherein the area for storing computer-readable instructions may store computer-readable instructions required to operate at least one function may include computer-readable instructions, and the computer-readable instructions include computer-readable instructions. operating instructions.
  • the memory 830 can also temporarily store messages and the like that the communication module 820 needs to send.
  • the memory 830 may include high-speed RAM memory, and may also include non-volatile memory (non-volatile memory), such as at least one solid state disk (Solid State Disk, SSD for short).
  • the processor 810 is used to execute computer readable instructions stored in the memory 830 .
  • the steps of the customer service methods of the above embodiments are implemented.
  • Embodiments of the present application further provide one or more computer-readable storage media, where computer-readable instructions are stored on the computer-readable storage media, and when the computer-readable instructions are executed by a processor, each of the customer service methods of the foregoing embodiments is implemented. process, and can achieve the same technical effect, in order to avoid repetition, it will not be repeated here.
  • the computer-readable storage media such as read-only memory (Read-Only Memory, referred to as ROM), random access memory (Random Access Memory, referred to as RAM), SSD, electrically erasable programmable read-only memory (Electrically Erasable Programmable read-only memory) read only memory, referred to as EEPROM) or flash memory (Flash Memory, referred to as Flash) and so on.
  • ROM read-only memory
  • RAM random access memory
  • SSD electrically erasable programmable read-only memory (Electrically Erasable Programmable read-only memory) read only memory
  • EEPROM Electrically erasable programmable read-only memory
  • Flash memory Flash Memory

Abstract

一种客服服务系统，包括中控模块、信息处理模块、业务模式控制模块和图像生成模块。其中，中控模块用于与智能终端建立通话连接。信息处理模块用于获取中控模块与智能终端之间的通话数据，并根据通话数据确定向智能终端反馈的反馈信息。业务模式控制模块用于从通话数据中获取用户信息，并根据用户信息确定反馈业务模式。图像生成模块用于若反馈业务模式为数字人在线模式，则基于反馈信息获取包括数字人的数字人图像，并将数字人图像作为交互反馈结果发送至中控模块；中控模块还用于将所述交互反馈结果发送至所述智能终端。

