WO2023159525A1 - 客户服务方法、装置、系统及存储介质 - Google Patents

客户服务方法、装置、系统及存储介质 Download PDF

Info

Publication number
WO2023159525A1
WO2023159525A1 PCT/CN2022/078111 CN2022078111W WO2023159525A1 WO 2023159525 A1 WO2023159525 A1 WO 2023159525A1 CN 2022078111 W CN2022078111 W CN 2022078111W WO 2023159525 A1 WO2023159525 A1 WO 2023159525A1
Authority
WO
WIPO (PCT)
Prior art keywords
target customer
customer
target
image
staff
Prior art date
Application number
PCT/CN2022/078111
Other languages
English (en)
French (fr)
Inventor
钟盼
许景涛
Original Assignee
京东方科技集团股份有限公司
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 京东方科技集团股份有限公司 filed Critical 京东方科技集团股份有限公司
Priority to CN202280000320.XA priority Critical patent/CN116964610A/zh
Priority to PCT/CN2022/078111 priority patent/WO2023159525A1/zh
Publication of WO2023159525A1 publication Critical patent/WO2023159525A1/zh

Links

Images

Classifications

    • 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
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising

Definitions

  • the present disclosure relates to the field of data processing, and in particular to a customer service method, device, system and storage medium.
  • related technologies generally acquire customer face images and detect whether there is customer identity information matching the customer face images in the database based on face recognition technology. If there is matching customer identity information in the database, assign corresponding staff to the customer according to the identity information.
  • a customer service method includes: determining the overall image and partial image of the target customer; determining the group attribute of the target customer according to the overall image; the group attribute is used to represent the type of the group to which the target customer belongs;
  • the method further includes: sending the customer information of the target customer and the current location information of the target customer to the service terminal; determining the first positions of the staff and the target customer according to the overall image of the staff and the overall image of the target customer Distance: when the distance from the first location is greater than the first preset distance and the duration is greater than the first duration, send service prompt information to the service terminal.
  • the method further includes: receiving registration information of the target customer; the registration information includes partial images of the target customer and partial images of accompanying personnel of the target customer.
  • the method includes: determining the overall image of the accompanying person according to the partial image of the accompanying person; determining the second position distance between the target customer and the accompanying person according to the overall image of the target customer and the overall image of the accompanying person; When the location distance is greater than the second preset distance, according to the service priority of the target customer and the type of staff corresponding to the target customer, assign staff to the target customer.
  • the method includes: determining a target staff group, the type of the target staff group is the type of staff corresponding to the target customer, and the target staff group includes a plurality of target staff;
  • the personnel attribute of the group determines the personnel priority of the target staff group; assigns the target staff to the target customer according to the service priority of the target customer and the personnel priority of the target staff group.
  • the method includes: acquiring an image to be detected; the image to be detected includes a target customer; determining a first image area in the image to be detected, and a first value and a second value corresponding to the first image area; the first The numerical value indicates the probability that the first image area includes the overall image; the second numerical value indicates the probability that the first image area includes a partial image; the first image area is any area in the image to be detected; when the first numerical value is greater than the first threshold In this case, it is determined that the first image area includes the overall image of the target customer; when the second value is greater than the second threshold, it is determined that the first image area includes a partial image of the target customer.
  • the method includes: inputting the overall image into the group attribute prediction model, determining the probability of different group attributes of the target customer; and determining the group attribute of the target customer according to the probabilities of different group attributes of the target customer.
  • different group attributes include at least one of mutually exclusive group attributes and independent group attributes; the method includes: under the condition that different group attributes include mutually exclusive group attributes, determining the group of target customers The attributes include the first mutually exclusive group attribute; the first mutually exclusive group attribute is the group attribute with the highest probability among the mutually exclusive group attributes; when different group attributes include independent group attributes, it is determined that the group attribute of the target customer includes the first Independent group attributes; the first independent group attribute is a group attribute whose probability is greater than the attribute probability threshold among the independent group attributes.
  • the method includes: inputting partial images into the individual attribute prediction model, determining the gender probability corresponding to each gender of the target customer and the age probability corresponding to each age; determining the target customer's gender as the one with the highest gender probability Gender; determine the age of the target customer as the age with the greatest age probability.
  • the method includes: detecting whether the target customer exists in the user database according to the partial image of the target customer; and obtaining the customer level of the target customer if the target customer exists in the user database.
  • the method further includes: if the target customer exists in the user database, obtaining the basic information of the target customer; sending the basic information of the target customer to the service terminal; the service terminal is a terminal used by the staff; When the customer authorizes to query the historical business, obtain the historical business of the target customer; send the historical business of the target customer to the service terminal.
  • the method includes: determining the attribute factor of the target customer according to group attributes and individual attributes; the attribute factor is used to represent the influence degree of the attribute of the target customer on the service priority; obtaining the queue information of the target customer, and according to the queue The information determines the queue factor; the queue factor is used to indicate the influence degree of the target customer's queue information on the service priority; the target customer's reservation business is obtained, and the business factor is determined according to the reservation business; the business factor is used to represent the target customer's reservation business on the service The degree of influence of priority; determine the service priority of target customers according to attribute factors, queue factors and business factors.
  • the method further includes: determining a plurality of overall images of the target customer at different moments in the images collected by the multiple image acquisition devices; and determining the route trajectory of the target customer according to the position information of the multiple overall images.
  • the method further includes: determining the residence time of the target customer in at least one specific area according to the route track of the target customer; at least one specific area corresponds to at least one business; determining the target customer's concerned business as the residence time The business corresponding to the longest specific area; sending business information of concerned business to the service terminal; the service terminal is the terminal used by the staff.
  • a customer service device including a processing unit configured to determine the overall image and partial image of the target customer; the processing unit is also configured to determine the group attribute of the target customer according to the overall image; the group attribute is used to represent The type of the group to which the target customer belongs; the processing unit is also configured to determine the individual attributes of the target customer according to the partial image; the individual attributes include at least one of gender, age, and customer level; the processing unit is also configured to determine the target customer according to the group of the target customer attribute and the individual attributes of the target customer, determine the service priority of the target customer and the staff type corresponding to the target customer; the processing unit is also configured to, according to the service priority of the target customer and the staff type corresponding to the target customer, as Assign staff to target customers.
  • the communication unit is configured to send the customer information of the target customer and the current location information of the target customer to the service terminal; the processing unit is configured to determine the staff member according to the overall image of the staff member and the overall image of the target customer The first location distance from the target customer; the communication unit is also configured to send service prompt information to the service terminal when the first location distance is greater than the first preset distance and the duration is longer than the first duration.
  • the communication unit is configured to receive registration information of the target customer; the registration information includes partial images of the target customer and partial images of accompanying persons of the target customer.
  • the processing unit is configured to determine the overall image of the accompanying person according to the partial image of the accompanying person; the processing unit is configured to determine the relationship between the target customer and the accompanying person according to the overall image of the target customer and the overall image of the accompanying person The second location distance; a processing unit configured to assign staff to the target customer according to the service priority of the target customer and the staff type corresponding to the target customer when the second location distance is greater than the second preset distance.
  • the processing unit is configured to determine a target staff group, the type of the target staff group is the staff type corresponding to the target customer, and the target staff group includes a plurality of target staff; the processing unit, It is also configured to determine the personnel priority of the target staff group according to the personnel attributes of the target staff group; the processing unit is also configured to determine the personnel priority of the target staff group according to the service priority of the target customer and the personnel priority of the target staff group, Assign target workers to target accounts.
  • the communication unit is configured to acquire the image to be detected; the image to be detected includes the target customer; the processing unit is configured to determine the first image area in the image to be detected, and the first image area corresponding to the first image area numerical value and second numerical value; the first numerical value represents the probability that the first image area includes the overall image; the second numerical value represents the probability that the first image area includes the partial image; the first image area is any area in the image to be detected; the processing unit , configured to determine that the first image region includes the overall image of the target customer when the first numerical value is greater than a first threshold; the processing unit is configured to determine that the first image area includes a second numerical value greater than a second threshold Regions include partial images of target customers.
  • the processing unit is configured to input the overall image into the group attribute prediction model to determine the probability of different group attributes of the target customer; the processing unit is configured to determine the probability of the target customer according to the probability of different group attributes of the target customer Customer group attributes.
  • different group attributes include at least one of mutually exclusive group attributes and independent group attributes; the processing unit is configured to determine the target customer under the condition that different group attributes include mutually exclusive group attributes
  • the group attributes include the first mutually exclusive group attribute; the first mutually exclusive group attribute is the group attribute with the highest probability among the mutually exclusive group attributes; the processing unit is configured to determine when different group attributes include independent group attributes
  • the group attributes of the target customers include a first independent group attribute; the first independent group attribute is a group attribute with a probability greater than an attribute probability threshold among the independent group attributes.
  • the processing unit is configured to input the partial image into the individual attribute prediction model, and determine the gender probability corresponding to each gender of the target customer and the age probability corresponding to each age; the processing unit is configured to determine the target customer The gender of the customer is the gender with the highest gender probability; the processing unit is configured to determine that the age of the target customer is the age with the highest age probability.
  • the processing unit is configured to detect whether the target customer exists in the user database according to the partial image of the target customer; the communication unit is configured to obtain the customer level of the target customer if the target customer exists in the user database .
  • the communication unit is configured to obtain the basic information of the target customer if the target customer exists in the user database; the communication unit is configured to send the basic information of the target customer to the service terminal; A terminal used by personnel; a communication unit configured to obtain the historical business of the target customer under the condition that the target customer authorizes the inquiry of historical business; the communication unit is configured to send the historical business of the target customer to the service terminal.
  • the processing unit is configured to determine the attribute factor of the target customer according to the group attribute and the individual attribute; the attribute factor is used to indicate the influence degree of the attribute of the target customer on the service priority; the processing unit is configured to acquire the target customer The customer's queue information, and determine the queue factor according to the queue information; the queue factor is used to indicate the influence degree of the target customer's queue information on the service priority; the processing unit is configured to obtain the reservation business of the target customer, and determine the business according to the reservation business The factor; the service factor is used to indicate the degree of influence of the reservation service of the target customer on the service priority; the processing unit is configured to determine the service priority of the target customer according to the attribute factor, the queue factor and the business factor.
  • the processing unit is configured to determine multiple overall images of the target customer at different moments in the images captured by multiple image acquisition devices; the processing unit is configured to, according to the position information of the multiple overall images, Determine the route trajectory of target customers.
  • the processing unit is configured to determine the residence time of the target customer in at least one specific area according to the route trajectory of the target customer; at least one specific area corresponds to at least one business; the processing unit is configured to determine the target customer
  • the concerned business is the business corresponding to the specific area with the longest residence time;
  • the communication unit is configured to send the business information of the concerned business to the service terminal;
  • the service terminal is a terminal used by the staff.
  • a non-transitory computer readable storage medium stores computer program instructions.
  • the computer program instructions run on a computer (eg, a customer service device), the computer executes the customer service method as described in any one of the above embodiments.
  • a computer program product includes computer program instructions.
  • the computer program instructions When the computer program instructions are executed on a computer (eg, a customer service device), the computer program instructions cause the computer to execute the customer service method as described in any of the above embodiments.
  • a computer program is provided.
  • the computer program When the computer program is executed on a computer (for example, a customer service device), the computer program causes the computer to execute the customer service method as described in any of the above embodiments.
  • a chip in yet another aspect, includes a processor and a communication interface, the communication interface is coupled to the processor, and the processor is used to run computer programs or instructions to implement the customer service method as described in any of the above embodiments.
  • the chip provided in the present disclosure further includes a memory for storing computer programs or instructions.
  • all or part of the above computer instructions may be stored on a computer-readable storage medium.
  • the computer-readable storage medium may be packaged together with the processor of the device, or may be packaged separately with the processor of the device, which is not limited in the present disclosure.
  • a customer service system including: a customer service device and at least one image acquisition device, wherein the customer service device is configured to execute the customer service method as described in any one of the above embodiments.
  • Fig. 1 is a structural diagram of a customer service system provided according to some embodiments
  • Fig. 2 is an example diagram of image data acquired by an image acquisition device provided according to some embodiments.
  • Fig. 3 is a structural diagram of a customer service device provided according to some embodiments.
  • FIG. 4 is a flow chart providing another method of customer service according to some embodiments.
  • Fig. 5 is a flowchart of another customer service method provided according to some embodiments.
  • Fig. 6 is a flowchart of another customer service method provided according to some embodiments.
  • Fig. 7 is a flowchart of another customer service method provided according to some embodiments.
  • Fig. 8 is a flowchart of another customer service method provided according to some embodiments.
  • Fig. 9 is an example diagram of an image to be detected according to some embodiments.
  • Fig. 10 is a flowchart of another customer service method provided according to some embodiments.
  • Fig. 11 is a flowchart of another customer service method provided according to some embodiments.
  • Fig. 12 is a flowchart of another customer service method provided according to some embodiments.
  • Fig. 13 is a flowchart of another customer service method provided according to some embodiments.
  • Fig. 14 is a function diagram of age evaluation value and age according to some embodiments.
  • Fig. 15 is a flowchart of another customer service method provided according to some embodiments.
  • Fig. 16 is a mapping diagram of an overall image in a target area according to some embodiments.
  • Fig. 17 is a structural diagram of another customer service device provided according to some embodiments.
  • Fig. 18 is a structural diagram of another customer service device provided according to some embodiments.
  • first and second are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of indicated technical features. Thus, a feature defined as “first” and “second” may explicitly or implicitly include one or more of these features. In the description of the embodiments of the present disclosure, unless otherwise specified, "plurality” means two or more.
  • the expressions “coupled” and “connected” and their derivatives may be used.
  • the term “connected” may be used in describing some embodiments to indicate that two or more elements are in direct physical or electrical contact with each other.
  • the term “coupled” may be used when describing some embodiments to indicate that two or more elements are in direct physical or electrical contact.
  • the terms “coupled” or “communicatively coupled” may also mean that two or more elements are not in direct contact with each other, but yet still co-operate or interact with each other.
  • the embodiments disclosed herein are not necessarily limited by the context herein.
  • At least one of A, B and C has the same meaning as “at least one of A, B or C” and both include the following combinations of A, B and C: A only, B only, C only, A and B A combination of A and C, a combination of B and C, and a combination of A, B and C.