Description

客服服务系统、方法、装置、电子设备及存储介质
本申请要求于2021年03月16日提交中国专利局，申请号为2021102783203，申请名称为“客服服务系统、方法、装置、电子设备及存储介质”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
本申请涉及人机交互技术领域，更具体地，涉及一种客服服务系统、方法、装置、电子设备及存储介质。
背景技术
近些年来，随着网络信息技术的不断发展和应用，传统的PSTN(公共交换电话网络，Public Switched Telephone Network)通话方式逐渐不能满足用户的需求，进而导致PSTN通话逐渐被高清通话所取代，基于此，市场上出现了允许客服人员通过视频交互为用户服务的服务模式，该视频交互的方式能够让用户感觉更加亲切，但是，在咨询高峰期时，很容易出现客服人员数量不足的状况，用户需较长时间等待，导致用户的需求得不到及时响应。
发明内容
鉴于上述问题，本申请实施例提出了一种客服服务系统、方法、装置、电子设备及存储介质。
第一方面，本申请实施例提供了一种客服服务系统，该客服服务系统包括中控模块、信息处理模块、业务模式控制模块和图像生成模块。其中，中控模块用于与智能终端建立通话连接。信息处理模块用于获取中控模块与智能终端之间的通话数据，并根据通话数据确定向智能终端反馈的反馈信息。业务模式控制模块用于从通话数据中获取用户信息，并根据用户信息确定反馈业务模式。图像生成模块用于若反馈业务模式为数字人在线模式，则基于反馈信息获取包括数字人的数字人图像，并将数字人图像作为交互反馈结果发送至中控模块；其中，数字人图像中数字人的形态与反馈信息对应，中控模块还用于将交互反馈结果发送至智能终端。
第二方面，本申请实施例提供了一种客服服务方法，由电子设备执行，该客服服务方法包括：与智能终端建立通话连接；获取与智能终端之间的通话数据，并根据通话数据确定向智能终端反馈的反馈信息；从通话数据中获取用户信息，并根据用户信息确定反馈业务模式；若反馈业务模式为数字人在线模式，则基于反馈信息获取包括数字人的数字人图像，并将数字人图像作为交互反馈结果；其中，数字人图像中数字人的形态与反馈信息对应；以及将交互反馈结果发送至智能终端。
第三方面，本申请实施例提供了一种客服服务装置，设置于电子设备中，该客服服务装置包括连接模块、反馈信息确定模块、业务模式确定模块、数字人驱动模块和交互反馈结果发送模块。其中，连接模块用于与智能终端建立通话连接。反馈信息确定模块用于获取与智能终端之间的通话数据，并根据通话数据确定向智能终端反馈的反馈信息。业务模式确定模块用于从通话数据中获取用户信息，并根据用户信息确定反馈业务模式。数字人驱动模块用于若反馈业务模式为数字人在线模式，则基于反馈信息获取包括数字人的数字人图像，并将数字人图像作为交互反馈结果；其中，数字人图像中数字人的形态与反馈信息对应。交互反馈结果发送模块用于将交互反馈结果发送至智能终端。
第四方面，本申请实施例提供了一种电子设备，该电子设备包括：一个或多个处理器；存储器；一个或多个计算机可读指令，其中一个或多个计算机可读指令被存储在存储器中并被配置为由一个或多个处理器执行，一个或多个计算机可读指令配置用于执行第一方面提供的客服服务方法的步骤。
第五方面，本申请实施例提供了一个或多个计算机可读存储介质，计算机可读存储介质存储有计算机可读指令，计算机可读指令被处理器执行第二方面提供的客服服务方法的步骤。
本申请的一个或多个实施例的细节在下面的附图和描述中提出。基于本申请的说明书、附图以及权利要求书，本申请的其它特征、目的和优点将变得更加明显。
附图说明
为了更清楚地说明本申请实施例中的技术方案，下面将对实施例描述中所需要使用的附图作 简单地介绍，显而易见地，下面描述中的附图仅仅是本申请的一些实施例，对于本领域技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1示出了本申请实施例提供的客服服务系统的一种应用环境示意图。
图2示出了本申请实施例提供的客服服务系统的一模块示意图。
图3示出了本申请实施例提供的客服服务系统的另一模块示意图。
图4示出了本申请实施例提供的客服服务方法的一流程示意图。
图5示出了一个实施例中建立通话连接的一流程示意图。
图6示出了一个实施例中建立通话连接的另一流程示意图。
图7示出了一个实施例中获取反馈信息的一流程示意图。
图8示出了一个实施例中确定数字人在线模式的一流程示意图。
图9示出了一个实施例中获取数字人图像的一流程示意图。
图10示出了一个实施例中发送交互反馈结果的一流程示意图。
图11示出了本申请实施例提出的一种客服服务装置的功能模块框图。
图12示出了本申请实施例提出的一种电子设备的功能模块框图。
具体实施方式
为了使本技术领域的人员更好地理解本申请方案，下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。
近些年来，随着通信技术的不断发展，传统的PSTN通话方式逐渐被高清通话方式所取代，比如VOLTE(长期演进语音承载，Voice over Long-Term Evolution)通话方式，相对于PSTN通话方式，高清通话方式的语音视频通话更清晰、掉线率更低，在通话过程中能够同时使用流量数据业务，同时无需借助安装于智能终端的程序和软件即可实现视频通话，从而使通话双方的通话体验更好。基于此，现有技术中出现了客服人员通过视频交互的方式为用户提供视频客服服务，在这个过程中，客服人员所持有的智能终端能够将客服人员的音视频信息发送至用户所持有的智能终端，从而达到将客服人员直播呈现给用户的效果，该种方式能够让用户感觉更加亲切，提高用户向客服人员咨询的体验。该种方式虽然能够有效提高用户的体验，但是，在咨询高峰期，很容易出现客服人员数量不足的状况，导致用户需要等待较长的时间，使用户的体验较差。
为了提高用户的体验，本申请发明人投入研发，发明人发现可以提前录制视频，当处于咨询高峰期时，若获取到用户的需求，则基于该需求将提前录制的视频发送至用户持有的智能终端，能够实现在咨询高峰期时，用户持有的智能终端呈现的画面始终为客服人员的视频，但是，该种方式会导致各个录制的视频之间衔接不流畅，显示的内容也比较单一，导致用户所看到的客服人员的画面较为生硬。
基于此，本申请发明人继续投入研发，为了进一步提高用户的体验，使用户能够看到流畅的关于客服人员的视频画面，本申请发明人提出了本申请实施例的客服服务系统、方法、装置、电子设备及存储介质，该客服服务系统包括中控模块、信息处理模块、业务模式控制模块和图像生成模块。其中，中控模块用于与智能终端建立通话连接。信息处理模块用于获取中控模块与智能终端之间的通话数据，并根据通话数据确定向智能终端反馈的反馈信息。业务模式控制模块用于从通话数据中获取用户信息，并根据用户信息确定反馈业务模式，反馈业务模式用于表征向智能终端发送反馈信息的业务模式，反馈业务模式包括数字人在线模式，数字人在线模式用于表征在智能终端呈现包含数字人的画面的模式。图像生成模块用于若反馈业务模式为数字人在线模式，则基于反馈信息获取包括数字人的数字人图像，并将数字人图像作为交互反馈结果发送至中控模块；其中，数字人图像中数字人的形态与反馈信息对应，中控模块还用于将交互反馈结果发送至智能终端。通过本申请实施例的实施，可以通过中控模块与智能终端形成视频通话，从通话数据中获取用户信息，基于用户信息确定反馈业务模式，并若反馈业务模式为在线数字人模式，则通过基于通话数据获得的反馈信息得到数字人图像，从而在智能终端显示数字人，当将本客服服务系统应用于客服服务时，能够有效提高用户的亲切度，同时可以代替人工服务，当处于咨询高峰期时，可以快速响应用户的咨询需求，同时，可以考虑到用户需求，选择对应的反馈业务模式向智能终端反馈信息，提升客服服务的形式多样性，从而 满足不同用户的需求。
为了更好理解本申请实施例提供的一种客服服务系统、方法、装置、电子设备及存储介质，下面先对适用于本申请实施例的应用环境进行描述。
请参阅图1，图1示出了一种适用于本申请实施例所提供的客服服务系统的应用环境示意图。本申请实施例提供的客服服务系统、方法、装置、电子设备及存储介质可以应用于如图1所示的交互系统300。该交互系统300包括智能终端301以及服务端302，服务端302与智能终端301通信连接。其中，服务端302可以用独立的服务器或者是多个服务器组成的服务器集群来实现。服务器可以是云端服务器，还可以是传统机房服务器，在此不作具体限定。
在一些实施例中，智能终端301可以是具有显示屏且支持数据输入的各种电子设备，包括但不限于智能手机、平板电脑、膝上型便携计算机、台式计算机和可穿戴式电子设备等。具体地，数据输入可以是基于智能终端301所配置的语音模块输入语音、字符输入模块输入字符、图像输入模块输入图像、视频输入模块输入视频等，还可以是基于智能终端301所配置的手势识别模块，使得用户可以实现手势输入等交互方式。
其中，智能终端301上可以安装有客户端应用程序，用户可以基于客户端应用程序(例如微信等APP、微信小程序等)与服务端302进行通信，具体地，服务端302上安装有对应的服务端应用程序，用户可以基于客户端应用程序在服务端302注册一个用户帐号，并基于该用户帐号与服务端302进行通信，例如用户在客户端应用程序登录用户帐号，并基于该用户帐号通过客户端应用程序进行输入，可以输入文字信息、语音信息、图像信息或视频信息等，客户端应用程序接收到用户输入的信息后，可以将该信息发送至服务端302，使得服务端302可以接收该信息并进行处理及存储，服务端302还可以接收该信息并根据该信息将对应的输出信息返回至智能终端301。另外，智能终端301可以接收服务端302的拨号呼叫请求，也可以向服务端302发送拨号呼叫请求。
在一些实施例中，服务端302可以用于接收用户输入的信息，并根据该信息生成仿真数字人的画面，并将该画面发送至智能终端301，从而向智能终端301提供客服服务，与用户进行客服沟通。具体地，智能终端301可以接收用户输入的信息，并呈现服务端302向智能终端301发送的仿真数字人的画面。其中，仿真数字人是基于可视化图形的软件程序，该软件程序被执行后可以向用户呈现出模拟生物行为或思想的机器人形态。仿真数字人可以是模拟真人式的仿真数字人，例如根据用户自身或其他自然人的形态建立的形似真人的仿真数字人，也可以是动漫效果式的仿真数字人，例如动物形态或卡通人物形态的仿真数字人。另外，服务端302可以接收智能终端301的拨号呼叫请求，也可以向智能终端发送拨号呼叫请求。
在一些实施例中，如图1所示，智能终端301在获取与用户输入的信息对应的回复信息后，可以在智能终端301的显示屏或与其连接的其他图像输出设备上显示对应与该回复信息的仿真数字人图像。作为一种方式，在播放数字人图像的同时，可以通过智能终端301的扬声器或与其连接的其他音频输出设备播放与数字人图像对应的音频，还可以在智能终端301的显示屏上显示与该回复信息对应的文字或图形，实现在图像、语音、文字等多个方面上与用户的多态交互。
上述应用环境仅为方便理解所作的示例，可以理解的是，本申请实施例不仅局限于上述应用环境。
下面将通过具体实施例对本申请实施例提供的客服服务系统、方法、装置、电子设备及存储介质进行详细说明。
请参阅图2，本申请实施例提供一种客服服务系统400，该客服服务系统400可以部署于如图1所示的交互系统300的服务端302，该客服服务系统400包括中控模块410、信息处理模块420、业务模式控制模块430和图像生成模块440。其中，中控模块410用于与智能终端建立通话连接。信息处理模块420用于获取中控模块410与智能终端之间的通话数据，并根据通话数据确定向智能终端反馈的反馈信息。业务模式控制模块430用于从通话数据中获取用户信息，并根据用户信息确定反馈业务模式。图像生成模块440用于若反馈业务模式为数字人在线模式，则基于反馈信息获取包括数字人的数字人图像，并将数字人图像作为交互反馈结果发送至中控模块410；其中，数字人图像中数字人的形态与反馈信息对应，中控模块410还用于将交互反馈结果发送至智能终端。
在本实施例中，反馈业务模式可以用于表征客服服务系统400发送反馈信息的业务模式， 反馈业务模式包括数字人在线模式，数字人在线模式用于表征在智能终端呈现包含数字人的画面的模式
其中，数字人在线模式可以包括向智能终端反馈包含数字人图像的交互反馈结果的模式。例如，当反馈业务模式为数字人在线模式时，智能终端与本实施例提供的客服服务系统400建立通话连接，客服服务系统400可以向智能终端反馈交互反馈结果，该交互反馈结果可以包括连续性的多张数字人图像，由此，智能终端可以呈现该多张数字人图像，达到数字人为用户提供客服服务的效果，而无需专门安排客服工作人员为客户提供客服服务，避免客服人员人力紧张，实现快速响应用户的咨询需求。
在本实施例中，中控模块410可以与智能终端之间建立通话连接。其中，中控模块410与智能终端之间的通话方式可以为高清通话，也就是说，部署有客服服务系统400的服务端302可以与智能终端进行高清通话，该高清通话过程无需依赖在智能终端上安装的应用程序(例如，该应用程序可以包括具有视频通话功能的软件)。例如，高清通话包括VOLTE(Voice over Long-Term Evolution，长期演进语音承载)、畅连通话等。另外，中控模块410可以直接与智能终端建立通话连接，也可以间接地与智能终端建立通话连接，此处对中控模块410与智能终端建立通话连接的方式不做具体限制。
在一些示例中，当中控模块410直接与智能终端建立通话连接时，中控模块410可以具备呼叫功能和/或接听功能。具体地，中控模块410可以通过SIM卡号码向智能终端发出呼叫请求，当智能终端一侧接受该呼叫请求后，中控模块410与智能终端建立高清通话连接，中控模块410可以从智能终端获取用户信息，并将数字人图像发送至智能终端；智能终端可以向中控模块410发送呼叫请求，当中控模块410一侧接受该呼叫请求后，中控模块410与智能终端建立高清通话连接。
在一些示例中，当中控模块410间接与智能终端建立通话连接时，可以在中控模块410与智能终端形成的通信链路之间，设置呼叫平台和SIP(会话初始协议，Session Initiation Protocol)服务器。具体地，呼叫平台可以通过SIM卡号码向智能终端发出呼叫请求，当智能终端一侧接受该呼叫请求后，呼叫平台与智能终端建立高清通话连接，同时，呼叫平台可以通过RTP(实时传输，Real-time Transport Protocol)协议与SIP服务器建立通话连接，SIP服务器与中控模块410建立数据连接，呼叫平台和SIP服务器可以作为智能终端与中控模块410之间的传输媒介，同时RTP协议作为数据传输的规则。呼叫平台可以通过SIM卡号码接收由智能终端发出呼叫请求，当允许智能终端发出的呼叫请求后，呼叫平台可以与智能终端建立高清通话连接，同时，呼叫平台可以通过RTP协议与SIP服务器建立通话连接，SIP服务器与中控模块410建立数据连接，呼叫平台和SIP服务器可以作为智能终端与中控模块410之间的传输媒介，同时RTP协议作为数据传输的规则。
在本实施例中，通话数据可以包括中控模块410接收的智能终端发送的信息以及向智能终端发送的信息。例如，该信息可以为智能终端与中控模块410之间以往的通话记录，还可以为智能终端实时向中控模块410发送的信息。另外，当通话数据包括通话记录时，服务端可以预存有智能终端与中控模块410之间以往的通话记录，当需要获取通话数据时，可以从服务端提取得到该通话记录。
在本实施例中，反馈信息可以包括基于用户的意图确定向用户反馈的信息。具体而言，可以基于包括通话记录和实时接收到智能终端发送的信息的通话数据确定用户的意图，进而确定向用户反馈的信息。
示例性地，可以将通话记录和智能终端实时发送的数据进行结合，并基于结合后的信息确定用户的意图。例如，基于通话记录确定用户想要办理的套餐有套餐A和套餐B，不想办理的套餐有套餐C和套餐D，当智能终端杆实时发送的数据包含套餐办理时，可以确定用户想要办理套餐A或套餐B，并将此作为用户的意图。
在本实施例中，业务模式控制模块430可以从通话数据中获取用户信息。用户信息可以包括与用户相关的信息。例如，用户信息可以为用户在智能终端301键入的信息、用户的语音，用户的图像等，此处对用户信息的类型不做具体限制。
示例性地，当用户信息为用户在智能终端301键入的信息时，可以将该信息与预先存储的信息进行匹配，当匹配成功时，可以将与预先存储的信息对应的业务模式作为反馈业务模 式。例如，用户在智能终端301键入的信息可以为符号、数字等。当用户信息为用户的语音时，可以提取该语音的声纹特征，将该声纹特征与预先存储的该用户的声纹特征进行匹配，当匹配成功时，可以将与预先存储的用户的声纹特征对应的业务模式作为反馈业务模式。当用户信息为用户的图像时，可以提取该图像的特征，将该特征与预先存储的用户的特征进行匹配，当匹配成功时，可以将与预先存储的用户的特征对应的业务模式作为反馈业务模式。
在一些示例中，可以预先存储用户所需要的业务模式，并将业务模式与用户进行关联，当获取到用户信息时，可以将与用户关联的业务模式作为反馈业务模式。具体而言，反馈业务模式可以包括数字人在线模式，当反馈业务模式为数字人在线模式时，可以向智能终端发送数字人图像，智能终端可以播放连续的多张数字人图像，从而达到数字人为用户提供客服服务的效果。
另外，在本实施例中，反馈信息可以包括驱动信息，该驱动信息可以用于驱动数字人。具体而言，可以基于反馈信息获取用于驱动数字人的驱动信息，通过该驱动信息驱动数字人，并获取数字人图像。其中驱动信息的形式可以包括文字、语音、语义等，此处对驱动信息的类型不做具体限制。另外，数字人的形态可以包括数字人的动作状态。例如，数字人的形态可以为数字人的眼球、嘴型、表情、双臂等形态。需要注意的是，驱动信息可以是具有时间属性的连续性参数，按照时序控制数字人的动作，驱动信息可以与数字人的动作状态对应关联。例如，驱动信息可以包括在时刻t1控制3D数字人头部上扬15度，在时刻t2控制3D数字人头部左右摆动，在时刻t3控制3D数字人微笑等。
需要注意的是，数字人图像可以被理解为在某一角度拍摄数字人所得到的图像，该数字人图像中数字人的驱动信息与动作状态对应。
进一步地，本实施例中的数字人可以是通过3D建模得到，也可以是通过深度学习模型生成每一帧画质近乎于相机拍摄的逼真图像，数字人如同相机拍摄的真人的效果。可选地，可以由连贯的逼真图像生成视频数字人。
具体地，通过3D建模方式获得数字人的方法可以包括：获取多张包括目标模特的样本图像；根据多张样本图像获取目标模特的形态；获取原始数字人以及建模信息，建模信息包括原始数字人的原始关键点；根据形态获取目标模特的目标关键点，将原始关键点与目标关键点对应，生成数字人。
在本实施例中，目标模特可以包括与数字人相关的模特。例如，当数字人为某播音主持人C时，目标模特可以是该播音主持人C，还可以是与该播音主持人C面容、骨架、身材相仿的人物，还可以是与该播音主持人C面容、骨架、身材相仿的假人(例如人物蜡像)。
在本实施例中，样本图像可以包括多张目标模特的不同角度的图像。具体地，样本图像可以是在各个角度下，目标模特各种动作、发音、表情等情况下的图像。
在一些示例中，可以将用于采集目标模特的图像的摄像装置环球形布置于目标模特周边，其中，相对于目标模特的同一方位，可以设置不同焦距大小的摄像装置。当目标模特发出声音、变化动作、变化面部表情等时，可以采用各个摄像装置同时采集包括目标模特的图像，从而得到样本图像。
在本实施例中，目标模特的形态可以包括目标模特的躯体变化情况相关的信息。例如，形态可以是嘴角下垂、眼球右偏、头部上扬、右手抬起等。
在一些示例中，可以通过目标检测算法从样本图像中获取到目标模特的各个部位，再基于多张连续性的样本图像中的相同部位的变化状态确定该部位的形态，从而得到目标模特各个部位的形态。例如，目标检测算法可以是滑动窗口目标检测、two stage目标检测算法、one stage目标检测算法等算法。
在本实施例中，原始数字人可以包括已经构建好的数字人的模型。例如，该原始数字人可以是某一地域的平均人脸模型，也可以是工业动漫中的3D动画模型，此处对原始数字人的类型不做具体限制。另外，建模信息可以包括用于构建原始数字人的参数信息，通过该建模信息可以还原该原始数字人，使该原始数字人能够被呈现。
在本实施例中，可以将目标模特的形态与建模信息相结合，使得目标模特的形态特征添加至原始数字人，从而得到与目标模特的形态大致相同的数字人。
在本实施例中，原始数字人的原始关键点可以包括用于辨识、定位和控制原始数字人各 个部位的位置。例如，原始关键点可以是原始数字人的左眼角位置、右眼角位置、嘴角位置、面部轮廓位置、眉毛位置、鼻翼位置、大拇指位置、肩部位置等。需要注意的是，建模信息中原始关键点在原始数字人的各部位越密集，最终构建的数字人越准确。
在本实施例中，目标关键点可以包括用于辨识和定位目标模特各个部位的位置。例如，目标关键点可以是目标模特的左眼角位置、右眼角位置、嘴角位置、面部轮廓位置、眉毛位置、鼻翼位置、大拇指位置、肩部位置等。需要注意的是，目标关键点在目标模特的各个部位越密集，最终构建的数字人越准确。
在本实施例中，可以将原始数字人的各个部位的原始关键点与目标模特相同部位的目标关键点一一对应。例如，原始关键点包括原始数字人脸部轮廓的各个位置，目标关键点包括目标模特脸部轮廓的各个位置，将处于原始数字人脸部上庭的各个位置分别与处于目标模特脸部上庭的各个位置一一对应，将处于原始数字人脸部中庭的各个位置分别与处于目标模特脸部中庭的各个位置一一对应，将处于原始数字人脸部下庭的各个位置分别与处于目标模特脸部下庭的各个位置一一对应，原始数字人脸部的其他原始关键点与目标模特脸部的其他目标关键点一一对应的方式同理，此处不再赘述。
在一些示例中，可以将处于动态的目标关键点与原始关键点对应。具体地，可以获取目标模特的目标关键点，对连续性样本图像中目标模特的面部进行标记，并按照连续性样本图像的时序将目标模特面部的同一目标关键点进行关联，从而获得目标模特面部的各个目标关键点的动态变化轨迹；将各个目标关键点的动态变化轨迹与原始数字人面部的原始关键点对应，从而获得原始数字人面部的原始关键点的动态变化轨迹，计算同一时刻下，原始关键点的动态变化轨迹的幅度与目标关键点的动态变化轨迹之间的幅度差值，在该幅度差值大于预设幅度阈值，则确定需要对原始数字人的原始关键点进行修正。
示例地，可以获取目标模特的目标关键点，对连续性样本图像中目标模特的面部进行标记，并按照连续性样本图像的时序将目标模特面部的同一目标关键点进行关联，从而获得目标模特面部的各个目标关键点的动态变化轨迹，并将各个目标关键点的动态变化轨迹与原始数字人面部的原始关键点对应，从而获得原始数字人面部的原始关键点的动态变化轨迹，基于该原始关键点的动态变化轨迹获取到原始数字人在各时刻的面部变化幅度，基于该目标关键点的动态变化轨迹获取到目标模特在各个时刻的面部变化幅度，比较原始数字人在各时刻的面部变化幅度和目标模特在各个时刻的面部变化幅度，若在同一时刻下原始数字人的面部变化幅度与目标模特的面部变化幅度之差大于预设幅度阈值，则确定需要对原始数字人的原始关键点进行修正；若在同一时刻下原始数字人的面部变化幅度与目标模特的面部变化幅度之差小于或等于预设幅度阈值，则确定当前数字人的面部构建符合预期。
具体地，也可以通过预先构建仿真数字人模型，通过仿真数字人生成每一帧画质近乎于相机拍摄的逼真图像，构建仿真数字人生成模型可以包括：获取多张包括目标模特的样本图像以及与每张样本图像对应的相机参数；获取与相机参数对应的样本图像配置参数；根据样本图像获取目标模特的角度信息，并将角度信息与样本图像配置参数关联；根据样本图像配置参数和角度信息构建仿真数字人模型，得到预设的仿真数字人模型。
在本实施例中，相机参数可以包括用于拍摄样本图像的拍摄装置拍摄目标模特时所采用的参数。例如，相机参数可以是焦距、光圈大小等。样本图像配置参数可以包括用于拍摄样本图像的拍摄装置拍摄目标模特所生成的样本图像的参数。例如，样本图像配置参数可以是像素大小、图像曝光度、目标模特在图像中的占比、目标模特与地面接触的位置等。角度信息可以包括样本图像中目标模特呈现的角度。例如，当样本图像中目标模特的面部朝向与预设轴线方向之间的夹角为15度时，可以将15度作为该角度信息。