  • a and/or B includes the following three combinations: A only, B only, and a combination of A and B.
  • the term “if” is optionally interpreted to mean “when” or “at” or “in response to determining” or “in response to detecting,” depending on the context.
  • the phrases “if it is determined that " or “if [the stated condition or event] is detected” are optionally construed to mean “when determining ! or “in response to determining ! depending on the context Or “upon detection of [stated condition or event]” or “in response to detection of [stated condition or event]”.
  • Pedestrian re-identification refers to the technology of detecting specific pedestrians in image or video stream data through computer vision technology. ReID is usually used to determine pedestrian images from image information acquired by at least one image acquisition device (ie across devices), so as to detect or track the target.
  • Neural networks also known as artificial neural networks (ANNs) are a mathematical model algorithm that imitates the behavior characteristics of animal neural networks and performs distributed parallel information processing.
  • Neural networks include deep learning networks, such as convolutional neural networks (CNN), residual networks (resNet), long short-term memory networks (long short-term memory, LSTM), etc.
  • Intersection over union refers to the ratio of the intersection area of two image areas to the union area of two image areas, and is usually used to detect the relationship between two image areas, such as calculating the predicted image area and the actual The IoU of an image region determines the accuracy of the prediction.
  • L1 norm also known as the Manhattan distance
  • L1 norm regularization L1 regularization
  • L1 regularization is usually applied to the loss function in machine learning algorithms to predict or classify target customers.
  • the present disclosure provides a customer service method, which determines the group attribute and individual attribute of the target customer based on the overall image and partial image of the target customer, and assigns staff to the target customer from the two dimensions of the group attribute and the individual attribute. Therefore, the scope of application of the present disclosure is wider, and at the same time, it can improve the accuracy and intelligent level of service business for target customers, and guarantee the user experience of target customers.
  • FIG. 1 is a structural diagram of a customer service system 10 provided according to some embodiments. As shown in FIG. 1
  • customer service device 101 there may be one customer service device 101 , image acquisition device 102 , service terminal 103 and user terminal 104 in the present disclosure, or there may be multiple ones.
  • image acquisition device 102 image acquisition device 102
  • service terminal 103 user terminal 104
  • FIG. 1 For ease of understanding, only one customer service device 101 , image acquisition device 102 , service terminal 103 and user terminal 104 are shown in FIG. 1 .
  • the customer service device 101 is connected to the image acquisition device 102 through a communication link, the customer service device 101 is connected to the service terminal 103 through a communication link, and the customer service device 101 is connected to the user terminal 104 through a communication link.
  • the communication link may be a wired communication link or a wireless communication link, which is not limited in the present disclosure.
  • the customer service device 101 is configured to determine the overall image and partial image of the target customer, and determine the relevant attributes of the target customer according to the overall image and partial image of the target customer, so as to assign staff to the target customer according to the relevant attributes of the target customer.
  • the customer service device 101 respectively determines the overall image and partial image of the multiple target customers, so as to assign staff to the multiple target customers.
  • the overall image is an image representing the overall screen information of the target customer
  • the partial image is an image representing part of the screen information of the target customer.
  • the customer service device 101 receives the image data sent by the image acquisition device 102, and determines the overall image and partial image of the target customer according to the image data.
  • FIG. 2 is an example diagram of image data 20 acquired by an image acquisition device provided according to some embodiments.
  • the customer service device 101 receives the image data 20 sent by the image acquisition device 102.
  • the customer service device 101 determines that the overall image of the target customer 1 is the image 201 , and the partial images are the image 203 , the image 204 and the image 205 .
  • the customer service device 101 determines that the overall image of the target customer 2 is the image 202 , and the partial images are the image 206 , the image 207 and the image 208 .
  • the staff may be service personnel providing services to target customers, or service equipment providing services to target customers.
  • the device is a robot, a terminal device, an outlet facility, and the like.
  • the customer service device 101 is also configured to send the customer information of the target customer to the service terminal 103 .
  • the service terminal 103 receives the customer information sent by the customer service device 101 .
  • the service terminal 103 may be a terminal used by staff.
  • the customer information includes at least one item of relevant attributes, basic information, historical business, or concerned business of the target customer.
  • the customer information may include a partial image of the target customer, information such as gender, age, and partial images of accompanying personnel.
  • the service terminal 103 may be coupled to the device.
  • the customer service device 101 can be a server, or the customer service device 101 can be an application program installed on the server to provide customer service functions, or the customer service device 101 can be a central processing unit in the server, or , the customer service apparatus 101 may be a control module configured to execute a customer service method in the server.
  • FIG. 3 is a structural diagram of a customer service apparatus 101 provided according to some embodiments.
  • the customer service device 101 includes: a data collection module 301 , a detection and tracking module 302 , an attribute analysis module 303 , a data analysis module 304 and a data storage and sending module 305 .
  • the data acquisition module 301 is configured to acquire image data sent by the image acquisition device 102 .
  • the detection and tracking module 302 is configured to determine the overall image and the partial image of the target customer from the image data.
  • the detection and tracking module 302 is also configured to track the target customer and determine the route track of the target customer.
  • the attribute analysis module 303 is configured to determine group attributes of target customers according to the overall image.
  • the attribute analysis module is also configured to determine individual attributes of the target customer based on the partial image.
  • the group attribute is used to represent the type of group to which the target customer belongs, and the individual attribute includes at least one of gender, age and customer level.
  • the data analysis module 304 is configured to assign staff to the target customer according to the group attribute of the target customer and the individual attribute of the target customer.
  • the data storage and sending module 305 is configured to store the data to be stored during execution of the above modules.
  • the data storage and sending module 305 is also configured to send the customer information of the target customer to the service terminal 103 .
  • the image acquisition device 102 is configured to acquire image data of the target area, and send the image data to the customer service device 101 .
  • the customer service device 101 receives the image data sent by the image acquisition device 102 .
  • the target area may be a business place that provides services, for example, the target area may be a bank, a communication business hall, a clothing store, a hotel, and the like.
  • the target area may also be an exhibition place, for example, the target area is a museum, a science and technology museum, a calligraphy and painting exhibition, a car exhibition, and the like.
  • the target area may also be an event place, for example, the target area is a ball game exhibition hall, a car game exhibition hall, and the like.
  • the image acquisition device 102 may acquire image data of the target area in real time, and send the image data to the customer service device 101 .
  • the image acquisition device 102 can also acquire image data of the target area according to a preset frequency, and send the image data to the customer service device 101 .
  • the image acquisition device 102 in the embodiment of the present disclosure is a device that converts image data into an analog signal or a digital signal through a photoreceptor, and can be deployed on land, including indoor or outdoor, handheld or vehicle-mounted. It can also be deployed on water (such as ships, etc.). It can also be deployed in the air (for example, on aircraft, balloons and satellites, etc.).
  • the image acquisition device 102 includes a camera, a video camera, and a camera.
  • the image acquisition device 102 can also be a device with a camera function.
  • the image acquisition device 102 can be a mobile phone, a tablet computer, a notebook computer, a palmtop computer, a wearable device (such as a smart watch, a smart bracelet, a pedometer) with a camera function. devices, etc.), vehicle-mounted equipment and flight equipment (for example, intelligent robots, hot air balloons, drones, airplanes), etc.
  • the image acquisition device 102 in the embodiment of the present disclosure may also be an infrared imager or a night vision device, for acquiring image data of a dark area.
  • the service terminal 103 is configured to receive customer information sent by the customer service device 101 .
  • the kiosk 103 is also configured to output customer information.
  • the service terminal 103 may output customer information in the form of screen, text or sound, so that staff can obtain the customer information.
  • the user terminal 103 is configured to send reservation service information to the customer service device 101 , and the customer service device 101 receives the service reservation information sent by the user terminal 103 accordingly.
  • the user terminal 103 is also configured to receive the request message sent by the customer service device 101 .
  • the user terminal 103 may be a terminal device used by a target customer or an accompanying person.
  • Terminal equipment is also called user equipment (user equipment, UE), mobile station (mobile station, MS), mobile terminal (mobile terminal, MT), etc., and is a device that provides voice and/or data connectivity to users.
  • a terminal includes a handheld device with a wireless connection function, a vehicle-mounted device, and the like.
  • terminal devices can be: mobile phone, tablet computer, notebook computer, palmtop computer, mobile internet device (mobile internet device, MID), wearable device (such as smart watch, smart bracelet, pedometer, etc.) , vehicle-mounted equipment (such as automobiles, bicycles, electric vehicles, airplanes, ships, trains, high-speed rail, etc.), virtual reality (virtual reality, VR) equipment, augmented reality (augmented reality, AR) equipment, industrial control (industrial control) wireless terminals in smart home devices (for example, refrigerators, TVs, air conditioners, electric meters, etc.), intelligent robots, workshop equipment, wireless terminals in self driving, wireless terminals in remote medical surgery, Wireless terminals in smart grid, wireless terminals in transportation safety, wireless terminals in smart city, or wireless terminals in smart home, flight equipment (for example, intelligent robots, hot air balloons, drones, airplanes), etc.
  • vehicle-mounted equipment such as automobiles, bicycles, electric vehicles, airplanes, ships, trains, high-speed rail, etc.
  • virtual reality virtual reality
  • AR augmented reality
  • a terminal device is a terminal device that often works on the ground, such as a vehicle-mounted device.
  • chips deployed in the above devices such as System-On-a-Chip (SOC), baseband chips, etc., or other chips with communication functions may also be referred to as terminal devices.
  • Fig. 4 is a flowchart of a customer service method provided according to some embodiments. As shown in Figure 4, the method includes the following steps:
  • Step 401 the customer service device determines the overall image and partial image of the target customer.
  • the overall image is an image representing the overall screen information of the target customer
  • the partial image is an image representing part of the screen information of the target customer.
  • the overall image can be an image that includes the target customer (as shown in image 201 in Figure 2), and the partial image can be a face image of the target customer (such as image 203 in Figure 2), or it can be the target customer's hand
  • the body image (such as the image 204 in FIG. 2 ) can also be the foot image of the target customer (such as the image 205 in FIG. 2 ).
  • the present disclosure takes the partial image as a facial image as an example to describe the customer service method provided in the present disclosure in detail.
  • the customer service device acquires the image of the target area from the image acquisition device, and inputs the image of the target area into the target detection model to obtain the overall image and partial image of the target customer.
  • the target recognition model can be a neural network model, such as YOLO (you only look once) algorithm model, SSD (single shot multibox detector) algorithm model, R-CNN (region CNN) algorithm model, Faster R-CNN algorithm model .
  • YOLO young only look once
  • SSD single shot multibox detector
  • R-CNN region CNN
  • Faster R-CNN algorithm model a neural network model
  • the target customer may be one or multiple.
  • the customer service device respectively determines the overall image and the partial image of the multiple target customers.
  • the customer service device can determine that the overall image and partial image satisfying the correlation correspond to the target customer.
  • the customer service device determines to bind the target customer with the overall image and the partial image through the association relationship between the overall image and the partial image.
  • the customer service device determines that the overall image includes image 201 and image 202 , and the partial image includes image 203 and image 206 .
  • the customer service device determines the IoU of image 201 and image 203 , the IoU of image 201 and image 206 , the IoU of image 202 and image 203 , and the IoU of image 202 and image 206 .
  • the customer service device determines that target customer 1 corresponds to image 201 and image 203
  • target customer 2 corresponds to image 202 and image 206 .
  • Step 402 the customer service device determines the group attribute of the target customer according to the overall image.
  • the group attribute is used to characterize the type of group to which the target customer belongs.
  • the customer service device may classify groups according to occupation types, for example, group attributes include staff, special occupations and other personnel. Group attributes may also include other specific groups, such as pregnant women, people with disabilities, and other groups with limited mobility.
  • the customer service device may input the overall image into the group attribute prediction model to obtain the group attribute of the target customer.
  • the group attribute prediction model may be a neural network model, such as an EfficientNet algorithm model, a residual network (ResNet) algorithm model.
  • a neural network model such as an EfficientNet algorithm model, a residual network (ResNet) algorithm model.
  • customer service can also obtain object images associated with group attributes, and further determine the group attributes of target customers based on the object images and the overall image.
  • the customer service can detect object images such as wheelchairs and crutches, and calculate the IoU between the object image and the overall image, and when the IoU is greater than a preset threshold, it is confirmed that the group attribute of the target customer includes persons with disabilities.
  • the customer service device can also filter the target customer whose group attribute is the staff member, so as to avoid the false detection problem when detecting the target customer and improve the efficiency of the target customer. detection accuracy.
  • the overall image determined by the customer service device may be occluded or truncated.
  • the occluded and truncated overall image will affect the group attribute of the target customer determined by the customer service device. Therefore, the customer service device can obtain multiple overall images of the target customer, and select an overall image with better image quality, thereby improving the accuracy of the customer service device in determining group attributes.
  • the customer service device may also determine the occlusion factor and the truncation factor of the overall image according to the overall image, and determine the quality evaluation value of the overall image according to the occlusion factor and the truncation factor.
  • the occlusion factor is used to characterize the degree of occlusion of the overall image
  • the truncation factor is used to characterize the degree of truncation of the overall image.
  • the quality evaluation value of the overall image satisfies the following formula 1:
  • Q is the quality evaluation value of the overall image
  • P 1 is the occlusion factor
  • P 2 is the truncation factor
  • a is the weight value of the occlusion factor
  • b is the weight value of the truncation factor.
  • a and b can be set according to the actual situation, which is not limited in the present disclosure.
  • the customer service device removes the overall image.
  • the quality assessment value of the overall image is less than or equal to the preset quality threshold, it indicates that the image quality of the overall image is better, and the customer service device retains the overall image.
  • the above step 402 can be realized by the following steps: the customer service device determines the group attribute of the target customer according to the overall image satisfying the first preset condition.
  • the first preset condition is that the quality evaluation value of the overall image is less than or equal to a preset quality threshold.
  • Step 403 the customer service device determines the individual attributes of the target customer according to the partial image.
  • the individual attribute is used to characterize the information of the target customer individual.
  • individual attributes include at least one of gender, age, and customer level.
  • the partial image can reflect the characteristics of the details of the target customer, so the customer service device can determine the attributes representing individual information of the target customer through the partial image.