在一些示例中，可以对样本图像进行识别，从而获得目标模特的角度。具体地，可以通过目标检测算法从样本图像中获取到目标模特的各个部位，再基于多张连续性的样本图像中的相同部位的变化状态确定该部位的角度，从而得到目标模特各个部位的角度，并将该各个部位的角度作为角度信息。
在本实施例中，可以将样本图像视为由多个区域、多个点组成，基于样本图像配置参数和角度信息获取到目标模特在各个角度下多个区域、多个点位的状态，将各个角度下的区域、点位进行结合，构建得到仿真数字人模型，以使仿真数字人模型能够输出不同角度下的包括 目标模特的图像。
需要注意的是，相对于通过3D建模获取数字人，通过构建仿真数字人生成模型的过程无需3D建模，得到的仿真数字人也更接近真人模特，效果更加逼真，适用于实际应用中对可能需要对不同真人模特进行建模以获取仿真数字人的情况。
在本实施例中，中控模块410可以通过其与智能终端之间的通信链路，将交互反馈结果发送至智能终端，智能终端可以显示交互反馈结果中的数字人图像。具体地，当数字人图像包括连续的图像帧时，可以在智能终端呈现出数字人视频的效果。
在本实施例中，通过采用包括中控模块410、信息处理模块420、业务模式控制模块430和图像生成模块440的客服服务系统400，能够通过中控模块410与智能终端形成视频通话，从通话数据中获取用户信息，基于用户信息确定反馈业务模式，并若反馈业务模式为在线数字人模式，则通过基于通话数据获得的反馈信息得到数字人图像，从而在智能终端显示数字人，当将本客服服务系统400应用于客服服务时，能够有效提高用户的亲切度，同时可以代替人工服务，当处于咨询高峰期时，可以快速响应用户的咨询需求，同时，可以考虑到用户需求，选择对应的反馈业务模式向智能终端反馈信息，提升客服服务的形式多样性，从而满足不同用户的需求。
进一步地，作为本实施例的一种实施方式，如图3所示，信息处理模块420可以包括语音识别单元421和文本反馈单元422。其中，语音识别单元421用于从通话数据中获取用户语音，并识别用户语音得到文本输入信息。文本反馈单元422用于根据文本输入信息确定用户意图，并基于用户意图获取文本输出信息；文本输出信息用于确定反馈信息。
在一个实施例中，反馈信息可以包括文本输出信息和反馈语音信息等中的至少一种。
在一个实施例中，可以直接将文本输出信息作为反馈信息。文本输出信息，是文本格式的反馈信息。
在另一个实施例中，可以基于文本输出信息进行语音合成，生成反馈语音信息，将反馈语音信息作为反馈信息。反馈语音信息是指语音格式的反馈信息。
在又一个实施例中，还可以将文本输出信息和反馈语音信息一并作为反馈信息。
在一个实施例中，语音识别单元421、语义单元422可以分别与中控模块410形成通信连接，语音识别单元421可以通过ASR(自动语音识别技术，Automatic Speech Recognition)技术将语音转换为文本输入信息，并将该文本输入信息发送至中控模块410，中控模块410可以接收该文本输入信息，并将该文本输入信息发送至语义单元422，语义单元422根据该文本输入信息确定用户意图，并获取向用户反馈的文本输出信息。
进一步地，语音识别单元421可以通过基于语言学和声学方式、随机模型方式、人工神经网络方式、概率语法分析等方式对用户语音信息进行识别，此处对用户语音信息的方式不做具体限制。
进一步地，可以在语义单元422中设置意图识别模型和反馈平台，通过意图识别模型对文本输入信息进行意图识别，得到用户意图，将该意图输送至反馈平台，获取向用户反馈的文本输出信息。
示例性地，可以采用意图识别模型对用户信息进行意图识别，意图识别模型可以采用RNN模型、CNN模型、VAE模型、BERT、支持向量机等机器学习模型，在此不做限定。例如，意图识别模型还可以是采用上述机器学习模型的变体或组合等。另外，意图识别模型可以部署于智能终端，也可以部署于服务端，此处对意图识别模型的部署载体不做具体限制，具体以本实施例中获取用户意图的实际场景确定。另外，该反馈平台可以包括基于客服业务类型需求构建的问答库、客服话术库等，此处对反馈平台的构建不做具体限制。
需要注意的是，数字人的形态可以与文本一一对应，当获取到文本输出信息时，可以通过文本输出信息中的文本驱动数字人，使输出的数字人图像中数字人的形态与文本输出信息对应。
在本实施例中，通过在信息处理模块420中设置语义单元422和语音识别单元421，能够对用户的语音进行识别，得到文本输入信息，并基于文本输入信息得到用于向用户反馈的文本输出信息，并通过该文本输出信息驱动数字人，将向用户反馈的内容用数字人的形式进行呈现，从而模拟出用户与客服人员面对面交流的场景，提高用户的体验。
进一步地，作为本实施例的一种实施方式，反馈信息包括反馈语音信息；如图3所示，信息处理模块420可以包括语音识别单元421、文本反馈单元422和语音合成单元423。其中，语音识别单元421用于从通话数据中获取用户语音，并识别用户语音得到文本输入信息。文本反馈单元422用于根据文本输入信息确定用户意图，并基于用户意图获取文本输出信息。语音合成单元423用于对文本输出信息进行语音合成处理，得到反馈语音信息。
在本实施例中，语音合成单元423可以与中控模块410形成通信连接，语音识别单元421可以通过ASR技术将语音转换为文本输入信息，并将该文本输入信息发送至中控模块410，中控模块410可以接收该文本输入信息，并将该文本输入信息发送至语义单元422，语义单元422根据该文本输入信息确定用户意图，获取向用户反馈的文本输出信息，并将文本输出信息发送至中控模块410，中控模块410将该文本输出信息发送至语音合成单元423，语音合成单元423通过TTS(从文本到语音，Text To Speech)技术将该文本输出信息生成反馈语音信息。
进一步地，语音合成单元423可以通过参数合成法、波形合成法、规则合成法等方法对文本输出信息进行合成，得到反馈语音信息。
需要注意的是，数字人的形态可以与语音一一对应，当获取到反馈语音信息时，可以通过语音反馈信息中的语音驱动数字人，使输出的数字人图像中数字人的形态与语音反馈信息对应。即，反馈语音信息，用于驱动数字人图像中的数字人，使数字人图像中数字人的形态与反馈语音信息对应。
另外，在通过语音驱动数字人的过程中，可以通过语音控制数字人的嘴型参数。例如，可以获取数字人在发出与反馈语音信息对应的声音时数字人的嘴部关键点的变化情况，从而得到用于表征嘴部关键点的变化情况的嘴型参数。其中，嘴部关键点可以包括用于辨识、定位和控制数字人嘴部各个部分的位置。例如，嘴部关键点可以包括左嘴角、右嘴角、颏唇沟、鼻底等。同时，由于反馈语音信息中的语音可以为连续性语音，反馈语音信息可以具备时间属性，可以在反馈语音信息中语音对应的各个时间节点控制数字人的嘴型变化，从而能够在智能终端准确地呈现数字人的嘴型的变化过程。
在本实施例中，通过在信息处理模块420中设置语音识别单元421、文本反馈单元422和语音合成单元423，能够对用户的语音进行识别，得到文本输入信息，并基于文本输入信息得到用于向用户反馈的文本输出信息，并通过该文本输出信息，将该文本输出信息转换为反馈语音信息，通过反馈语音信息驱动数字人，从而模拟出用户与客服人员面对面交流的场景，提高用户的体验。
进一步地，为了能够更好地提示用户，可以在生成数字人图像的同时，生成电子卡片，如图3所示，图像生成模块440可以包括卡片单元441，卡片单元441用于根据反馈信息获取提示信息，生成包括提示信息的电子卡片，并将电子卡片发送至中控模块410，以使中控模块410将电子卡片作为交互反馈结果发送至智能终端。
在本实施例中，电子卡片可以用于展示提示内容(比如业务套餐、套餐费用等)。例如，电子卡片中可以当驱动信息为用户可办理的套餐包括套餐A、套餐B、套餐C时，提示内容可以为按照预设格式排布的套餐A、套餐B、套餐C，并将该提示内容呈现于电子卡片。
在本实施例中，电子卡片的卡片类型可以包括静态类型和动态类型。当电子卡片为静态类型时，卡片单元441可以生成一张电子卡片，使智能终端呈现该一张电子卡片，从而得到电子卡片处于静态的呈现效果；当电子卡片为动态类型时，卡片单元441可以生成多张电子卡片，该多张电子卡片可以为连续性的电子卡片，当智能终端按照电子卡片的时序呈现多张电子卡片时，可以呈现出变化播放电子卡片视频的效果。
在本实施例中，电子卡片可以与数字人图像结合后作为交互反馈结果发送至智能终端，也可以独立作为反馈结果发送至智能终端。
示例性地，当电子卡片与数字人图像结合后作为交互反馈结果发送至智能终端时，可以将电子卡片与数字人图像重叠，使电子卡片覆盖数字人图像的非核心区域，并将重叠后形成的图像作为交互反馈结果发送至智能终端。其中，非核心区域可以为数字人图像中数字人以外的区域，还可以为数字人图像中数字人头部以外的区域。此时由于电子卡片与数字人图像融合形成包括电子卡片和数字人的图像，因此，在将该图像通过数据包的形式传输至智能终 端的过程中，无需将电子卡片与数字人图像分开传输，同时，当智能终端在呈现交互反馈结果时，能够直接显示包括电子卡片和数字人的图像。
示例性地，当电子卡片作为反馈结果发送至智能终端时，可以将电子卡片与数字人图像分别通过数据包的形式传输至智能终端，当智能终端接收到电子卡片和数字人图像时，可以将电子卡片作为第一图层，将数字人图像作为第二图层，在智能终端需要同时显示第一图层与第二图层时，可以将第一图层置于顶层，将第二图层置于底层。
在本实施例中，通过在图像生成模块440中设置电子卡片生成单元131，可以通过电子卡片生成单元131生成电子卡片，客服服务系统400可以向智能终端发送数字人图像的同时发生电子卡片，使用户在与数字人客服进行高清通话过程中，可以同时查看电子卡片的提示内容，帮助用户更好地了解和办理业务。
进一步地，作为本实施例的一种实施方式，如图3所示，反馈业务模式还可以包括数字人离线模式，数字人离线模式用于表征在智能终端呈现不包含数字人的画面的模式；图像生成模块440还用于若反馈业务模式为数字人离线模式，则按照预设的信息排布规则将反馈信息生成推介图像，并将推介图像发送至中控模块410；其中，中控模块410还用于将推介图像作为交互反馈结果发送至智能终端。
其中，数字人离线模式可以包括向智能终端反馈包含推介图像的交互反馈结果的模式。例如，当反馈业务模式为数字人离线模式时，智能终端与本实施例提供的客服服务系统400建立通话连接，客服服务系统400可以向智能终端反馈交互反馈结果，该交互反馈结果可以包括连续性的多张推介图像，由此，智能终端可以呈现该多张推介图像，智能终终端所呈现的画面中不包含数字人，通过推介图像中的信息为用户提供客服服务，使得智能终端可以直接呈现包含反馈信息的画面，用户可以较为直观地获取到客服服务系统400反馈的信息，方便存在听觉障碍的用户获知信息，以实现多样性地为用户提供客服服务。
在本实施例中，可以先获取反馈信息所涉及的反馈业务类型，根据反馈业务类型获取排布模板，将反馈信息排布于排布模板，从而得到包含反馈信息的推介图像。其中，反馈业务类型可以用于表征反馈信息中的反馈内容的种类。例如，反馈业务类型可以包括消费流水、可办理套餐、套餐剩余使用量、优惠活动等。排布模板可以包括用于排布各种反馈业务类型的反馈信息的模板。
示例性地，获取反馈信息中的反馈内容，获取反馈内容所涉及到的反馈业务类型包括消费流水和优惠活动，获取能够排布反馈业务类型包括消费流水和优惠活动的排布模板，将与消费流水对应的反馈信息排布于排布模板中的第一区域，该第一区域用于排布反馈业务类型为消费流水的反馈信息，将与优惠活动对应的反馈信息排布于排布模板中的第二区域，该第二区域用于排布反馈业务类型为优惠活动的反馈信息。也就是说，排布模板中可以排布多种反馈业务类型的反馈信息。
在本实施例中，图像生成模块440可以在反馈业务模式处于数字人离线模式时，将反馈信息排布于推介图像，并将推介图像反馈至智能终端，从而能够在智能终端呈现出推介图像，便于用户直接查看推介图像中的反馈信息。
请参阅图4，本申请实施例提供一种客服服务方法，该客服服务方法可以应用于如图1所示的交互系统300。具体地，该客服服务方法可以包括以下步骤S11至步骤S15。
步骤S11：与智能终端建立通话连接。
在本实施例中，步骤S11的具体实施操作可以与上述实施例提供的相应操作类似，因此，步骤11可以参考上述中控模块410与智能终端建立通话连接的具体步骤，此处不再赘述。
步骤S12：获取与智能终端之间的通话数据，并根据通话数据确定向智能终端反馈的反馈信息。
在本实施例中，步骤S12的具体实施操作可以与上述实施例提供的相应操作类似，因此，步骤S12可以参考上述信息处理模块420获取中控模块410与智能终端之间的通话数据、根据通话数据确定向智能终端反馈的反馈信息的具体步骤，此处不再赘述。
步骤S13：从通话数据中获取用户信息，并根据用户信息确定反馈业务模式。
在本实施例中，反馈业务模式用于表征向智能终端发送反馈信息的业务模式，反馈业务模式可以包括数字人在线模式，数字人在线模式用于表征在智能终端呈现包含数字人的画面 的模式。
在本实施例中，步骤S13的具体实施操作可以与上述实施例提供的相应操作类似，因此，步骤13可以参考上述业务模式控制模块430从通话数据中获取用户信息、根据用户信息确定反馈业务模式的具体步骤，此处不再赘述。
步骤S14：若反馈业务模式为数字人在线模式，则基于反馈信息获取包括数字人的数字人图像，并将数字人图像作为交互反馈结果；其中，数字人图像中数字人的形态与反馈信息对应。
在本实施例中，步骤S14的具体实施操作可以与上述实施例提供的相应操作类似，因此，步骤13可以参考上述图像生成模块440基于反馈信息获取包括数字人的数字人图像、将数字人图像作为交互反馈结果的具体步骤，此处不再赘述。
步骤S15：将交互反馈结果发送至智能终端。
在本实施例中，步骤S15的具体实施操作可以与上述实施例提供的相应操作类似，因此，步骤13可以参考上述中控模块410将交互反馈结果发送至智能终端的具体步骤，此处不再赘述。
在本实施例中，通过上述步骤S11至步骤S15的实施，能够与智能终端形成视频通话，从通话数据中获取用户信息，基于用户信息确定反馈业务模式，并若反馈业务模式为在线数字人模式，则通过基于通话数据获得的反馈信息得到数字人图像，从而在智能终端显示数字人，当将本客服服务方法应用于客服服务时，能够有效提高用户的亲切度，同时可以代替人工服务，当处于咨询高峰期时，可以快速响应用户的咨询需求，同时，可以考虑到用户需求，选择对应的反馈业务模式向智能终端反馈信息，提升客服服务的形式多样性，从而满足不同用户的需求。
本申请实施例还提供一种客服服务方法，该客服服务方法可以应用于如图1所示的交互系统300，该客服服务方法可以包括以下步骤S21至步骤S25。在本实施例中提供的客服服务方法中，可以包括与上述实施例相同或相似的步骤，对于相同或相似步骤的执行，可参考前文介绍，本说明书不再一一赘述。
步骤S21：与智能终端建立通话连接。
进一步地，服务端可以向智能终端发出呼叫请求，当智能终端允许该呼叫请求后，服务端可以与智能终端建立通话连接；如图5所示，上述步骤S21可以包括以下步骤S211至步骤S213。
步骤S211：获取智能终端的终端标识。
在本实施例中，终端标识可以为用于建立高清通话连接的标识。例如，该终端标识可以为标准SIM卡号、迷你SIM卡号、Micro SIM卡号等。
步骤S212：通过终端标识向智能终端发出呼叫请求。
在本实施例中，可以基于终端标识进行拨号，从而实现向智能终端发出呼叫请求。
步骤S213：若确定智能终端允许呼叫请求，则与智能终端建立通话连接。
在本实施例中，当用户在智能终端进行操作，允许呼叫请求时，可以在智能终端生成指令，用于使智能终端与服务端建立通话连接。
进一步地，智能终端也可以向服务端发送发出呼叫请求，当服务端允许该呼叫请求后，智能终端可以与服务端建立通话连接；如图6所示，上述步骤S21还可以包括以下步骤S214至步骤S215。
步骤S214：接收由智能终端发送的呼叫请求。
在本实施例中，可以通过智能终端呼叫客服标识，进而向服务端发出呼叫请求。其中，客服标识可以根据业务需求设定。
步骤S215：若确定允许呼叫请求，则与智能终端建立通话连接。
在本实施例中，服务端既可以向智能终端发送呼叫请求与智能终端建立通话连接，也可以接收由智能终端发送的呼叫请求以与智能终端建立通话连接。
步骤S22：获取与智能终端之间的通话数据，并根据通话数据确定向智能终端反馈的反馈信息。
进一步地，为了便于获知用户需求，可以将用户可输入的字符与业务反馈信息对应；如 图7所示，上述步骤S22可以包括以下步骤S221至步骤S224。
步骤S221：从通话数据中获取键入指令，键入指令用于表征用户在智能终端界面操作所生成的指令。
在本实施例中，当用户在智能终端界面操作输入字符(例如数字、符号等)时，生成与输入的字符对应的指令，可以将该指令作为键入指令。
步骤S222：获取与键入指令对应的字符。
步骤S223：从通话数据中获取历史反馈信息，历史反馈信息包括相互关联的业务反馈信息和预设字符。
在本实施例中，历史反馈信息中可以包含业务反馈信息以及与业务反馈信息关联的预设字符。例如，历史反馈信息可以包括业务反馈信息为可用余额，预设字符为1，业务反馈信息为套餐剩余量，预设字符为2，将余额与预设字符“1”关联，将业务反馈信息与预设字符“2”关联。
步骤S224：若字符与预设字符匹配，则将与预设字符关联的业务反馈信息作为反馈信息。
在本实施例中，可以从通话数据中获取历史反馈信息，进而获得业务反馈信息与预设字符之间的关联关系，当获取到用户在智能终端输入的字符时，可以将获取到与该字符关联的业务反馈信息，从而实现快速响应用户的需求。
步骤S23：从通话数据中获取用户信息，并根据用户信息确定反馈业务模式，反馈业务模式用于表征向智能终端发送反馈信息的业务模式，反馈业务模式包括数字人在线模式，数字人在线模式用于表征在智能终端呈现包含数字人的画面的模式。
进一步地，作为本实施例的一种实施方式，为了确定反馈业务模式，可以通过用户图像确定反馈业务模式；如图8所示，上述步骤S23可以包括以下步骤S231至步骤S233。
步骤S231：从用户信息中获取包含用户的用户图像。
在本实施例中，可以在智能终端上设置用于采集图像的图像采集装置，通过该图像采集装置获取包含用户的用户图像；也可以在智能终端上设置用于接收其他设备发送的信息以获得用户图像。
步骤S232：对用户图像进行图像识别，得到用户特征。
在本实施例中，可以通过图像识别算法对用户图像进行识别，得到用户特征。其中，图像识别算法可以包括最大流量算法、匈牙利算法、广度优先搜索算法、深度优先搜索算法等，此处不做具体限制。
步骤S233：若用户特征与预设的用户特征相符，则确定反馈业务模式为数字人在线模式。
在本实施例中，若用户特征与预设的用户特征相符，则确定该用户为目标用户，即可确定反馈业务模式为数字人在线模式。此时，预设的用户特征的作用为用户身份验证，从而防止非法用户使用智能终端使用客服服务。另外，也可以预先将用户特征与反馈业务模式进行关联，当用户特征与预设的用户特征关联(相符)时，即可将与用户特征关联的业务模式确定为数字人在线模式或者数字人离线模式。此时，预设的用户特征的作用为模式匹配。
在本实施例中，通过步骤S231至步骤S233的实施，可通过用户图像确定反馈业务模式。
步骤S24：若反馈业务模式为数字人在线模式，则基于反馈信息获取包括数字人的数字人图像，并将数字人图像作为交互反馈结果；其中，数字人图像中数字人的形态与反馈信息对应。
进一步地，作为本实施例的一种实施方式，可以在用户选择指定的数字人，并在将包含该数字人的形态的图像发送至智能终端；如图9所示，上述步骤S24可以包括以下步骤S241至步骤S243。
步骤S241：根据反馈信息获取数字人选择指令和数字人驱动信息。
在本实施例中，数字人选择指令可以包括用于从多个数字人中选择出目标数字人的指令。数字人驱动信息可以包括用于驱动目标数字人的指令。
在本实施例中，可以将反馈信息与数字人进行关联，当与智能终端建立通话连接后，可以确定与该反馈信息关联的数字人，从而生成用于选择该数字人的数字人选择指令；也可以从反馈信息中获取用户所指定的数字人，并生成用于选择该数字人的数字人选择指令，数字人选择指令的表现形式与获取数字人选择指令对应。
在一些示例中，与智能终端建立通话连接的过程中，可以基于所接收到的反馈信息实时确定用户当前所指定的数字人，并对指定的数字人进行驱动，因此，在智能终端呈现的数字人可以是实时变化的不同的数字人。
步骤S242：按照预设的数字人选择规则获取与数字人选择指令对应的目标数字人。
在本实施例中，针对不同数字人选择指令的表现形式，可以设置对应的预设的数字人选择规则。
示例性地，当从反馈信息中获取目标数字人时，数字人选择指令可以是反馈信息所描述的数字人特征，按照该数字人选择指令从多个数字人中进行筛选，进而得到目标数字人。例如，数字人包括形似真人的仿真数字人和动漫效果式的仿真数字人，对每个仿真数字人进行预先编号，当反馈信息确定用户选择形似真人的仿真数字人时，获取形似真人的仿真数字人，按照编号的先后顺序从多个形似真人的仿真数字人中确定目标数字人。
示例性地，当从反馈信息中获取目标数字人时，数字人选择指令可以是反馈信息中用于表征用户所指定的数字人的信息。需要注意的是，反馈信息与数字人之间的关联关系可以在服务端预先设定。例如，可以将SIM卡号码与数字人关联，当智能终端通过该SIM卡号码与服务端建立通话连接后，将与该SIM卡号码关联的数字人作为目标数字人。
步骤S243：基于数字人驱动信息驱动目标数字人，获取包括目标数字人的目标数字人图像。
在本实施例中，基于数字人驱动信息获取目标数字人图像的方式与步骤S14中基于反馈信息获取数字人图像的方式大致相同，此处不再赘述。
在本实施例中，通过上述步骤S241至步骤S243的实施，可以基于用户需求选择相应的目标数字人，并驱动该目标数字人，从而在智能终端呈现包含该目标数字人的数字人图像，从而能够给用户带来更加丰富的体验。
步骤S25：将交互反馈结果发送至智能终端。
进一步地，作为本实施例的一种实施方式，为了防止反馈交互结果中某些信息泄露，如图10所示，上述步骤S25可以包括以下步骤S251至步骤253。
步骤S251：获取交互反馈结果中的隐私信息。
在本实施中，隐私信息可以包括涉及用户隐私的信息。例如，该隐私信息可以是身份证号码、用户余额、拨号记录等。
在一些示例中，可以基于客服反馈结果中的内容类型，采取不同的方式获取客服反馈结果中的隐私信息。
示例性地，可以获取该电子卡片中所包含的文本输出信息，并将该文本输出信息中涉及用户隐私的信息作为隐私信息。当客服反馈结果包括语音反馈信息时，可以获取与该语音反馈信息对应的文本输出信息，并将该文本输出信息中涉及用户隐私的信息作为隐私信息。
步骤S252：按照预设的信息保护方式更新交互反馈结果，使更新后的交互反馈结果中的隐私信息被隐藏。
在本实施例中，预设的信息保护方式可以基于实际场景设定。
示例性地，当隐私信息为电子卡片中的文本时，可以隐藏该文本中一个或多个字符。例如，当该电子卡片中的文本为身份证号码，该身份证号码为110101200108150612，可以更新该身份证号码，更新后的电子卡片中的文本为1101012001XXXXXXXX。
当隐私信息为反馈语音信息时，可以获取与该反馈语音信息对应的文本输出信息，若该文本输出信息为用户余额，可以将用户余额中的数字与字母对应，并将对应后字母作为更新后的客服反馈结果。例如，可以将数字“1”与字母“A”对应，将数字“2”与字母“B”对应，将数字“3”与字母“C”对应，将数字“4”与字母“D”对应，将数字“5”与字母“E”对应，将数字“6”与字母“F”对应，当用户余额为12345时，可以将该用户余额更新为“ABCDE”，并将与“ABCDE”对应的语音作为更新后的客服反馈结果。其中，数字与字母对应的方式可以为用户预先设定。
步骤S253：将更新后的交互反馈结果发送至智能终端。
在本实施例中，当将更新后的客服反馈结果发送至智能终端后，智能终端能够播放该更新后的客服反馈结果，从而对客服反馈结果中的隐私信息进行保护。
本申请实施例还提供一种客服服务方法，该客服服务方法可以应用于如图1所示的交互系统300，该客服服务方法可以包括以下步骤(1)至(5)。在本实施例中提供的客服服务方法中，可以包括与上述实施例相同或相似的步骤，对于相同或相似步骤的执行，可参考前文介绍，本说明书不再一一赘述。
(1)：与智能终端建立通话连接。
(2)：获取与智能终端之间的通话数据，根据通话数据确定向智能终端反馈的反馈信息。