  • the customer service device can perform operations such as face key point detection and face alignment on the face image to obtain a standard face image of the target customer for subsequent determination of individual attributes.
  • operations such as face key point detection and face alignment on the face image to obtain a standard face image of the target customer for subsequent determination of individual attributes.
  • the customer service device may input the partial image into the individual attribute prediction model to obtain the individual attribute of the target customer.
  • the individual attribute prediction model may be a neural network model, such as a ShuffleNet algorithm model, a residual network (ResNet) algorithm model.
  • a neural network model such as a ShuffleNet algorithm model, a residual network (ResNet) algorithm model.
  • the customer service device may also determine the image quality of the partial images according to the partial images, and remove partial images with poor image quality.
  • the customer service device may also determine the image quality of the partial images according to the partial images, and remove partial images with poor image quality.
  • the customer service device may also determine the image quality of the partial images according to the partial images, and remove partial images with poor image quality.
  • the customer service apparatus may also perform corrections based on the individual attributes and group attributes of the target customers. For example, when the customer service device determines that the gender of the target customer is male, the customer service device may remove the attribute of pregnant women from the group attributes of the target customer. In this way, the customer service device can further improve the accuracy of determining the attribute of the target customer.
  • Step 404 the customer service device determines the service priority of the target customer and the type of staff corresponding to the target customer according to the group attribute of the target customer and the individual attribute of the target customer.
  • the service priority is used to indicate the priority order in which the customer service device assigns staff to the target customer.
  • the customer service device can increase the service priority of the target customers of the disabled groups such as pregnant women and the disabled, so as to ensure the service experience of each user group.
  • the customer service device may also determine the service priority of the target customer according to the target customer's group attributes, individual attributes, queue information, and reservation business. For details, reference may be made to subsequent descriptions, and details are not repeated here.
  • the staff can be service personnel providing services to target customers, or service equipment providing services to target customers.
  • the device is a robot, a terminal device, an outlet facility, and the like.
  • the customer service device can also classify the staff according to the group attributes of the target customers that the staff can serve and the individual attributes of the target customers, so as to determine multiple types of staff, so as to effectively provide corresponding services for different target customers.
  • Step 405 the customer service device allocates staff to the target customer according to the service priority of the target customer and the staff type corresponding to the target customer.
  • the customer service device assigns staff of the corresponding staff type to the target customer according to the priority order corresponding to the service priority.
  • the customer service device determines that the service priority of the target customer is high priority, and the staff type is the first staff type.
  • the customer service device preferentially allocates staff corresponding to the first staff type to the target customer.
  • the customer service device determines the correspondence between the customer level of the target customer that the staff can serve and the staff type.
  • the customer service device assigns staff of corresponding staff type to the target customer according to the customer level of the target customer.
  • the customer service device classifies the staff according to their business capabilities for handling different businesses, and assigns staff of the corresponding staff type to the target customer according to the reservation business type of the target customer.
  • the customer service device assigns staff capable of handling complex businesses to target customers, and for simple businesses, the customer service device allocates self-service equipment to target customers.
  • the customer service device may also determine the corresponding staff type according to the group to which the target customer belongs, and assign staff to the target customer.
  • the customer service device classifies the staff according to the relevant service experience for special occupations that the staff can serve, or groups with disabilities such as pregnant women and the disabled, and assigns corresponding types of staff to better provide Serve.
  • the customer service device can also classify workers according to their productivity, assigning workers with high productivity to younger groups.
  • the customer service device in the present disclosure determines the overall image and partial image of the target customer, determines the group attribute of the target customer according to the overall image, and determines the individual attribute of the target customer according to the partial image. Since the group attribute can represent the type of the group to which the target customer belongs, and the individual attribute can represent the individual information of the target customer, the customer service device can assign staff to the target customer from the two dimensions of the group attribute and the individual attribute, thereby improving the efficiency of the target customer.
  • the accuracy and intelligence level of the service business is provided to ensure the user experience of the target customers.
  • the customer service device also determines the service priority of the target customer and the staff type corresponding to the target customer according to the group attribute and individual attribute of the target customer, and assigns the corresponding staff type to the target customer according to the priority order of the service priority. staff member. In this way, the customer service device can provide the best service for different target customers in a targeted manner, ensuring the service experience of different target customers.
  • step 405 the above method further includes the following steps 501-503:
  • Step 501 the customer service device sends the customer information of the target customer and the current location information of the target customer to the service terminal.
  • the service terminal is a terminal used by staff.
  • the customer service device may send the customer information of the target customer and the current location information of the target customer to the service terminal, so that the staff can provide services to the target customer in time.
  • customer information may include information such as a partial image of a target customer, an overall image, group attributes, and individual attributes.
  • the location information of the target customer can be determined through the overall image of the target customer, and the specific implementation method can refer to the subsequent description, which will not be repeated here.
  • Step 502 the customer service device determines the first location distance between the staff and the target customer according to the overall image of the staff and the overall image of the target customer.
  • the customer service device may store the overall image and the partial image of the staff in advance, and may also obtain the overall image of the staff through the above steps 401-403.
  • the customer service device After the customer service device acquires the overall image and the partial image of the first object, it determines the group attribute of the first object according to the overall image of the first object.
  • the group attribute of the first object includes staff
  • the customer service device determines that the first object is a staff.
  • the overall image of the first object is the overall image of the staff member.
  • the customer service device can map the overall image to the coordinate system corresponding to the target area, obtain the position coordinates of the staff and the target customer, and determine the staff and target customer according to the position coordinates.
  • the target customer's first location distance can be determined.
  • Step 503 when the distance from the first location is greater than the first preset distance and the duration is greater than the first duration, the customer service device sends service prompt information to the service terminal.
  • the service reminder information is used to remind the staff to provide services for the target customers in time.
  • Specific values of the first preset distance and the first duration can be set according to actual conditions, which are not limited in the present disclosure.
  • the customer service device can determine whether the staff has reached the target customer based on the distance between the staff and the target customer. When the distance between the staff and the target customer is greater than the first preset distance, it means that the staff does not provide services to the target customer. When the duration is longer than the first duration, the customer service device may send a service prompt message to the service terminal to prompt staff to provide services to the target customer in time.
  • the customer service device may regularly determine whether the distance to the first location is greater than the first preset distance at preset intervals.
  • the customer service device may send service prompt information to the service terminal again, or reassign staff to the target customer.
  • the customer service device in the present disclosure can send the customer information of the target customer and the current location information of the target customer to the service terminal after assigning staff to the target customer, so that the staff can reach the target customer in time and provide service for the target customer. target customers to provide services.
  • the customer service device will also judge whether the staff is providing service to the target customer based on the distance between the staff and the target customer, and remind the staff to provide services when the duration is longer than the first time, thereby ensuring the service experience of the target customer.
  • step 401 the above method further includes the following step 601:
  • Step 601 the customer service device receives the registration information of the target customer.
  • the registration information includes a partial image of the target customer and a partial image of an accompanying person of the target customer.
  • the customer service device may store the acquired partial images of the target customer and the partial images of the target customer's accompanying personnel in the user database, and set the binding relationship between the target customer and the accompanying personnel, so that subsequent The customer service device obtains the relevant information of the target customer and the accompanying personnel.
  • the above method further includes the following steps 602-604:
  • Step 602 the customer service device determines the overall image of the accompanying person according to the partial image of the accompanying person.
  • the customer service device can match the partial image of the target customer with the partial image of each user in the user database.
  • the customer service device can further detect whether there is an accompanying person bound to the target customer, so as to obtain a partial image of the accompanying person.
  • the process of matching the partial image of the target customer with the partial image of each user in the user database by the customer service device can refer to the subsequent description, and will not be repeated here.
  • the customer service device can determine the overall image that has an association relationship with the partial image (for example, IoU greater than a preset threshold) in the images collected by the image acquisition device according to the partial image of the accompanying person, and use the overall image as the overall image of the accompanying person.
  • step 401 For a specific implementation manner, reference may be made to related descriptions in step 401, and details are not repeated here.
  • Step 603 the customer service device determines the second position distance between the target customer and the accompanying person according to the overall image of the target customer and the overall image of the accompanying person.
  • the specific method for the customer service device to determine the second location distance between the target customer and the accompanying person is the same as the implementation method for the customer service device to determine the first location distance between the staff member and the target customer in step 502 above, and will not be repeated here.
  • Step 604 when the second location distance is greater than the second preset distance, the customer service device assigns staff to the target customer according to the service priority of the target customer and the staff type corresponding to the target customer.
  • the second preset distance may be set according to actual conditions, which is not limited in the present disclosure.
  • the customer service device can assign staff to the target customer in time, so as to provide services for the target customer.
  • the customer service device in the present disclosure acquires the partial image of the accompanying person of the target customer to determine the second position distance between the target customer and the accompanying person, and when the distance between the target customer and the accompanying person is too large, timely provide the target customer Allocate staff to ensure the service experience of target customers.
  • step 405 the process of assigning staff to the target customer by the customer service device will be specifically introduced.
  • step 405 may also be implemented through the following steps 701-703:
  • Step 701 the customer service device determines the target staff group.
  • the type of the target staff group is the staff type corresponding to the target customer, and the target staff group includes multiple target staff.
  • the customer service device may confirm the target staff group corresponding to the type based on the staff type.
  • Step 702 the customer service device determines the personnel priority of the target staff group according to the personnel attributes of the target staff group.
  • the personnel attribute is used to represent the workload of each target worker in the target worker group.
  • the personnel attribute includes at least one of working hours, working intensity and working frequency.
  • Staff priority is used to indicate the order in which each target worker is assigned within the target worker group.
  • the customer service device can determine the personnel priority of the target staff group.
  • the customer service device determines that the target worker's personnel priority is lower. Conversely, when the target worker's working hours are shorter, the work intensity is lower, and the work frequency is lower, the customer service device determines that the target worker has a higher personnel priority.
  • the customer service device may set the target worker's personnel priority as a high priority, thereby improving the working efficiency of the service device.
  • Step 703 the customer service device assigns target staff to the target customer according to the service priority of the target customer and the personnel priority of the target staff group.
  • the customer service device assigns staff to the target customer preferentially.
  • the customer service device preferentially assigns the target worker to the corresponding target customer.
  • the customer service device in the present disclosure can determine the target staff group to be assigned, and assign target staff to the target customer based on the customer priority of the target customer and the personnel priority of the target staff group. In this way, the customer service device can adjust the work intensity of each staff member, avoiding the problem that some staff members are overworked due to uneven distribution of staff members.
  • step 401 the process of determining the overall image and partial image of the target customer by the customer service device will be specifically introduced.
  • step 401 may also be implemented through the following steps 801 to 804:
  • Step 801 the customer service device acquires the image to be detected.
  • the image to be detected includes the target customer.
  • the image to be detected is an image of the target area collected by the image acquisition device. There can be one or more image acquisition devices.
  • the target area may be a business place that provides services, for example, the target area is a bank, a communication business hall, a clothing store, a hotel, and the like.
  • the target area may also be an exhibition place, for example, the target area is a museum, a science and technology museum, a calligraphy and painting exhibition, a car exhibition, and the like.
  • the target area may also be an event place, for example, the target area is a ball game exhibition hall, a car game exhibition hall, and the like.
  • the customer service device receives images to be detected sent by one or more image acquisition devices.
  • Step 802 the customer service device determines a first image area in the image to be detected, and a first value and a second value corresponding to the first image area.
  • the first numerical value represents the probability that the first image area includes the whole image.
  • the second value represents the probability that the first image region includes a partial image.
  • the first image area is any area in the image to be detected.
  • FIG. 9 is an example diagram of an image to be detected 90 provided according to some embodiments.
  • the customer service device determines the first image area (that is, any dashed box in FIG. 9 ) from the image 90 to be detected, and the first value and the second value corresponding to the first image area.
  • Step 803 in the case that the first value is greater than the first threshold, the customer service device determines that the first image area includes the overall image of the target customer.
  • the first numerical value represents the probability that the first image area includes the overall image, therefore, the larger the first numerical value, the greater the probability that the first image area includes the overall image of the target customer.
  • the customer service device determines that the image included in the first image area (such as image 901 in FIG. 9 ) is the overall image of the target customer.
  • Step 804 in the case that the second value is greater than the second threshold, the customer service device determines that the first image area includes a partial image of the target customer.
  • the customer service device determines that the image included in the first image area (such as the image 902 in FIG. 9 ) is a partial image of the target customer.
  • the customer service device in the present disclosure can acquire the image to be detected and determine the first image area in the image to be detected and the corresponding first value and second value. Since the first numerical value represents the probability that the first image area includes the overall image, and the second numerical value represents the probability that the first image area includes the partial image, the customer service device can then use the determined probability that the first image area includes the entire image and the probability that the first image area includes the entire image.
  • the probability that an image area includes a partial image determines the overall image and partial image of the target customer in the image to be detected, which improves the accuracy of image detection and facilitates the subsequent assignment of staff to the target customer based on the overall image and the partial image.
  • step 402 can also be implemented through the following steps 1001-1002:
  • Step 1001 the customer service device inputs the overall image into the group attribute prediction model, and determines the probability of different group attributes of the target customer.
  • the group attribute prediction model is used to determine the probability of different group attributes of the target customers.
  • the group attribute prediction model may be a neural network model, such as an EfficientNet algorithm model, a residual network (ResNet) algorithm model.
  • a neural network model such as an EfficientNet algorithm model, a residual network (ResNet) algorithm model.
  • the different group attributes include at least one of mutually exclusive group attributes and independent group attributes.
  • a mutually exclusive group attribute refers to an attribute corresponding to multiple groups that do not have an intersection.
  • a target customer cannot have multiple group attributes in mutually exclusive group attributes at the same time.
  • mutually exclusive group attributes include workers, special occupations, and others.
  • Target customers can be any one of staff, special occupations and others.
  • Independent group attributes refer to attributes that are independent of each other and are not affected by other group attributes.
  • separate group attributes include pregnant women, people with disabilities.
  • Target customers can be pregnant women, people with disabilities, pregnant women and people with disabilities.
  • the customer service device inputs the overall image into the group attribute prediction model to obtain probability factors of different group attributes.
  • the probability factor is used to characterize the credibility of the target customer's attributes of the group.
  • the customer service device determines the probability of different group attributes of the target customer according to the probability factors of different group attributes.
  • the customer service device inputs the probability factors of mutually exclusive group attributes into the Softmax function to obtain the probability of mutually exclusive group attributes.
  • p i is the probability of the i-th attribute in the mutually exclusive group attributes
  • v i is the probability factor of the i-th attribute in the mutually exclusive group attributes
  • v j is the probability factor of the j-th attribute in the mutually exclusive group attributes.
  • the customer service device converts the probability factor of the mutually exclusive group attribute into a non-negative number through an exponential function with the natural constant e as the base, and uses the converted probability factor of the ith mutually exclusive group attribute as a numerator, and converts the mutually exclusive
  • the sum of the converted probability factors of each attribute in the exclusive group attribute is used as the denominator, so as to obtain the probability of the ith mutually exclusive group attribute.
  • the probability factor of mutually exclusive group attribute 1 is 0.5
  • the probability factor of mutually exclusive group attribute 2 is 2
  • the probability factor of mutually exclusive group attribute 3 is 3.
  • the transformed probability factor of mutually exclusive group attribute 1 is 1.648721
  • the transformed probability factor of mutually exclusive group attribute 2 is 7.389056
  • the transformed probability factor of mutually exclusive group attribute 3 is 20.085537. Therefore, the probability of mutually exclusive group attribute 1 is 5.6612%
  • the probability of mutually exclusive group attribute 2 is 25.3716%
  • the probability of mutually exclusive group attribute 3 is 68.9672%.
  • the customer service device inputs the probability factors of independent group attributes into the Sigmoid function to obtain the probability of independent group attributes.
  • the sigmoid function satisfies the following formula 3:
  • x is the probability factor of any attribute in the independent group attributes
  • S(x) is the probability of this attribute.
  • the probability factor of independent group attribute 1 is 3, the probability factor of independent group attribute 2 is 2, the probability factor of independent group attribute 3 is 4, the obtained probability of independent group attribute 1 is 95.2574%, and the probability of independent group attribute 2 is 88.0797%, and the probability of independent group attribute 3 is 98.2014%.
  • Step 1002 the customer service device determines the group attributes of the target customers according to the probabilities of different group attributes of the target customers.
  • the customer service apparatus determines that the group attributes of the target customer include the first mutually exclusive group attributes.
  • the first mutually exclusive group attribute is the group attribute with the highest probability among the mutually exclusive group attributes.
  • probabilities of mutually exclusive group attributes affect each other. The greater the probability value of one of the mutually exclusive group attributes, the smaller the probability values of other group attributes in the mutually exclusive group attributes. The sum of the probabilities of mutually exclusive group attributes is 100%.
  • the probability of mutually exclusive group attribute 1 is 5.6612%
  • the probability of mutually exclusive group attribute 2 is 25.3716%
  • the probability of mutually exclusive group attribute 3 is 68.9672%.
  • the group attribute of the target customer determined by the customer service device includes the mutually exclusive group attribute 3 .
  • the customer service apparatus determines that the group attributes of the target customer include the first independent group attribute.
  • the first independent group attribute is a group attribute whose probability is greater than an attribute probability threshold among the independent group attributes.
  • the probability of an independent group attribute is only related to the independent group attribute and has nothing to do with other independent group attributes.
  • the target customer may have one or more group attributes in the independent group attributes at the same time.
  • the attribute probability threshold is set to 90%
  • the probability of independent group attribute 1 is 95.2574%
  • the probability of independent group attribute 2 is 88.0797%
  • the probability of independent group attribute 3 is 98.2014%.
  • the customer service device determines that the group attributes of the target customers include both the independent group attribute 1 and the independent group attribute 3 .
  • the customer service device in the present disclosure can input the overall image into the group attribute prediction model to determine the probability of different group attributes, so as to determine the group attributes of target customers according to the probability of different group attributes, so as to facilitate subsequent target customers Assign staff.
  • different group attributes include at least one of mutually exclusive group attributes and independent group attributes. Among them, the probabilities of mutually exclusive group attributes affect each other, and the probability of independent group attributes is only related to the independent group attributes. Therefore, in the present disclosure, the customer service device determines the probability of corresponding group attributes based on group attributes in different ways, improving The accuracy of the customer service device in determining the group attribute of the target customer is improved, so as to better assign staff to the target customer.
  • step 403 the process of determining the individual attribute of the target customer by the customer service device according to the partial image will be specifically introduced.
  • step 403 can also be implemented through the following steps 1101-1103:
  • Step 1101 the customer service device inputs the partial image into the individual attribute prediction model, and determines the gender probability corresponding to each gender and the age probability corresponding to each age of the target customer.
  • the individual attribute prediction model is used to determine the gender probability corresponding to each gender and the age probability corresponding to each age of the target customer.
  • the individual attribute prediction model may be a neural network model, such as a ShuffleNet algorithm model, a residual network (ResNet) algorithm model.
  • a neural network model such as a ShuffleNet algorithm model, a residual network (ResNet) algorithm model.
  • the customer service device can determine the gender of the target customer through the Softmax function.
  • the specific implementation method can refer to the above related content, and will not be repeated here.
  • the customer service device in this disclosure can use L1norm to regress the age features extracted from the individual attribute prediction model, so as to obtain the age probability corresponding to each age of the target customer.
  • Step 1102 the customer service device determines that the gender of the target customer is the gender with the highest gender probability.
  • the probability that the gender of the target customer is male is 40%
  • the probability that the gender of the target customer is female is 60%
  • the customer service device determines that the gender of the target customer is female.
  • Step 1103 the customer service device determines that the age of the target customer is the age with the greatest age probability.
  • the manner in which the customer service device determines the age of the target customer is similar to the manner in which the gender is determined above, and will not be repeated here.
  • the customer service device in the present disclosure can input partial images into the individual attribute prediction model to determine the individual attributes of the target customer, such as the gender and age of the target customer.
  • the customer service device in the present disclosure determines the probability of the corresponding individual attribute in different ways, and determines the individual attribute of the target customer based on the attribute probability, which improves the accuracy of the customer service device in determining the individual attribute of the target customer. In order to facilitate the subsequent assignment of staff to target customers.
  • step 403 the process of determining the individual attribute of the target customer by the customer service device according to the partial image will be specifically introduced.
  • step 403 further includes the following steps 1201-1202:
  • Step 1201 the customer service device detects whether there is a target customer in the user database according to the partial image of the target customer.
  • the user information of the user is pre-stored in the user database.
  • the user's partial image, customer level, gender, age, historical business and other information For example, the user's partial image, customer level, gender, age, historical business and other information.
  • the customer service device can match the determined partial image of the target customer with the partial image of each user in the user database, thereby detecting whether there is a target customer in the user database.
  • the customer service device inputs the partial image of the target customer and the partial image of each user in the user database into the partial image recognition model to obtain the partial image features of the partial image.
  • the customer service device determines the image similarity between the target customer and each user according to the partial image features of the target customer and the partial image features of each user in the user database.
  • the customer service device determines that the target customer is a user corresponding to the first image similarity.
  • the image features of each user's partial image may be pre-stored in the user database.
  • the customer service device can directly acquire the image features of each user's partial image from the user database, which improves detection efficiency.
  • the customer service device may determine the image similarity between the target customer and each user by calculating the cosine distance between the target customer's local image features and the local image features of each user in the user database.
  • Step 1202 if the target customer exists in the user database, the customer service device acquires the customer level of the target customer.
  • customer levels may be classified according to multiple levels, such as level 1 customers, level 2 customers, and level 3 customers. Customer levels can also be divided according to customer categories, such as VIP customers and non-VIP customers. This disclosure does not limit it.
  • the customer service device in this disclosure can detect the corresponding user from the user database according to the partial image of the target customer, so as to obtain more detailed user information of the target customer so that the subsequent customer service device can better assign the target customer staff member.
  • the above method further includes the following steps 1203-1206:
  • Step 1203 if the target customer exists in the user database, the customer service device obtains the basic information of the target customer.
  • the basic information may be gender, age, contact information, and service-related information.
  • the service-related information is determined according to the scene corresponding to the target area. For example, when the target area is a bank location, the service-related information may be business in the financial field.
  • Step 1204 the customer service device sends the basic information of the target customer to the service terminal.
  • the service terminal receives the basic information of the target customer sent by the customer service device.
  • the service terminal is a terminal used by staff.
  • the customer service device helps the staff to understand the needs of the target customers in a timely manner by sending the basic information of the target customers to the assigned staff, so as to provide better services for the target customers.
  • the service terminal may be coupled to the device.
  • Step 1205 under the condition that the target customer authorizes to inquire about the historical business, the customer service device obtains the historical business of the target customer.
  • the customer service device may send a service query request to the user terminal of the target customer.
  • the service query request is used to confirm whether the target customer is authorized to query historical services.
  • the customer service device When the customer service device receives the authorization message sent by the user terminal, it confirms that the target customer authorizes the inquiry of historical services.
  • user authorization information may be pre-stored in the user database.
  • the customer service device confirms from the authorization information whether the target customer is authorized to inquire about historical services.
  • Step 1206 the customer service device sends the historical business of the target customer to the service terminal.
  • the customer service device in the present disclosure can further obtain the basic information of the target customer from the user database and obtain the historical business of the target customer under the authorization of the target customer, and send it to the service terminal of the staff.
  • This solution not only protects the privacy data of target customers, but also helps to provide better services for target users.
  • step 404 further includes the following steps 1301-1305:
  • Step 1301 the customer service device determines the attribute factor of the target customer according to the group attribute and the individual attribute.
  • the attribute factor is used to represent the impact degree of the target customer's attribute on the service priority.
  • attribute factors of target customers may include one or more.
  • the attribute factor of the target customer includes one, the attribute factor satisfies the following formula 4:
  • S 1 is the attribute factor of the target customer
  • w 1 is the weight value of the attribute factor
  • the attribute factor is the sum of multiple sub-attribute factors.
  • the weight values corresponding to each sub-attribute factor among the multiple sub-attribute factors may be the same value or different values.
  • the method of determining the sub-attribute factors is the same as the above-mentioned method of determining the attribute factors when the attribute factors of the target customer include one, and will not be repeated here.
  • the attribute factors include at least one of a customer level factor, a group attribute factor, and an age factor. That is, the value of the attribute factor is the sum of at least one of the customer level factor, the group attribute factor, and the age factor.
  • attribute factors including any one of customer level factor, group attribute factor and age factor as an example.
  • the customer level factor is determined by the customer level evaluation value and the customer level weight value of the target customer.
  • the customer level evaluation value of the target customer is 1.
  • the target customer's customer level is a non-VIP customer
  • the target customer's customer level evaluation value is 0.
  • the group attribute factor is determined by the group attribute evaluation value and the group attribute weight value of the target customer.
  • the group attribute of the target customer may be one or multiple.
  • the realization method is the same as the method of determining the customer level factor, which will not be repeated here.
  • the first group attribute factor of each group attribute may be determined respectively, and the sum of the first group attribute factors of each group attribute may be used as the group attribute factor of the target customer.
  • the group attribute weight values corresponding to each group attribute may be the same value or different values.
  • the manner in which the customer service apparatus determines the first group attribute factor of each group attribute is the same as the above-mentioned determination method when there is one group attribute of the target customer, and will not be repeated here.
  • the customer service device determines that the group attribute factor of the target customer is the group attribute factor corresponding to the pregnant woman.
  • the customer service device determines that the group attribute factor of the target customer is the sum of the first group attribute factor corresponding to the pregnant woman and the first group attribute factor corresponding to the disabled person.
  • group attribute factors in the present disclosure is described only by taking group attributes including pregnant women and disabled persons as examples above.
  • the group attributes in this disclosure may also include other attributes, which can be set according to actual conditions.
  • the age factor is determined by the target customer's age evaluation value and age weight value.
  • the estimated age value of the target customer can be determined by a segment function.
  • the function graph of age evaluation value and age shown in FIG. 14 can also be determined by the curve function. This disclosure does not limit it.
  • Step 1302 the customer service device obtains the queue information of the target customer, and determines the queue factor according to the queue information.
  • the queue factor is used to represent the degree of influence of the target customer's queue information on the service priority.
  • the target customer can register for a service reservation through a user registration device set in the target area, and the customer service device obtains the queue information of the target customer from the user registration device.
  • the target customer can also remotely reserve a service through the user terminal, and the customer service device receives the service reservation information sent by the user terminal, and determines the queue information of the target customer therefrom.
  • the queue factor is determined by the target customer's queue information evaluation value and queue information weight value.
  • the queue factor satisfies the following formula 5:
  • S 2 is the queue factor of the target customer
  • w 2 is the weight value of the queue information
  • the customer service device may update the queue information evaluation value in real time according to the queue information of the target customer, so as to determine the queue factor of the target customer.
  • the queue information evaluation value satisfies the following formula 6:
  • N is the queuing order in the target customer's queue information.
  • the customer service device may also determine the expected waiting time of the target customer according to the queue information of the target customer, and send the estimated waiting time to the user terminal of the target customer, so that the target customer can know the current situation in time.
  • the customer service device obtains the reservation business of the target customer, and determines the business factor according to the reservation business.
  • the business factor is used to indicate the impact degree of the target customer's reservation business on the service priority.
  • the customer service device may obtain the reservation service of the target customer from the user registration device or the user terminal of the target customer.
  • the customer service device may obtain the reservation service of the target customer from the user registration device or the user terminal of the target customer.
  • the relevant description in step 1302 please refer to the relevant description in step 1302, which will not be repeated here.
  • the business factor is determined by the evaluation value of the reservation business of the target customer and the weight value of the reservation business.
  • the service factor satisfies the following formula 7:
  • S 3 is the business factor of the target customer
  • w 3 is the weight value of the reservation business
  • the reservation business evaluation value is related to the grade of the reservation business. The higher the grade of the reservation service, the greater the evaluation value of the reservation business. On the contrary, the lower the grade of the reservation business, the smaller the evaluation value of the reservation business.
  • Step 1304 the customer service device determines the service priority of the target customer according to the attribute factor, queue factor and business factor.
  • S is the service priority of the target customer
  • S 1 is the attribute factor of the target customer
  • S 2 is the queue factor of the target customer
  • S 3 is the business factor of the target customer.
  • Step 1305 the customer service device determines the type of staff corresponding to the target customer according to the group attribute of the target customer and the individual attribute of the target customer.