(3)：从通话数据中获取用户信息，并根据用户信息确定反馈业务模式，反馈业务模式用于表征向智能终端发送反馈信息的业务模式，反馈业务模式包括数字人在线模式，数字人在线模式用于表征在智能终端呈现包含数字人的画面的模式。
(4)：若反馈业务模式为数字人在线模式，则基于反馈信息获取包括数字人的数字人图像，并将数字人图像作为交互反馈结果；其中，数字人图像中数字人的形态与反馈信息对应。
(5)：将交互反馈结果发送至智能终端。
进一步地，作为本实施例的一种实施方式，反馈业务模式还可以包括数字人离线模式，数字人离线模式用于表征在智能终端呈现不包含数字人的画面的模式；本实施例提供的客服服务方法还可以包括：若反馈业务模式为数字人离线模式，则按照预设的信息排布规则将反馈信息生成推介图像，并将推介图像作为交互反馈结果。
其中，数字人离线模式可以包括向智能终端反馈包含推介图像的交互反馈结果的模式。例如，当反馈业务模式为数字人离线模式时，智能终端与本实施例提供的客服服务系统400建立通话连接，客服服务系统400可以向智能终端反馈交互反馈结果，该交互反馈结果可以包括连续性的多张推介图像，由此，智能终端可以呈现该多张推介图像，智能终终端所呈现的画面中不包含数字人，通过推介图像中的信息为用户提供客服服务，使得智能终端可以直接呈现包含反馈信息的画面，用户可以较为直观地获取到客服服务系统400反馈的信息，方便存在听觉障碍的用户获知信息，以实现多样性地为用户提供客服服务。
在本实施例中，可以先获取反馈信息所涉及的反馈业务类型，根据反馈业务类型获取排布模板，将反馈信息排布于排布模板，从而得到包含反馈信息的推介图像。其中，反馈业务类型可以用于表征反馈信息中的反馈内容的种类。例如，反馈业务类型可以包括消费流水、可办理套餐、套餐剩余使用量、优惠活动等。排布模板可以包括用于排布各种反馈业务类型的反馈信息的模板。
进一步地，为了有序排布反馈信息，可以将反馈信息排布至目标模板；按照预设的信息排布规则将反馈信息生成推介图像可以包括：根据反馈信息从预设的多个模板中获取目标模板；将反馈信息排布至目标模板，得到推介图像。
在本实施例中，目标模板可以包括多个用于排布反馈信息的区域，其中，可以将该区域与反馈信息所涉及的反馈业务类型一一对应，该区域用于排布反馈信息中与该区域对应的反馈信息。
示例性地，获取反馈信息中的反馈内容，获取反馈内容所涉及到的反馈业务类型包括消费流水和优惠活动，获取能够排布反馈业务类型包括消费流水和优惠活动的排布模板，将与消费流水对应的反馈信息排布于排布模板中的第一区域，该第一区域用于排布反馈业务类型为消费流水的反馈信息，将与优惠活动对应的反馈信息排布于排布模板中的第二区域，该第二区域用于排布反馈业务类型为优惠活动的反馈信息。也就是说，排布模板中可以排布多种反馈业务类型的反馈信息。
进一步地，为了能够更好地提示用户，可以在生成数字人图像的同时，生成电子卡片；本实施例提供的客服服务方法还可以包括：根据反馈信息获取用于在智能终端显示的提示内容；生成包括提示内容的电子卡片，并将电子卡片作为交互反馈结果。
在本实施例中，电子卡片可以用于展示提示内容(比如业务套餐、套餐费用等)。例如，电子卡片中可以当驱动信息为用户可办理的套餐包括套餐A、套餐B、套餐C时，提示内容可以为按照预设格式排布的套餐A、套餐B、套餐C，并将该提示内容呈现于电子卡片。
在本实施例中，电子卡片的卡片类型可以包括静态类型和动态类型。当电子卡片为静态类型时，可以生成一张电子卡片，使智能终端呈现该一张电子卡片，从而得到电子卡片处于静态的呈现效果；当电子卡片为动态类型时，可以生成多张电子卡片，该多张电子卡片可以 为连续性的电子卡片，当智能终端按照电子卡片的时序呈现多张电子卡片时，可以呈现出变化播放电子卡片视频的效果。
在本实施例中，电子卡片可以与数字人图像结合后作为交互反馈结果发送至智能终端，也可以独立作为反馈结果发送至智能终端。
示例性地，当电子卡片与数字人图像结合后作为交互反馈结果发送至智能终端时，可以将电子卡片与数字人图像重叠，使电子卡片覆盖数字人图像的非核心区域，并将重叠后形成的图像作为交互反馈结果发送至智能终端。其中，非核心区域可以为数字人图像中数字人以外的区域，还可以为数字人图像中数字人头部以外的区域。此时由于电子卡片与数字人图像融合形成包括电子卡片和数字人的图像，因此，在将该图像通过数据包的形式传输至智能终端的过程中，无需将电子卡片与数字人图像分开传输，同时，当智能终端在呈现交互反馈结果时，能够直接显示包括电子卡片和数字人的图像。
示例性地，当电子卡片作为反馈结果发送至智能终端时，可以将电子卡片与数字人图像分别通过数据包的形式传输至智能终端，当智能终端接收到电子卡片和数字人图像时，可以将电子卡片作为第一图层，将数字人图像作为第二图层，在智能终端需要同时显示第一图层与第二图层时，可以将第一图层置于顶层，将第二图层置于底层。
在本实施例中，可向智能终端发送数字人图像的同时发生电子卡片，使用户在与数字人客服进行高清通话过程中，可同时查看电子卡片的提示内容，帮助用户更好了解和办理业务。
所属领域的技术人员可以清楚地了解到，为描述的方便和简洁，上述描述交互系统的具体工作过程，也可以参考前述客服服务方法实施例中的对应过程，在此不再赘述。
请参阅图11，示出了本申请实施例提供的一种客服服务装置的结构框图，该客服服务装置可以包括连接模块61、反馈信息确定模块62、业务模式确定模块63、数字人驱动模块64和交互反馈结果发送模块65。其中，连接模块61用于与智能终端建立通话连接。反馈信息确定模块62用于获取与智能终端之间的通话数据，并根据通话数据确定向智能终端反馈的反馈信息。业务模式确定模块63用于从通话数据中获取用户信息，并根据用户信息确定反馈业务模式。数字人驱动模块64用于若反馈业务模式为数字人在线模式，则基于反馈信息获取可以包括数字人的数字人图像，并将数字人图像作为交互反馈结果；其中，数字人图像中数字人的形态与反馈信息对应。交互反馈结果发送模块65用于将交互反馈结果发送至智能终端。
进一步地，作为本实施例的一种实施方式，连接模块61可以包括终端标识获取单元、呼叫单元和第一连接单元。其中，终端标识获取单元用于获取智能终端的终端标识。呼叫单元用于通过终端标识向智能终端发出呼叫请求。第一连接单元用于若确定智能终端允许呼叫请求，则与智能终端建立通话连接。
进一步地，作为本实施例的一种实施方式，连接模块61可以包括呼叫接收单元和第二连接单元。其中，呼叫接收单元用于接收由智能终端发送的呼叫请求。第二连接单元用于若确定允许呼叫请求，则与智能终端建立通话连接。
进一步地，作为本实施例的一种实施方式，数字人驱动模块64可以包括指令获取单元、目标数字人获取单元和目标数字人图像获取单元。其中，指令获取单元用于根据反馈信息获取数字人选择指令和数字人驱动信息。目标数字人获取单元用于按照预设的数字人选择规则获取与数字人选择指令对应的目标数字人。目标数字人图像获取单元用于基于数字人驱动信息驱动目标数字人，获取可以包括目标数字人的目标数字人图像。
进一步地，作为本实施例的一种实施方式，客服服务装置还可以包括数字人离线处理模块。其中，数字人离线处理模块用于若反馈业务模式为数字人离线模式，则按照预设的信息排布规则将反馈信息生成推介图像，并将推介图像作为交互反馈结果。
进一步地，作为本实施例的一种实施方式，数字人离线处理模块可以包括目标模板获取单元和推介图像获取单元。其中，目标模板获取单元用于根据反馈信息从预设的多个模板中获取目标模板。推介图像获取单元用于将反馈信息排布至目标模板，得到推介图像。
进一步地，作为本实施例的一种实施方式，反馈信息确定模块62可以包括键入指令获取单元、字符获取单元、历史反馈信息获取单元和字符关联单元。其中，键入指令获取单元用于从通话数据中获取键入指令，键入指令用于表征用户在智能终端界面操作所生成的指令。字符获取单元用于获取与键入指令对应的字符。历史反馈信息获取单元用于从通话数据中获 取历史反馈信息，历史反馈信息可以包括相互关联的业务反馈信息和预设字符。字符关联单元用于若字符与预设字符匹配，则将与预设字符关联的业务反馈信息作为反馈信息。
进一步地，作为本实施例的一种实施方式，业务模式确定模块63可以包括用户图像获取单元、用户特征获取单元和在线模式确定单元。其中，用户图像获取单元用于从用户信息中获取包含用户的用户图像。用户特征获取单元用于对用户图像进行图像识别，得到用户特征。在线模式确定单元用于若用户特征与预设的用户特征相符，则确定反馈业务模式为数字人在线模式。
进一步地，作为本实施例的一种实施方式，交互反馈结果发送模块65可以包括隐私信息获取单元、交互反馈结果更新单元和交互反馈结果发送单元。其中，隐私信息获取单元用于获取交互反馈结果中的隐私信息。交互反馈结果更新单元用于按照预设的信息保护方式更新交互反馈结果，使更新后的交互反馈结果中的隐私信息被隐藏。交互反馈结果发送单元用于将更新后的交互反馈结果发送至智能终端。
进一步地，作为本实施例的一种实施方式，客服服务装置还可以包括提示内容获取模块和交互反馈结果生成模块。其中，提示内容获取模块用于根据反馈信息获取用于在智能终端显示的提示内容。交互反馈结果生成模块用于生成可以包括提示内容的电子卡片，并将电子卡片作为交互反馈结果。
请参阅图12，其示出了本申请实施例提供的一种电子设备，该电子设备可以包括处理器810、通信模块820、存储器830和总线。总线可以是ISA总线、PCI总线、EISA总线或CAN总线等。总线可以分为地址总线、数据总线、控制总线等。其中：
存储器830，用于存放计算机可读指令。具体地，存储器830可用于存储软件计算机可读指令以及各种数据。存储器830可主要包括存储计算机可读指令区和存储数据区，其中，存储计算机可读指令区可存储操作至少一个功能所需的计算机可读指令可以包括计算机可读指令，计算机可读指令包括计算机操作指令。除了存放计算机可读指令之外，存储器830还可以暂存通信模块820需要发送的消息等。存储器830可能包含高速RAM存储器，也可能还包括非易失性存储器(non-volatile memory)，例如至少一个固态硬盘(Solid State Disk，简称SSD)。
处理器810用于执行存储器830存放的计算机可读指令。计算机可读指令被处理器执行时实现上述各实施例的客服服务方法的步骤。
本申请实施例还提供一个或多个计算机可读存储介质，计算机可读存储介质上存储有计算机可读指令，该计算机可读指令被处理器执行时实现上述各实施例的客服服务方法的各个过程，且能达到相同的技术效果，为避免重复，这里不再赘述。其中，的计算机可读存储介质，如只读存储器(Read-Only Memory，简称ROM)、随机存取存储器(Random Access Memory，简称RAM)、SSD、带电可擦可编程只读存储器(Electrically Erasable Programmable read only memory，简称EEPROM)或快闪存储器(Flash Memory，简称Flash)等。
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、方法、物品或者装置不仅包括那些要素，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、方法、物品或者装置所固有的要素。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、方法、物品或者装置中还存在另外的相同要素。
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的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一个存储介质(如ROM/RAM、SSD、Flash)中，包括若干计算机可读指令用以使得一台终端(可以是手机，计算机，服务器，空调器，或者网络设备等)执行本申请各个实施例的方法。
最后应说明的是：以上实施例仅用以说明本申请的技术方案，而非对其限制；尽管参照前述实施例对本申请进行了详细的说明，本领域的普通技术人员当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分技术特征进行等同替换；而这些修改或者替换，并不驱使相应技术方案的本质脱离本申请各实施例技术方案的精神和范围。