  • the customer service device in this disclosure can respectively determine the attribute factor, queue factor and business factor of the target customer, and determine the service priority of the target customer according to the attribute factor, queue factor and business factor, so that according to the service priority Prioritize the allocation of the corresponding staff to the target customers.
  • the customer service device can assign orders to target customers from the attributes of the target customer, the current queuing status, and the type of business to be handled, which makes the process of assigning staff to the customer service device more reasonable and ensures that different target customers are satisfied. service experience.
  • the above method further includes the following steps 1501-1502:
  • Step 1501 the customer service device determines multiple overall images of the target customer at different times from the images collected by multiple image capture devices.
  • the customer service device detects multiple overall images of the target customer at different times from the images collected by multiple image acquisition devices, It is conducive to more accurate and comprehensive determination of the route trajectory of target customers in the follow-up.
  • the customer service device may determine the overall image of the target customer through the method in step 401, which will not be repeated here.
  • the customer service device can determine the overall image of the target customer from the image information collected by any one of the image collection devices. Afterwards, the customer service device can input the overall image into the pedestrian re-identification model to determine the overall image of the target customer in the images collected by multiple image acquisition devices.
  • the pedestrian re-identification model may be a multiple granularity network (MGN) model.
  • the pedestrian re-identification model may also be a model for pedestrian detection, which is not limited in the present disclosure.
  • the customer service device can also correct the position of the overall image according to the detected overall image of the target customer.
  • the customer service device corrects the position information of the overall image of the target customer detected at the next time according to the position information of the overall image of the target customer at the current time, so as to obtain the predicted target customer at the next time overall image.
  • the customer service device may track the overall image of the target customer through the SORT algorithm.
  • the customer service device performs Kalman filtering on the overall image of the detected target customer to estimate the movement state of the target customer, and then performs position matching through the Hungarian matching algorithm to obtain multiple corrected overall images of the target customer at different times image.
  • Step 1502 the customer service device determines the route track of the target customer according to the location information of multiple overall images.
  • the customer service device may map multiple overall images in the coordinate system corresponding to the target area to obtain the coordinates corresponding to each overall image.
  • the customer service device connects each coordinate in chronological order to obtain the route track of the target customer.
  • the customer service device may also determine the image of the target item, and determine whether the target customer has lost the item by detecting the route trajectory of the target item and the target customer.
  • the customer service device detects the relationship between the target item and the target customer's route trajectory.
  • the distance between the target item and the target customer is less than a preset threshold and the duration exceeds a certain period of time, it is determined that the target item belongs to the target customer and the target customer has lost the item.
  • the customer service device in this disclosure can acquire multiple overall images of the target customer at different times from multiple image acquisition devices, so as to determine the route trajectory of the target customer according to the position information corresponding to the multiple overall images, so that In the follow-up, according to the route trajectory of the target customer, analyze the relevant business that the target customer is concerned about, so as to provide relevant business recommendations for the target customer.
  • the customer service device can also guide the staff to find the target customer to be served in time according to the location information of the overall image, so as to provide the service faster.
  • the above method further includes the following steps 1503-1505:
  • Step 1503 the customer service device determines the residence time of the target customer in at least one specific area according to the target customer's route track.
  • At least one specific area corresponds to at least one service.
  • the specific area may be set according to actual conditions, which is not limited in the present disclosure.
  • the customer service device may use the time period corresponding to the route trajectory of the target customer in the specific area as the residence time of the target customer in the specific area. Since each specific area corresponds to a business, the length of time a target customer stays in a specific area can represent the degree of attention of the target customer to the corresponding business in the specific area.
  • Step 1504 the customer service device determines that the target customer's concerned service is the service corresponding to the specific area with the longest residence time.
  • the customer service apparatus may regard the business corresponding to the specific area with the longest residence time as the concerned business of the target customer, and obtain the business information of the business.
  • Step 1505 the customer service device sends the service information of the concerned service to the service terminal.
  • the customer service device in this disclosure can analyze the business that the target customer is concerned about according to the route track of the target customer, and send the business information of the business to the service terminal of the assigned staff, so that the staff can provide targeted Provide services for target customers and improve the service experience of target customers.
  • the embodiments of the present disclosure can divide the customer service device into functional modules or functional units according to the above method examples.
  • each functional module or functional unit can be divided corresponding to each function, or two or more functions can be integrated into one processing module.
  • the above-mentioned integrated modules can be implemented not only in the form of hardware, but also in the form of software function modules or functional units.
  • the division of modules or units in the embodiments of the present disclosure is schematic, and is only a logical function division, and there may be another division manner in actual implementation.
  • FIG. 17 it is a schematic structural diagram of a customer service device 170 provided according to some embodiments, the device includes:
  • the processing unit 1701 is configured to determine the overall image and partial image of the target customer.
  • the processing unit 1701 is further configured to determine group attributes of target customers according to the overall image.
  • the group attribute is used to characterize the type of group to which the target customer belongs.
  • the processing unit 1701 is further configured to determine the individual attributes of the target customer according to the partial image.
  • the individual attribute includes at least one of gender, age and customer level.
  • the processing unit 1701 is further configured to determine the service priority of the target customer and the staff type corresponding to the target customer according to the group attribute of the target customer and the individual attribute of the target customer.
  • the processing unit 1701 is further configured to assign staff to the target customer according to the service priority of the target customer and the staff type corresponding to the target customer.
  • the communication unit 1702 is configured to send the customer information of the target customer and the current location information of the target customer to the service terminal; the processing unit 1701 is configured to determine according to the overall image of the staff and the overall image of the target customer The first location distance between the staff and the target customer; the communication unit 1702 is further configured to send service prompt information to the service terminal when the first location distance is greater than the first preset distance and the duration is longer than the first duration.
  • the communication unit 1702 is configured to receive registration information of the target customer; the registration information includes a partial image of the target customer and partial images of an accompanying person of the target customer.
  • the processing unit 1701 is configured to determine the overall image of the accompanying person according to the partial image of the accompanying person; the processing unit 1701 is configured to determine the target customer and the accompanying person according to the overall image of the target customer and the overall image of the accompanying person. The second position distance of the person; the processing unit 1701 is configured to assign the target customer to staff member.
  • the processing unit 1701 is configured to determine a target staff group, the type of the target staff group is the type of staff corresponding to the target customer, and the target staff group includes multiple target staff; the processing unit 1701. It is also configured to determine the personnel priority of the target staff group according to the personnel attributes of the target staff group; the processing unit 1701 is also configured to Prioritize, assign target workers to target customers.
  • the communication unit 1702 is configured to acquire the image to be detected; the image to be detected includes the target customer; the processing unit 1701 is configured to determine the first image area in the image to be detected, and the corresponding The first numerical value and the second numerical value; the first numerical value represents the probability that the first image region includes the overall image; the second numerical value represents the probability that the first image region includes a partial image; the first image region is any region in the image to be detected; The processing unit 1701 is configured to determine that the first image region includes the overall image of the target customer when the first numerical value is greater than the first threshold; the processing unit 1701 is configured to, when the second numerical value is greater than the second threshold, It is determined that the first image area includes a partial image of the target customer.
  • the processing unit 1701 is configured to input the overall image into the group attribute prediction model to determine the probability of different group attributes of the target customer; the processing unit 1701 is configured to, according to the probability of different group attributes of the target customer, Determine the group attributes of target customers.
  • different group attributes include at least one of mutually exclusive group attributes and independent group attributes; the processing unit 1701 is configured to determine the target when different group attributes include mutually exclusive group attributes
  • the customer's group attributes include a first mutually exclusive group attribute; the first mutually exclusive group attribute is the group attribute with the highest probability among the mutually exclusive group attributes; the processing unit 1701 is configured to include independent group attributes among different group attributes , determining that the group attribute of the target customer includes a first independent group attribute; the first independent group attribute is a group attribute with a probability greater than an attribute probability threshold among the independent group attributes.
  • the processing unit 1701 is configured to input the partial image into the individual attribute prediction model, and determine the gender probability corresponding to each gender of the target customer and the age probability corresponding to each age; the processing unit 1701 is configured to Determine the gender of the target customer as the gender with the highest gender probability; the processing unit 1701 is configured to determine the age of the target customer as the age with the highest age probability.
  • the processing unit 1701 is configured to detect whether the target customer exists in the user database according to the partial image of the target customer; the communication unit 1702 is configured to obtain the target customer's information if the target customer exists in the user database. customer level.
  • the communication unit 1702 is configured to obtain the basic information of the target customer if the target customer exists in the user database; the communication unit 1702 is configured to send the basic information of the target customer to the service terminal; the service terminal A terminal used by the staff; the communication unit 1702 is configured to obtain the historical business of the target customer under the condition that the target customer authorizes to inquire about the historical business; the communication unit 1702 is configured to send the historical business of the target customer to the service terminal.
  • the processing unit 1701 is configured to determine the attribute factor of the target customer according to the group attribute and the individual attribute; the attribute factor is used to represent the influence degree of the attribute of the target customer on the service priority; the processing unit 1701 is configured to Obtain the queue information of the target customer, and determine the queue factor according to the queue information; the queue factor is used to indicate the influence degree of the queue information of the target customer on the service priority; the processing unit 1701 is configured to obtain the reservation business of the target customer, and determine the The business determines the business factor; the business factor is used to indicate the impact degree of the target customer's reservation business on the service priority; the processing unit 1701 is configured to determine the target customer's service priority according to the attribute factor, the queue factor and the business factor.
  • the processing unit 1701 is configured to determine multiple overall images of the target customer at different moments in the images captured by multiple image acquisition devices; the processing unit 1701 is configured to Information to determine the route trajectory of target customers.
  • the processing unit 1701 is configured to determine the residence time of the target customer in at least one specific area according to the route track of the target customer; at least one specific area corresponds to at least one business; the processing unit 1701 is configured to determine The concerned service of the target customer is the service corresponding to the specific area with the longest residence time; the communication unit 1702 is configured to send the service information of the concerned service to the service terminal.
  • the communication unit 1702 in the embodiment of the present disclosure may be integrated on a communication interface, and the processing unit 1701 may be integrated on a processor.
  • the specific implementation is shown in Figure 18.
  • Fig. 18 shows another possible structural diagram of the customer service device involved in the above embodiment.
  • the customer service device 180 includes: a processor 1802 and a communication interface 1803 .
  • the processor 1802 is configured to control and manage the actions of the customer service device 180, for example, to execute the steps executed by the above-mentioned processing unit 1701, and/or configured to execute other processes of the technologies described herein.
  • the communication interface 1803 is configured to support the communication between the customer service device 180 and other network entities, for example, to perform the steps performed by the above-mentioned communication unit 1702 .
  • the customer service device 180 may further include a memory 1801 and a bus 1804 , and the memory 1801 is configured to store program codes and data of the customer service device 180 .
  • the memory 1801 may be a memory in the customer service device 180, etc., and the memory may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a read-only memory, a flash memory, Hard disk or solid state disk; the storage may also include a combination of the above-mentioned types of storage.
  • a volatile memory such as a random access memory
  • the memory may also include a non-volatile memory, such as a read-only memory, a flash memory, Hard disk or solid state disk
  • the storage may also include a combination of the above-mentioned types of storage.
  • the aforementioned processor 1802 may implement or execute various exemplary logic blocks, modules and circuits described in conjunction with the present disclosure.
  • the processor may be a central processing unit, a general purpose processor, a digital signal processor, an application specific integrated circuit, a field programmable gate array or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. It may implement or execute the various illustrative logical blocks, modules and circuits described in connection with the present disclosure.
  • the processor may also be a combination of computing functions, for example, a combination of one or more microprocessors, a combination of DSP and a microprocessor, and the like.
  • the bus 1804 may be an Extended Industry Standard Architecture (Extended Industry Standard Architecture, EISA) bus or the like.
  • EISA Extended Industry Standard Architecture
  • the bus 1804 can be divided into address bus, data bus, control bus and so on. For ease of representation, only one thick line is used in FIG. 18 , but it does not mean that there is only one bus or one type of bus.
  • the customer service device 180 in FIG. 18 can also be a chip.
  • the chip includes one or more than two (including two) processors 1802 and a communication interface 1803 .
  • the chip further includes a memory 1801 , which may include read-only memory and random access memory, and provides operation instructions and data to the processor 1802 .
  • a part of the memory 1801 may also include a non-volatile random access memory (non-volatile random access memory, NVRAM).
  • the memory 1801 stores the following elements, execution modules or data structures, or their subsets, or their extended sets.
  • the corresponding operation is executed by calling the operation instruction stored in the memory 1801 (the operation instruction may be stored in the operating system).
  • Some embodiments of the present disclosure provide a computer-readable storage medium (for example, a non-transitory computer-readable storage medium) in which computer program instructions are stored, and the computer program instructions are executed on a computer (for example, a client When running on the service device), the computer is made to execute the customer service method as described in any one of the above embodiments.
  • a computer-readable storage medium for example, a non-transitory computer-readable storage medium
  • the computer for example, a client When running on the service device
  • the computer is made to execute the customer service method as described in any one of the above embodiments.
  • the above-mentioned computer-readable storage medium may include, but is not limited to: magnetic storage devices (such as hard disks, floppy disks, or magnetic tapes, etc.), optical disks (such as CDs (Compact Disks, compact disks), DVDs (Digital Versatile Disks, Digital Versatile Disk), etc.), smart cards and flash memory devices (for example, EPROM (Erasable Programmable Read-Only Memory, Erasable Programmable Read-Only Memory), card, stick or key drive, etc.).
  • Various computer-readable storage media described in this disclosure can represent one or more devices and/or other machine-readable storage media for storing information.
  • the term "machine-readable storage medium” may include, but is not limited to, wireless channels and various other media capable of storing, containing and/or carrying instructions and/or data.
  • Some embodiments of the present disclosure also provide a computer program product, for example, the computer program product is stored on a non-transitory computer-readable storage medium.
  • the computer program product includes computer program instructions.
  • the computer program instructions When the computer program instructions are executed on a computer (eg, a customer service device), the computer program instructions cause the computer to execute the customer service method as described in the above-mentioned embodiments.
  • Some embodiments of the present disclosure also provide a computer program.
  • the computer program When the computer program is executed on a computer (for example, a customer service device), the computer program causes the computer to execute the customer service method as described in the above-mentioned embodiments.
  • the disclosed system, device and method may be implemented in other ways.
  • the device embodiments described above are only illustrative.
  • the division of the units is only a logical function division. In actual implementation, there may be other division methods.
  • multiple units or components can be combined or May be integrated into another system, or some features may be ignored, or not implemented.
  • the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, and the indirect coupling or communication connection of devices or units may be in electrical, mechanical or other forms.
  • the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.
  • each functional unit in each embodiment of the present disclosure may be integrated into one processing unit, each unit may exist separately physically, or two or more units may be integrated into one unit.

Abstract

一种客户服务方法、装置、系统及存储介质，包括：确定目标客户的整体图像与局部图像；根据整体图像确定目标客户的群体属性；群体属性用于表征目标客户所属群体的类型；根据局部图像确定目标客户的个体属性；个体属性包括性别、年龄以及客户级别中的至少一个；根据目标客户的群体属性和目标客户的个体属性，确定目标客户的服务优先级以及与目标客户对应的工作人员类型；根据目标客户的服务优先级以及与目标客户对应的工作人员类型，为目标客户分配工作人员。

Description

客户服务方法、装置、系统及存储介质 技术领域
本公开涉及数据处理领域，尤其涉及一种客户服务方法、装置、系统及存储介质。
背景技术
目前，为针对不同客户提供客户服务，相关技术一般通过获取客户人脸图像，并基于人脸识别技术检测数据库中是否存在与该客户人脸图像相匹配的客户身份信息。若数据库中存在相匹配的客户身份信息，则根据该身份信息为客户分配相应的工作人员。
发明内容
一方面，提供一种客户服务方法，该方法包括：确定目标客户的整体图像与局部图像；根据整体图像确定目标客户的群体属性；群体属性用于表征目标客户所属群体的类型；根据局部图像确定目标客户的个体属性；个体属性包括性别、年龄以及客户级别中的至少一个；根据目标客户的群体属性和目标客户的个体属性，确定目标客户的服务优先级以及与目标客户对应的工作人员类型；根据目标客户的服务优先级以及与目标客户对应的工作人员类型，为目标客户分配工作人员。
在一些实施例中，该方法还包括：向服务终端发送目标客户的客户信息和目标客户的当前位置信息；根据工作人员的整体图像与目标客户的整体图像确定工作人员与目标客户的第一位置距离；在第一位置距离大于第一预设距离且持续时长大于第一时长的情况下，向服务终端发送服务提示信息。
在一些实施例中，该方法还包括：接收目标客户的注册信息；注册信息包括目标客户的局部图像以及目标客户的陪同人员的局部图像。
在一些实施例中，该方法包括：根据陪同人员的局部图像确定陪同人员的整体图像；根据目标客户的整体图像与陪同人员的整体图像确定目标客户与陪同人员的第二位置距离；在第二位置距离大于第二预设距离的情况下，根据目标客户的服务优先级以及与目标客户对应的工作人员类型，为目标客户分配工作人员。
在一些实施例中，该方法包括：确定目标工作人员群组，目标工作人员群组的类型为目标客户对应的工作人员类型，目标工作人员群组包括多个目标工作人员；根据目标工作人员群组的人员属性，确定目标工作人员群组的人员优先级；根据目标客户的服务优先级以及目标工作人员群组的人员优先 级，为目标客户分配目标工作人员。
在一些实施例中，该方法包括：获取待检测图像；待检测图像包括目标客户；确定待检测图像中的第一图像区域，以及第一图像区域对应的第一数值和第二数值；第一数值表示第一图像区域包括整体图像的概率；第二数值表示第一图像区域包括局部图像的概率；第一图像区域为待检测图像中的任一个区域；在第一数值大于第一阈值的情况下，确定第一图像区域包括目标客户的整体图像；在第二数值大于第二阈值的情况下，确定第一图像区域包括目标客户的局部图像。
在一些实施例中，该方法包括：将整体图像输入群体属性预测模型中，确定目标客户的不同群体属性的概率；根据目标客户的不同群体属性的概率，确定目标客户的群体属性。
在一些实施例中，不同群体属性之间包括互斥群体属性和独立群体属性中的至少一项；该方法包括：在不同群体属性之间包括互斥群体属性的情况下，确定目标客户的群体属性包括第一互斥群体属性；第一互斥群体属性为互斥群体属性中概率最大的群体属性；在不同群体属性之间包括独立群体属性的情况下，确定目标客户的群体属性包括第一独立群体属性；第一独立群体属性为独立群体属性中概率大于属性概率阈值的群体属性。
在一些实施例中，该方法包括：将局部图像输入个体属性预测模型中，确定目标客户的每个性别对应的性别概率与每个年龄对应的年龄概率；确定目标客户的性别为性别概率最大的性别；确定目标客户的年龄为年龄概率最大的年龄。
在一些实施例中，该方法包括：根据目标客户的局部图像检测用户数据库中是否存在目标客户；在用户数据库中存在目标客户的情况下，获取目标客户的客户级别。
在一些实施例中，该方法还包括：在用户数据库中存在目标客户的情况下，获取目标客户的基本信息；向服务终端发送目标客户的基本信息；服务终端为工作人员使用的终端；在目标客户授权允许查询历史业务的情况下，获取目标客户的历史业务；向服务终端发送目标客户的历史业务。
在一些实施例中，该方法包括：根据群体属性与个体属性确定目标客户的属性因子；属性因子用于表示目标客户的属性对服务优先级的影响程度；获取目标客户的队列信息，并根据队列信息确定队列因子；队列因子用于表示目标客户的队列信息对服务优先级的影响程度；获取目标客户的预约业务，并根据预约业务确定业务因子；业务因子用于表示目标客户的预约业务对服 务优先级的影响程度；根据属性因子、队列因子以及业务因子确定目标客户的服务优先级。
在一些实施例中，该方法还包括：在多个图像采集装置采集的图像中，确定目标客户在不同时刻的多个整体图像；根据多个整体图像的位置信息，确定目标客户的路线轨迹。
在一些实施例中，该方法还包括：根据目标客户的路线轨迹确定目标客户在至少一个特定区域内的驻留时长；至少一个特定区域对应至少一个业务；确定目标客户的关注业务为驻留时长最长的特定区域对应的业务；向服务终端发送关注业务的业务信息；服务终端为工作人员使用的终端。
另一方面，提供一种客户服务装置，包括处理单元，被配置为确定目标客户的整体图像与局部图像；处理单元，还被配置为根据整体图像确定目标客户的群体属性；群体属性用于表征目标客户所属群体的类型；处理单元，还被配置为根据局部图像确定目标客户的个体属性；个体属性包括性别、年龄以及客户级别中的至少一个；处理单元，还被配置为根据目标客户的群体属性和目标客户的个体属性，确定目标客户的服务优先级以及与目标客户对应的工作人员类型；处理单元，还被配置为根据目标客户的服务优先级以及与目标客户对应的工作人员类型，为目标客户分配工作人员。
在一些实施例中，通信单元，被配置为向服务终端发送目标客户的客户信息和目标客户的当前位置信息；处理单元，被配置为根据工作人员的整体图像与目标客户的整体图像确定工作人员与目标客户的第一位置距离；通信单元，还被配置为在第一位置距离大于第一预设距离且持续时长大于第一时长的情况下，向服务终端发送服务提示信息。
在一些实施例中，通信单元，被配置为接收目标客户的注册信息；注册信息包括目标客户的局部图像以及目标客户的陪同人员的局部图像。
在一些实施例中，处理单元，被配置为根据陪同人员的局部图像确定陪同人员的整体图像；处理单元，被配置为根据目标客户的整体图像与陪同人员的整体图像确定目标客户与陪同人员的第二位置距离；处理单元，被配置为在第二位置距离大于第二预设距离的情况下，根据目标客户的服务优先级以及与目标客户对应的工作人员类型，为目标客户分配工作人员。
在一些实施例中，处理单元，被配置为确定目标工作人员群组，目标工作人员群组的类型为目标客户对应的工作人员类型，目标工作人员群组包括多个目标工作人员；处理单元，还被配置为根据目标工作人员群组的人员属性，确定目标工作人员群组的人员优先级；处理单元，还被配置为根据目标 客户的服务优先级以及目标工作人员群组的人员优先级，为目标客户分配目标工作人员。
在一些实施例中，通信单元，被配置为获取待检测图像；待检测图像包括目标客户；处理单元，被配置为确定待检测图像中的第一图像区域，以及第一图像区域对应的第一数值和第二数值；第一数值表示第一图像区域包括整体图像的概率；第二数值表示第一图像区域包括局部图像的概率；第一图像区域为待检测图像中的任一个区域；处理单元，被配置为在第一数值大于第一阈值的情况下，确定第一图像区域包括目标客户的整体图像；处理单元，被配置为在第二数值大于第二阈值的情况下，确定第一图像区域包括目标客户的局部图像。
在一些实施例中，处理单元，被配置为将整体图像输入群体属性预测模型中，确定目标客户的不同群体属性的概率；处理单元，被配置为根据目标客户的不同群体属性的概率，确定目标客户的群体属性。
在一些实施例中，不同群体属性之间包括互斥群体属性和独立群体属性中的至少一项；处理单元，被配置为在不同群体属性之间包括互斥群体属性的情况下，确定目标客户的群体属性包括第一互斥群体属性；第一互斥群体属性为互斥群体属性中概率最大的群体属性；处理单元，被配置为在不同群体属性之间包括独立群体属性的情况下，确定目标客户的群体属性包括第一独立群体属性；第一独立群体属性为独立群体属性中概率大于属性概率阈值的群体属性。
在一些实施例中，处理单元，被配置为将局部图像输入个体属性预测模型中，确定目标客户的每个性别对应的性别概率与每个年龄对应的年龄概率；处理单元，被配置为确定目标客户的性别为性别概率最大的性别；处理单元，被配置为确定目标客户的年龄为年龄概率最大的年龄。
在一些实施例中，处理单元，被配置为根据目标客户的局部图像检测用户数据库中是否存在目标客户；通信单元，被配置为在用户数据库中存在目标客户的情况下，获取目标客户的客户级别。
在一些实施例中，通信单元，被配置为在用户数据库中存在目标客户的情况下，获取目标客户的基本信息；通信单元，被配置为向服务终端发送目标客户的基本信息；服务终端为工作人员使用的终端；通信单元，被配置为在目标客户授权允许查询历史业务的情况下，获取目标客户的历史业务；通信单元，被配置为向服务终端发送目标客户的历史业务。
在一些实施例中，处理单元，被配置为根据群体属性与个体属性确定目 标客户的属性因子；属性因子用于表示目标客户的属性对服务优先级的影响程度；处理单元，被配置为获取目标客户的队列信息，并根据队列信息确定队列因子；队列因子用于表示目标客户的队列信息对服务优先级的影响程度；处理单元，被配置为获取目标客户的预约业务，并根据预约业务确定业务因子；业务因子用于表示目标客户的预约业务对服务优先级的影响程度；处理单元，被配置为根据属性因子、队列因子以及业务因子确定目标客户的服务优先级。
在一些实施例中，处理单元，被配置为在多个图像采集装置采集的图像中，确定目标客户在不同时刻的多个整体图像；处理单元，被配置为根据多个整体图像的位置信息，确定目标客户的路线轨迹。
在一些实施例中，处理单元，被配置为根据目标客户的路线轨迹确定目标客户在至少一个特定区域内的驻留时长；至少一个特定区域对应至少一个业务；处理单元，被配置为确定目标客户的关注业务为驻留时长最长的特定区域对应的业务；通信单元，被配置为向服务终端发送关注业务的业务信息；服务终端为工作人员使用的终端。
再一方面，提供一种非暂态计算机可读存储介质。所述计算机可读存储介质存储有计算机程序指令，所述计算机程序指令在计算机(例如，客户服务装置)上运行时，使得所述计算机执行如上述任一实施例所述的客户服务方法。
又一方面，提供一种计算机程序产品。所述计算机程序产品包括计算机程序指令，在计算机(例如，客户服务装置)上执行所述计算机程序指令时，所述计算机程序指令使计算机执行如上述任一实施例所述的客户服务方法。
又一方面，提供一种计算机程序。当所述计算机程序在计算机(例如，客户服务装置)上执行时，所述计算机程序使计算机执行如上述任一实施例所述的客户服务方法。
又一方面，提供一种芯片，芯片包括处理器和通信接口，通信接口和处理器耦合，处理器用于运行计算机程序或指令，以实现如上述任一实施例所述的客户服务方法。
示例性的，本公开中提供的芯片还包括存储器，用于存储计算机程序或指令。
需要说明的是，上述计算机指令可以全部或者部分存储在计算机可读存储介质上。其中，计算机可读存储介质可以与装置的处理器封装在一起的，也可以与装置的处理器单独封装，本公开对此不作限定。
又一方面，提供一种客户服务系统，包括：客户服务装置和至少一个图像采集装置，其中客户服务装置用于执行如上述任一实施例所述的客户服务方法。
在本公开中，上述客户服务装置的名字对设备或功能模块本身不构成限定，在实际实现中，这些设备或功能模块可以以其他名称出现。只要各个设备或功能模块的功能和本公开类似，属于本公开权利要求及其等同技术的范围之内。
附图说明
为了更清楚地说明本公开中的技术方案，下面将对本公开一些实施例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公开的一些实施例的附图，对于本领域普通技术人员来讲，还可以根据这些附图获得其他的附图。此外，以下描述中的附图可以视作示意图，并非对本公开实施例所涉及的产品的实际尺寸、方法的实际流程、信号的实际时序等的限制。
图1为根据一些实施例提供的一种客户服务系统的结构图；
图2为根据一些实施例提供的图像采集装置获取的图像数据的示例图；
图3为根据一些实施例提供的一种客户服务装置的结构图；
图4为根据一些实施例提供另一种客户服务方法的流程图；