Claims (18)

  1. 一种客服服务系统，其特征在于，包括：
    中控模块，用于与智能终端建立通话连接；
    信息处理模块，用于获取所述中控模块与所述智能终端之间的通话数据，并根据所述通话数据确定向所述智能终端反馈的反馈信息；
    业务模式控制模块，用于从所述通话数据中获取用户信息，并根据所述用户信息确定反馈业务模式；以及
    图像生成模块，用于若所述反馈业务模式为数字人在线模式，则基于所述反馈信息获取包括所述数字人的数字人图像，并将所述数字人图像作为交互反馈结果发送至所述中控模块；其中，所述数字人图像中所述数字人的形态与所述反馈信息对应，所述中控模块还用于将所述交互反馈结果发送至所述智能终端。
  2. 根据权利要求1所述的客服服务系统，其特征在于，所述反馈业务模式还包括数字人离线模式，所述数字人离线模式用于表征在所述智能终端呈现不包含所述数字人的画面的模式；所述图像生成模块还用于若所述反馈业务模式为所述数字人离线模式，则按照预设的信息排布规则基于所述反馈信息生成推介图像，并将所述推介图像发送至所述中控模块；其中，所述中控模块还用于将所述推介图像作为所述交互反馈结果发送至所述智能终端。
  3. 根据权利要求1所述的客服服务系统，其特征在于，所述图像生成模块包括卡片单元，所述卡片单元用于根据所述反馈信息获取提示信息，生成包括所述提示信息的电子卡片，并将所述电子卡片发送至所述中控模块，以使所述中控模块将所述电子卡片作为所述交互反馈结果发送至所述智能终端。
  4. 根据权利要求1至3中任一项所述的客服服务系统，其特征在于，所述信息处理模块包括：
    语音识别单元，用于从所述通话数据中获取用户语音，并识别所述用户语音得到文本输入信息；以及
    文本反馈单元，用于根据所述文本输入信息确定用户意图，并基于所述用户意图获取文本输出信息；所述文本输出信息用于确定所述反馈信息。
  5. 根据权利要求4所述的客服服务系统，其特征在于，所述反馈信息包括反馈语音信息；所述信息处理模块还包括：
    语音合成单元，用于对所述文本输出信息进行语音合成处理，得到反馈语音信息。
  6. 一种客服服务方法，其特征在于，由电子设备执行，包括：
    与所述智能终端建立通话连接；
    获取与所述智能终端之间的通话数据，并根据所述通话数据确定向所述智能终端反馈的反馈信息；
    从所述通话数据中获取用户信息，并根据所述用户信息确定反馈业务模式；以及
    若所述反馈业务模式为数字人在线模式，则基于所述反馈信息获取包括所述数字人的数字人图像，并将所述数字人图像作为交互反馈结果；其中，所述数字人图像中所述数字人的形态与所述反馈信息对应；以及
    将所述交互反馈结果发送至所述智能终端。
  7. 根据权利要求6所述的客服服务方法，其特征在于，所述与所述智能终端建立通话连接包括：
    获取所述智能终端的终端标识；
    通过所述终端标识向所述智能终端发出呼叫请求；以及
    若确定所述智能终端允许所述呼叫请求，则与所述智能终端建立通话连接。
  8. 根据权利要求6所述的客服服务方法，其特征在于，所述与所述智能终端建立通话连接包括：
    接收由所述智能终端发送的呼叫请求；以及
    若确定允许所述呼叫请求，则与所述智能终端建立通话连接。
  9. 根据权利要求6所述的客服服务方法，其特征在于，所述基于所述反馈信息获取包括所述数字人的数字人图像包括：
    根据所述反馈信息获取数字人选择指令和数字人驱动信息；
    按照预设的数字人选择规则获取与所述数字人选择指令对应的目标数字人；以及
    基于所述数字人驱动信息驱动所述目标数字人，获取包括所述目标数字人的目标数字人图像。
  10. 根据权利要求6所述的客服服务方法，其特征在于，所述客服服务方法还包括：
    若所述反馈业务模式为所述数字人离线模式，则按照预设的信息排布规则基于所述反馈信息生成推介图像，并将所述推介图像作为所述交互反馈结果。
  11. 根据权利要求10所述的客服服务方法，其特征在于，所述按照预设的信息排布规则基于所述反馈信息生成推介图像包括：
    根据所述反馈信息从预设的多个模板中获取目标模板；以及
    将所述反馈信息排布至所述目标模板，得到所述推介图像。
  12. 根据权利要求6所述的客服服务方法，其特征在于，所述根据所述通话数据确定向所述智能终端反馈的反馈信息包括：
    从所述通话数据中获取键入指令，所述键入指令用于表征用户在所述智能终端界面操作所生成的指令；
    获取与所述键入指令对应的字符；
    从所述通话数据中获取历史反馈信息，所述历史反馈信息包括相互关联的业务反馈信息和预设字符；以及
    若所述字符与所述预设字符匹配，则将与所述预设字符关联的所述业务反馈信息作为反馈信息。
  13. 根据权利要求6所述的客服服务方法，其特征在于，所述根据所述用户信息确定反馈业务模式包括：
    从所述用户信息中获取包含用户的用户图像；
    对所述用户图像进行图像识别，得到用户特征；以及
    若所述用户特征与预设的用户特征相符，则确定所述反馈业务模式为所述数字人在线模式。
  14. 根据权利要求6至13中任一项所述的客服服务方法，其特征在于，所述将所述交互反馈结果发送至所述智能终端包括：
    获取所述交互反馈结果中的隐私信息；
    按照预设的信息保护方式更新所述交互反馈结果，使更新后的交互反馈结果中的所述隐私信息被隐藏；以及
    将更新后的交互反馈结果发送至所述智能终端。
  15. 根据权利要求6至13中任一项所述的客服服务方法，其特征在于，所述客服服务方法还包括：
    根据所述反馈信息获取用于在所述智能终端显示的提示内容；以及
    生成包括所述提示内容的电子卡片，并将所述电子卡片作为所述交互反馈结果。
  16. 一种客服服务装置，其特征在于，设置于电子设备中，包括：
    连接模块，用于与所述智能终端建立通话连接；
    反馈信息确定模块，用于获取与所述智能终端之间的通话数据，并根据所述通话数据确定向所述智能终端反馈的反馈信息；
    业务模式确定模块，用于从所述通话数据中获取用户信息，并根据所述用户信息确定反馈业务模式；
    数字人驱动模块，用于在所述反馈业务模式为数字人在线模式时，基于所述反馈信息获取包括所述数字人的数字人图像，并将所述数字人图像作为交互反馈结果；其中，所述数字人图像中所述数字人的形态与所述反馈信息对应；以及
    交互反馈结果发送模块，用于将所述交互反馈结果发送至所述智能终端。
  17. 一种电子设备，其特征在于，包括：
    一个或多个处理器；
    存储器；
    一个或多个计算机可读指令，其中所述一个或多个计算机可读指令被存储在所述存储器中并被配置为由所述一个或多个处理器执行，所述一个或多个计算机可读指令配置用于执行如权利要求6至15中任一项所述的客服服务方法。
  18. 一个或多个计算机可读取存储介质，其特征在于，所述计算机可读取存储介质中存储有计算机可读指令，所述计算机可读指令可被处理器调用执行如权利要求6至15中任一项所述的客服服务方法。
PCT/CN2021/123865 2021-03-16 2021-10-14 客服服务系统、方法、装置、电子设备及存储介质 WO2022193635A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110278320.3A CN112669416B (zh) 2021-03-16 2021-03-16 客服服务系统、方法、装置、电子设备及存储介质
CN202110278320.3 2021-03-16