图5为根据一些实施例提供的另一种客户服务方法的流程图；
图6为根据一些实施例提供的另一种客户服务方法的流程图；
图7为根据一些实施例提供的另一种客户服务方法的流程图；
图8为根据一些实施例提供的另一种客户服务方法的流程图；
图9为根据一些实施例提供的待检测图像的示例图；
图10为根据一些实施例提供的另一种客户服务方法的流程图；
图11为根据一些实施例提供的另一种客户服务方法的流程图；
图12为根据一些实施例提供的另一种客户服务方法的流程图；
图13为根据一些实施例提供的另一种客户服务方法的流程图；
图14为根据一些实施例提供的一种年龄评估值与年龄的函数图；
图15为根据一些实施例提供的另一种客户服务方法的流程图；
图16为根据一些实施例提供的整体图像在目标区域中的映射图；
图17为根据一些实施例提供的另一种客户服务装置的结构图；
图18为根据一些实施例提供的另一种客户服务装置的结构图。
具体实施方式
下面将结合附图，对本公开一些实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本公开一部分实施例，而不是全部的实施例。基于本公开所提供的实施例，本领域普通技术人员所获得的所有其他实施例，都属于本公开保护的范围。
除非上下文另有要求，否则，在整个说明书和权利要求书中，术语“包括(comprise)”及其其他形式例如第三人称单数形式“包括(comprises)”和现在分词形式“包括(comprising)”被解释为开放、包含的意思，即为“包含，但不限于”。在说明书的描述中，术语“一个实施例(one embodiment)”、“一些实施例(some embodiments)”、“示例性实施例(exemplary embodiments)”、“示例(example)”、“特定示例(specific example)”或“一些示例(some examples)”等旨在表明与该实施例或示例相关的特定特征、结构、材料或特性包括在本公开的至少一个实施例或示例中。上述术语的示意性表示不一定是指同一实施例或示例。此外，所述的特定特征、结构、材料或特点可以以任何适当方式包括在任何一个或多个实施例或示例中。
以下，术语“第一”、“第二”仅用于描述目的，而不能理解为指示或暗示相对重要性或者隐含指明所指示的技术特征的数量。由此，限定有“第一”、“第二”的特征可以明示或者隐含地包括一个或者更多个该特征。在本公开实施例的描述中，除非另有说明，“多个”的含义是两个或两个以上。
在描述一些实施例时，可能使用了“耦接”和“连接”及其衍伸的表达。例如，描述一些实施例时可能使用了术语“连接”以表明两个或两个以上部件彼此间有直接物理接触或电接触。又如，描述一些实施例时可能使用了术语“耦接”以表明两个或两个以上部件有直接物理接触或电接触。然而，术语“耦接”或“通信耦合(communicatively coupled)”也可能指两个或两个以上部件彼此间并无直接接触，但仍彼此协作或相互作用。这里所公开的实施例并不必然限制于本文内容。
“A、B和C中的至少一个”与“A、B或C中的至少一个”具有相同含义，均包括以下A、B和C的组合：仅A，仅B，仅C，A和B的组合，A和C的组合，B和C的组合，及A、B和C的组合。
“A和/或B”，包括以下三种组合：仅A，仅B，及A和B的组合。
如本文中所使用，根据上下文，术语“如果”任选地被解释为意思是“当……时”或“在……时”或“响应于确定”或“响应于检测到”。类似地，根据上下文，短语“如果确定……”或“如果检测到[所陈述的条件或事 件]”任选地被解释为是指“在确定……时”或“响应于确定……”或“在检测到[所陈述的条件或事件]时”或“响应于检测到[所陈述的条件或事件]”。
本文中“适用于”或“被配置为”的使用意味着开放和包容性的语言，其不排除适用于或被配置为执行额外任务或步骤的设备。
另外，“基于”的使用意味着开放和包容性，因为“基于”一个或多个所述条件或值的过程、步骤、计算或其他动作在实践中可以基于额外条件或超出所述的值。
如本文所使用的那样，“约”、“大致”或“近似”包括所阐述的值以及处于特定值的可接受偏差范围内的平均值，其中所述可接受偏差范围如由本领域普通技术人员考虑到正在讨论的测量以及与特定量的测量相关的误差(即，测量系统的局限性)所确定。
以下，对本公开实施例涉及的名词进行解释，以方便读者理解。
(1)行人重识别
行人重识别(pedestrian re-identification，ReID)是指通过计算机视觉(computer vision)技术对图像或视频流数据中的特定行人进行检测的技术。ReID通常用于从至少一个图像采集装置(即跨设备)获取的图像信息中确定行人图像，从而对该目标进行检测或跟踪。
(2)神经网络
神经网络(neural networks，NNs)也称作人工神经网络(artificial neural networks，ANNs)，是一种模仿动物神经网络行为特征，进行分布式并行信息处理的数学模型算法。神经网络包括深度学习网络，例如卷积神经网络(convolutional neural networks，CNN)、残差网络(residual network，ResNet)、长短期记忆网络(long short-term memory，LSTM)等。
(3)交并比
交并比(intersection over union，IoU)是指两个图像区域的交集面积与两个图像区域的并集面积的比值，通常用于检测两个图像区域的关系，例如通过计算预测图像区域与实际图像区域的IoU确定预测的准确性。
(4)L1范数(L1 norm)
L1范数也称作曼哈顿距离，用于表示向量空间或矩阵中向量的长度。L1范数正则化(L1 regularization)通常应用于机器学习算法中的损失函数，以对目标客户进行预测或分类。
通常，为了针对不同的客户提供相应的客户服务，一般通过获取客户人脸图像，并基于人脸识别技术检测数据库中是否存在与该客户人脸图像相匹 配的客户身份信息。若数据库中存在相匹配的客户身份信息，则根据该身份信息为客户分配相应的工作人员。然而该方案需要预先在数据库中存储客户的相关信息，无法适用于未在数据库中注册的客户群体，导致客户服务效果较差。
鉴于此，本公开提供了一种客户服务方法，基于目标客户的整体图像与局部图像确定目标客户的群体属性与个体属性，并通过从群体属性与个体属性两个维度为目标客户分配工作人员。因此，本公开的适用范围更广，同时能够提高为目标客户提供服务业务的精准度与智能化水平，保障了目标客户的用户体验。
下面将结合说明书附图，对本公开实施例的实施方式进行详细描述。
图1为根据一些实施例提供的一种客户服务系统10的结构图，如图1所示，该客户服务系统10包括：客户服务装置101、图像采集装置102、服务终端103以及用户终端104。
需要说明的是，本公开中的客户服务装置101、图像采集装置102、服务终端103以及用户终端104可以为一个，也可以为多个。为方便理解，图1中仅示出了一个客户服务装置101、图像采集装置102、服务终端103以及用户终端104。
如图1所示，客户服务装置101与图像采集装置102通过通信链路连接，客户服务装置101与服务终端103通过通信链路连接，客户服务装置101与用户终端104通过通信链路连接。该通信链路可以为有线通信链路，也可以为无线通信链路，本公开对此不予限制。
客户服务装置101被配置为确定目标客户的整体图像与局部图像，并根据目标客户的整体图像与局部图像确定目标客户的相关属性，从而根据目标客户的相关属性为目标客户分配工作人员。
示例性的，目标客户可以为一个，也可以为多个。当目标客户为多个时，客户服务装置101分别确定多个目标客户的整体图像与局部图像，从而为多个目标客户分配工作人员。
整体图像为表征目标客户整体画面信息的图像，局部图像为表征目标客户部分画面信息的图像。
一种可能的实现方式中，客户服务装置101接收图像采集装置102发送的图像数据，并根据图像数据确定目标客户的整体图像与局部图像。
示例性的，如图2所示，图2为根据一些实施例提供的图像采集装置获取的图像数据20的示例图。客户服务装置101接收图像采集装置102发送的 图像数据20。客户服务装置101根据图像数据20确定目标客户1的整体图像为图像201，局部图像为图像203、图像204以及图像205。客户服务装置101根据图像数据20确定目标客户2的整体图像为图像202，局部图像为图像206、图像207以及图像208。
需要说明的是，工作人员可以为向目标客户提供服务的服务人员，也可以为向目标客户提供服务的服务设备。例如，该设备为机器人、终端设备、网点设施等。
客户服务装置101还被配置为向服务终端103发送目标客户的客户信息。相应的，服务终端103接收客户服务装置101发送的客户信息。
示例性的，服务终端103可以为工作人员使用的终端。客户信息包括目标客户的相关属性、基本信息、历史业务、或关注业务中的至少一项。
示例性的，客户信息可以包括目标客户的局部图像，性别、年龄、陪同人员的局部图像等信息。
当工作人员为向目标客户提供服务的设备时，服务终端103可以耦合在该设备中。
需要说明的是，客户服务装置101可以为服务器，或者，客户服务装置101可以为该服务器安装的提供客户服务功能的应用程序，或者，客户服务装置101可以为服务器中的中央处理器，又或者，客户服务装置101可以为该服务器中被配置为执行客户服务方法的控制模块。
作为一种可能的实现方式，如图3所示，图3为根据一些实施例提供的一种客户服务装置101的结构图。该客户服务装置101包括：数据采集模块301、检测与跟踪模块302、属性分析模块303、数据分析模块304以及数据存储与发送模块305。
数据采集模块301被配置为获取图像采集装置102发送的图像数据。
检测与跟踪模块302被配置为从图像数据中确定目标客户的整体图像与局部图像。检测与跟踪模块302还被配置为对目标客户进行跟踪，确定目标客户的路线轨迹。
属性分析模块303被配置为根据整体图像确定目标客户的群体属性。属性分析模块还被配置为根据局部图像确定目标客户的个体属性。
其中，群体属性用于表征目标客户所属群体的类型，个体属性包括性别、年龄以及客户级别中的至少一个。
数据分析模块304被配置为根据目标客户的群体属性和目标客户的个体属性，为目标客户分配工作人员。
数据存储与发送模块305被配置为存储上述模块在执行过程中待存储的数据。数据存储与发送模块305还被配置为向服务终端103发送目标客户的客户信息。
图像采集装置102被配置为获取目标区域的图像数据，并向客户服务装置101发送该图像数据。相应的，客户服务装置101接收图像采集装置102发送的图像数据。
示例性的，目标区域可以为提供服务的经营场所，例如目标区域可以为银行、通信营业厅、服装店、酒店等场所。目标区域还可以为展示场所，例如目标区域为博物馆、科技馆、书画展、车展等场所。目标区域还可以为赛事场所，例如目标区域为球赛展厅、车赛展厅等场所。
一种可能的实现方式中，图像采集装置102可以实时获取目标区域的图像数据，并向客户服务装置101发送该图像数据。图像采集装置102还可以按照预设频率获取目标区域的图像数据，并向客户服务装置101发送该图像数据。
本公开实施例中的图像采集装置102为通过感光器将图像数据转换为模拟信号或者数字信号的装置，可以部署在陆地上，包括室内或室外、手持或车载。也可以部署在水面上(如轮船等)。还可以部署在空中(例如飞机、气球和卫星上等)。例如，图像采集装置102包括摄像头、摄像机、相机。图像采集装置102也可以是具有摄像功能的设备，例如，图像采集装置102可以为具有摄像功能的手机、平板电脑、笔记本电脑、掌上电脑、可穿戴设备(例如智能手表、智能手环、计步器等)，车载设备飞行设备(例如，智能机器人、热气球、无人机、飞机)等。
示例性的，本公开实施例中的图像采集装置102还可以是红外成像仪或者夜视仪，用于获取黑暗区域的图像数据。
服务终端103被配置为接收客户服务装置101发送的客户信息。服务终端103还被配置为输出客户信息。
示例性的，服务终端103可以通过画面、文字或者声音的形式输出客户信息，以使得工作人员获取该客户信息。
用户终端103被配置为向客户服务装置101发送预约业务信息，相应的，客户服务装置101接收用户终端103发送的预约业务信息。
用户终端103还被配置为接收客户服务装置101发送的请求消息。
其中，用户终端103可以为目标客户或陪同人员使用的终端设备。
终端设备又称之为用户设备(user equipment，UE)，移动台(mobile station， MS)、移动终端(mobile terminal，MT)等，是一种向用户提供语音和/或数据连通性的设备。例如，终端包括具有无线连接功能的手持式设备、车载设备等。目前，终端设备可以是：手机(mobile phone)、平板电脑、笔记本电脑、掌上电脑、移动互联网设备(mobile internet device，MID)、可穿戴设备(例如智能手表、智能手环、计步器等)，车载设备(例如，汽车、自行车、电动车、飞机、船舶、火车、高铁等)、虚拟现实(virtual reality，VR)设备、增强现实(augmented reality，AR)设备、工业控制(industrial control)中的无线终端、智能家居设备(例如，冰箱、电视、空调、电表等)、智能机器人、车间设备、无人驾驶(self driving)中的无线终端、远程手术(remote medical surgery)中的无线终端、智能电网(smart grid)中的无线终端、运输安全(transportation safety)中的无线终端、智慧城市(smart city)中的无线终端，或智慧家庭(smart home)中的无线终端、飞行设备(例如，智能机器人、热气球、无人机、飞机)等。本公开一种可能的应用的场景中终端设备为经常工作在地面的终端设备，例如车载设备。在本公开中，为了便于叙述，部署在上述设备中的芯片，例如片上系统(System-On-a-Chip，SOC)、基带芯片等，或者其他具备通信功能的芯片也可以称为终端设备。
需要指出的是，本公开各实施例之间可以相互借鉴或参考，例如，相同或相似的步骤，方法实施例、系统实施例和装置实施例之间，均可以相互参考，不予限制。
图4为根据一些实施例提供的一种客户服务方法的流程图。如图4所示，该方法包括以下步骤：
步骤401、客户服务装置确定目标客户的整体图像与局部图像。
其中，整体图像为表征目标客户整体画面信息的图像，局部图像为表征目标客户部分画面信息的图像。
示例性的，整体图像可以为包括目标客户的图像(如图2中的图像201)，局部图像可以为目标客户的脸部图像(如图2中的图像203)，也可以为目标客户的手部图像(如图2中的图像204)，还可以为目标客户的脚部图像(如图2中的图像205)。本公开以局部图像为脸部图像为例，对本公开提供的客户服务方法进行具体描述。
一种可能的实现方式中，客户服务装置从图像采集装置中获取目标区域的图像，并将目标区域的图像输入目标检测模型中，得到目标客户的整体图像与局部图像。
示例性的，目标识别模型可以为神经网络模型，例如YOLO(you only look  once)算法模型、SSD(single shot multibox detector)算法模型、R-CNN(region CNN)算法模型、Faster R-CNN算法模型。
其中，目标客户可以为一个，也可以为多个。当目标客户为多个时，客户服务装置分别确定多个目标客户的整体图像与局部图像。
需要说明的是，由于同一目标客户的整体图像与局部图像具有一定的关联关系(例如IoU大于预设阈值)，因此客户服务装置可以确定满足关联关系的整体图像、局部图像与目标客户相对应。
一种可能的实现方式中，客户服务装置通过整体图像与局部图像的关联关系确定将目标客户与整体图像、局部图像进行绑定。
示例性的，如图2所示，客户服务装置确定整体图像包括图像201与图像202，局部图像包括图像203与图像206。客户服务装置确定图像201与图像203的IoU，图像201与图像206的IoU，图像202与图像203的IoU，图像202与图像206的IoU。根据上述确定的整体图像与局部图像的IoU，客户服务装置确定目标客户1与图像201、图像203相对应，目标客户2与图像202、图像206相对应。
步骤402、客户服务装置根据整体图像确定目标客户的群体属性。
其中，群体属性用于表征目标客户所属群体的类型。
示例性的，客户服务装置可以按照职业类型划分群体，例如群体属性包括工作人员、特殊职业人员以及其他人员。群体属性还可以包括其他特定群体，例如孕妇、残疾人士等行动不便的群体。
一种可能的实现方式中，客户服务装置可以将整体图像输入群体属性预测模型中，得到目标客户的群体属性。
示例性的，群体属性预测模型可以为神经网络模型，例如EfficientNet算法模型、残差网络(ResNet)算法模型。
此外，客户服务还可以获取与群体属性相关联的物体图像，根据物体图像与整体图像进一步确定目标客户的群体属性。
示例性的，客户服务可以检测轮椅、拐杖等物体图像，并计算该物体图像与整体图像的IoU，当IoU大于预设阈值时，确认目标客户的群体属性包括残疾人士。
一种可能的实现方式中，当目标客户的群体属性为工作人员时，客户服务装置还可以将群体属性为工作人员的目标客户过滤，以避免检测目标客户时出现误检问题，提高了目标客户检测的准确性。
需要说明的是，由于光照变化、目标客户的位置角度、其他物体等因素， 客户服务装置确定的整体图像会中存在遮挡或截断的情况。被遮挡、截断的整体图像会对客户服务装置确定目标客户的群体属性造成影响。因此，客户服务装置可以获取目标客户的多张整体图像，并从中筛选出图像质量较好的整体图像，从而提高客户服务装置确定群体属性的准确度。
一种可能的实现方式中，在步骤402之前，客户服务装置还可以根据整体图像确定整体图像的遮挡因子与截断因子，并根据遮挡因子与截断因子确定整体图像的质量评估值。
其中，遮挡因子用于表征整体图像的遮挡程度，截断因子用于表征整体图像的截断程度。
示例性的，整体图像的质量评估值满足以下公式1：
Q＝a*P 1+b*P 2                公式1
其中，Q为整体图像的质量评估值，P 1为遮挡因子，P 2为截断因子，a为遮挡因子的权重值，b为截断因子的权重值。
a、b的取值可根据实际情况设置，本公开对此不作限定。
当整体图像的质量评估值大于预设质量阈值时，则说明整体图像的图像质量较差，客户服务装置去除该整体图像。
当整体图像的质量评估值小于或等于预设质量阈值时，则说明整体图像的图像质量较好，客户服务装置保留该整体图像。
上述步骤402可以通过以下步骤实现：客户服务装置根据满足第一预设条件的整体图像确定目标客户的群体属性。
其中，第一预设条件为整体图像的质量评估值小于或等于预设质量阈值。
步骤403、客户服务装置根据局部图像确定目标客户的个体属性。
其中，个体属性用于表征目标客户个体的信息。例如，个体属性包括性别、年龄以及客户级别中的至少一个。
需要说明的是，局部图像能够体现目标客户细节方面的特点，因此通过局部图像，客户服务装置可以确定目标客户的表征个体信息的属性。
以局部图像为脸部图像为例，客户服务装置可以对脸部图像进行人脸关键点检测、人脸对齐等操作，获取目标客户的标准脸部图像，以便于后续确定个体属性。具体操作可参考相关技术，本公开对此不作过多详述。
一种可能的实现方式中，客户服务装置可以将局部图像输入个体属性预测模型中，得到目标客户的个体属性。
示例性的，个体属性预测模型可以为神经网络模型，例如ShuffleNet算法模型、残差网络(ResNet)算法模型。
一种可能的实现方式中，在步骤403之前，客户服务装置还可以根据局部图像确定局部图像的图像质量，并去除图像质量差的局部图像。具体实现方式可以参照上述步骤402中去除图像质量差的整体图像的相关实现方式，此处不再赘述。
示例性的，客户服务装置在确定目标客户的个体属性之后，还可以通过目标客户的个体属性与群体属性进行校正。例如，当客户服务装置确定目标客户的性别为男时，客户服务装置可以将目标客户的群体属性中的孕妇属性去除。如此一来，客户服务装置可以进一步提高确定目标客户属性的准确性。
步骤404、客户服务装置根据目标客户的群体属性和目标客户的个体属性，确定目标客户的服务优先级以及与目标客户对应的工作人员类型。
其中，服务优先级用于表示客户服务装置为目标客户分配工作人员的优先次序。根据目标客户的群体属性和目标客户的个体属性，客户服务装置可以提高例如孕妇、残疾人士等行动不便群体的目标客户的服务优先级，以保障各个用户群体的服务体验。
一种可能的实现方式中，客户服务装置还可以根据目标客户的群体属性、个体属性、队列信息、预约业务确定目标客户的服务优先级。具体可参考后续描述，此处不作过多赘述。
工作人员可以为向目标客户提供服务的服务人员，也可以为向目标客户提供服务的服务设备。例如，该设备为机器人、终端设备、网点设施等。
另外，客户服务装置还可以根据工作人员可服务的目标客户的群体属性和目标客户的个体属性对工作人员进行分类，以确定多个工作人员类型，从而有效地针对不同目标客户提供相应的服务。
步骤405、客户服务装置根据目标客户的服务优先级以及与目标客户对应的工作人员类型为目标客户分配工作人员。
在一些实施例中，客户服务装置根据服务优先级对应的优先次序，为目标客户分配对应的工作人员类型的工作人员。
示例性的，客户服务装置确定目标客户的服务优先级为高优先级，工作人员类型为第一工作人员类型。客户服务装置优先为该目标客户分配第一工作人员类型对应的工作人员。
一种可能的实现方式中，客户服务装置确定工作人员可服务的的目标客户的客户级别与工作人员类型的对应关系。客户服务装置根据目标客户的客户级别，为目标客户分配对应工作人员类型的工作人员。
又一种可能的实现方式中，客户服务装置按照处理不同业务的业务能力 对工作人员进行分类，并根据目标客户的预约业务类型为目标客户分配对应工作人员类型的工作人员。
示例性的，对于复杂业务，客户服务装置为目标客户分配处理复杂业务能力强的工作人员，对于简单业务，客户服务装置为目标客户分配自助服务设备。
需要说明的是，客户服务装置还可以根据目标客户所属群体确定对应的工作人员类型，并为目标客户分配工作人员。
示例性的，客户服务装置为工作人员可服务的特殊职业人员或者孕妇、残疾人士等行动不便的群体按照相关服务经验对工作人员进行分类，并分配对应类型的工作人员，以便于更好地提供服务。客户服务装置还可以按照工作效率对工作人员进行分类，为年轻群体分配工作效率高的工作人员。
基于上述技术方案，本公开中客户服务装置通过确定目标客户的整体图像与局部图像，并分别根据整体图像确定目标客户的群体属性，根据局部图像确定目标客户的个体属性。由于群体属性能够表征目标客户所属群体的类型，个体属性能够表征目标客户个体的信息，因此，客户服务装置可以从群体属性与个体属性两个维度为目标客户分配工作人员，从而能够提高为目标客户提供服务业务的精准度与智能化水平，保障了目标客户的用户体验。同时，客户服务装置还根据目标客户的群体属性和个体属性，确定目标客户的服务优先级以及与目标客户对应的工作人员类型，并按照服务优先级的优先次序为目标客户分配相应工作人员类型的工作人员。如此一来，客户服务装置可以针对性地为不同目标客户提供最佳的服务，保障了不同目标客户的服务体验。
以下，在客户服务装置为目标对象分配工作人员之后，对客户服务装置确定工作人员的服务状态的过程进行具体介绍。
作为本公开的一种可能的实施例，结合图4，如图5所示，在步骤405之后，上述方法还包括以下步骤501-步骤503：
步骤501、客户服务装置向服务终端发送目标客户的客户信息和目标客户的当前位置信息。
其中，服务终端为工作人员使用的终端。
需要说明的是，在为目标客户分配工作人员之后，客户服务装置可以向服务终端发送目标客户的客户信息和目标客户的当前位置信息，以便于工作人员及时为目标客户提供服务。
示例性的，客户信息可以包括目标客户的局部图像，整体图像、群体属 性、个体属性等信息。
其中，目标客户的位置信息可以通过目标客户的整体图像确定，具体实现方式可参考后续描述，此处不再赘述。
步骤502、客户服务装置根据工作人员的整体图像与目标客户的整体图像确定工作人员与目标客户的第一位置距离。
其中，客户服务装置可以预先存储工作人员的整体图像与局部图像，也可以通过上述步骤401-步骤403中的方式获取工作人员的整体图像。
例如，客户服务装置获取第一对象的整体图像与局部图像之后，根据第一对象的整体图像确定第一对象的群体属性。在第一对象的群体属性中包括工作人员时，客户服务装置确定该第一对象为工作人员。第一对象的整体图像即为该工作人员的整体图像。
在确定工作人员的整体图像与目标客户的整体图像之后，客户服务装置可以将整体图像映射在目标区域对应的坐标系中，得到工作人员与目标客户的位置坐标，从而根据位置坐标确定工作人员与目标客户的第一位置距离。
步骤503、在第一位置距离大于第一预设距离且持续时长大于第一时长的情况下，客户服务装置向服务终端发送服务提示信息。
其中，服务提示信息用于提示工作人员及时为目标客户提供服务。
第一预设距离与第一时长的具体数值可根据实际情况设置，本公开对此不作限定。
需要说明的是，客户服务装置可以通过工作人员与目标客户的距离判断工作人员是否到达目标客户身边。在工作人员与目标客户的距离大于第一预设距离时，说明工作人员并未向目标客户提供服务。当持续时长大于第一时长时，客户服务装置可以向服务终端发送服务提示信息以提示工作人员及时为目标客户提供服务。
一种可能的实现方式中，在客户服务装置向服务终端发送服务提示信息之后，客户服务装置可以按照预设间隔定时判断第一位置距离是否大于第一预设距离。
在第一位置距离仍然大于第一预设距离的情况下，客户服务装置可以再次向服务终端发送服务提示信息，或者重新为目标客户分配工作人员。
基于上述技术方案，本公开中的客户服务装置能够在为目标客户分配工作人员后，向服务终端发送目标客户的客户信息和目标客户的当前位置信息，以便于工作人员及时达到目标客户身边并为目标客户提供服务。同时，客户服务装置还会基于工作人员与目标客户的位置距离判断工作人员是否在为目 标客户提供服务，并在持续时间大于第一时长时提醒工作人员提供服务，从而保障目标客户的服务体验。
以下，对客户服务装置接收目标客户的注册信息的过程进行具体介绍。
作为本公开的一种可能的实施例，结合图4，如图6所示，在步骤401之前，上述方法还包括以下步骤601：
步骤601、客户服务装置接收目标客户的注册信息。
其中，注册信息包括目标客户的局部图像以及目标客户的陪同人员的局部图像。
一种可能的实现方式中，客户服务装置可以将获取的目标客户的局部图像以及目标客户的陪同人员的局部图像存储在用户数据库中，并设置目标客户与陪同人员的绑定关系，以便于后续客户服务装置获取目标客户与陪同人员的相关信息。
作为本公开的又一种可能的实施例，结合图4，如图6所示，在步骤405之后，上述方法还包括以下步骤602-步骤604：
步骤602、客户服务装置根据陪同人员的局部图像确定陪同人员的整体图像。
在确定目标客户的整体图像与局部图像之后，客户服务装置可以将目标客户的局部图像与用户数据库中每个用户的局部图像进行匹配。
在用户数据库中存在目标客户的情况下，客户服务装置可以进一步检测该目标客户是否存在与其绑定的陪同人员，从而获取陪同人员的局部图像。
客户服务装置将目标客户的局部图像与用户数据库中每个用户的局部图像进行匹配的过程可参考后续描述，此处不再赘述。
客户服务装置可以根据陪同人员的局部图像确定图像采集装置采集的图像中与该局部图像具有关联关系(例如IoU大于预设阈值)的整体图像，并将该整体图像作为陪同人员的整体图像。
具体实现方式可参考步骤401中的相关描述，此处不再赘述。
步骤603、客户服务装置根据目标客户的整体图像与陪同人员的整体图像确定目标客户与陪同人员的第二位置距离。
客户服务装置确定目标客户与陪同人员的第二位置距离的具体方式与上述步骤502中客户服务装置确定工作人员与目标客户的第一位置距离的实现方式相同，此处不再赘述。
步骤604、在第二位置距离大于第二预设距离的情况下，客户服务装置根据目标客户的服务优先级以及与目标客户对应的工作人员类型，为目标客户 分配工作人员。
第二预设距离可以根据实际情况设置，本公开对此不作限定。
需要说明的是，在第二位置距离大于第二预设距离的情况下，说明陪同人员不在目标客户身边，客户服务装置可以及时为目标客户分配工作人员，以便于为目标客户提供服务。
基于上述技术方案，本公开中的客户服务装置获取目标客户的陪同人员的局部图像从而确定目标客户与陪同人员的第二位置距离，在目标客户与陪同人员距离相差过大时，及时为目标客户分配工作人员，保障了目标客户的服务体验。
以下，结合上述步骤405，对客户服务装置为目标客户分配工作人员的过程进行具体介绍。
作为本公开的一种可能的实施例，结合图4，如图7所示，上述步骤405还可以通过以下步骤701-步骤703实现：
步骤701、客户服务装置确定目标工作人员群组。
其中，目标工作人员群组的类型为目标客户对应的工作人员类型，目标工作人员群组包括多个目标工作人员。
需要说明的是，在客户服务装置确定与目标客户对应的工作人员类型