Publications (1)

Publication Number Publication Date
WO2022193635A1 true WO2022193635A1 (zh) 2022-09-22

Family

ID=75399330

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/123865 WO2022193635A1 (zh) 2021-03-16 2021-10-14 客服服务系统、方法、装置、电子设备及存储介质

Country Status (2)

Country Link
CN (1) CN112669416B (zh)
WO (1) WO2022193635A1 (zh)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112669416B (zh) * 2021-03-16 2021-06-22 深圳追一科技有限公司 客服服务系统、方法、装置、电子设备及存储介质
CN114040142A (zh) * 2021-11-05 2022-02-11 深圳壹账通智能科技有限公司 智能外呼机器人的视频通话方法、装置、设备及存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108268555A (zh) * 2017-01-03 2018-07-10 中国移动通信有限公司研究院 一种信息处理方法及装置
CN110348387A (zh) * 2019-07-12 2019-10-18 腾讯科技（深圳）有限公司 一种图像数据处理方法、装置以及计算机可读存储介质
CN110400251A (zh) * 2019-06-13 2019-11-01 深圳追一科技有限公司 视频处理方法、装置、终端设备及存储介质
CN110647636A (zh) * 2019-09-05 2020-01-03 深圳追一科技有限公司 交互方法、装置、终端设备及存储介质
CN112669416A (zh) * 2021-03-16 2021-04-16 深圳追一科技有限公司 客服服务系统、方法、装置、电子设备及存储介质

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108848281B (zh) * 2018-06-13 2020-11-24 上海携程商务有限公司 一种客服方法、系统、设备及存储介质
CN110826441B (zh) * 2019-10-25 2022-10-28 深圳追一科技有限公司 交互方法、装置、终端设备及存储介质

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108268555A (zh) * 2017-01-03 2018-07-10 中国移动通信有限公司研究院 一种信息处理方法及装置
CN110400251A (zh) * 2019-06-13 2019-11-01 深圳追一科技有限公司 视频处理方法、装置、终端设备及存储介质
CN110348387A (zh) * 2019-07-12 2019-10-18 腾讯科技（深圳）有限公司 一种图像数据处理方法、装置以及计算机可读存储介质
CN110647636A (zh) * 2019-09-05 2020-01-03 深圳追一科技有限公司 交互方法、装置、终端设备及存储介质
CN112669416A (zh) * 2021-03-16 2021-04-16 深圳追一科技有限公司 客服服务系统、方法、装置、电子设备及存储介质

Also Published As

Publication number Publication date
CN112669416B (zh) 2021-06-22
CN112669416A (zh) 2021-04-16

Similar Documents

Publication Publication Date Title
US7065711B2 (en) Information processing device and method, and recording medium
RU2488232C2 (ru) Сеть связи и устройства для преобразования текста в речь и текста в анимацию лица
WO2022193635A1 (zh) 客服服务系统、方法、装置、电子设备及存储介质
KR101992424B1 (ko) 증강현실용 인공지능 캐릭터의 제작 장치 및 이를 이용한 서비스 시스템
CN107704169B (zh) 虚拟人的状态管理方法和系统
CN110400251A (zh) 视频处理方法、装置、终端设备及存储介质
CN109521927B (zh) 机器人互动方法和设备
CN108833941A (zh) 人机交互处理方法、装置、用户终端、处理服务器及系统
KR20220027187A (ko) 장면 인터랙션 방법 및 장치, 전자 장치 및 컴퓨터 저장 매체
CN111290568A (zh) 交互方法、装置及计算机设备
CN112669846A (zh) 交互系统、方法、装置、电子设备及存储介质
CN110794964A (zh) 虚拟机器人的交互方法、装置、电子设备及存储介质
CN110287947A (zh) 互动课堂中的互动教室确定方法及装置
CN112669422A (zh) 仿真3d数字人生成方法、装置、电子设备及存储介质
KR102412823B1 (ko) 번역 기능을 제공하는 실시간 양방향 온라인 회의 시스템
CN113850898A (zh) 场景渲染方法及装置、存储介质及电子设备
JP2007030050A (ja) ロボット制御装置、ロボット制御システム、ロボット装置、およびロボット制御方法
JP7096626B2 (ja) 情報抽出装置
CN113050791A (zh) 交互方法、装置、电子设备及存储介质
KR20190002386A (ko) 캐릭터 서비스 시스템에서의 캐릭터 서비스를 제공하기 위한 장치
JP7130290B2 (ja) 情報抽出装置
CN113157241A (zh) 交互设备、交互装置及交互系统
CN111696182A (zh) 一种虚拟主播生成系统、方法和存储介质
CN113031768A (zh) 客服服务方法、装置、电子设备及存储介质
JP2021043723A (ja) 情報処理装置、情報処理方法およびプログラム

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: 21931210

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE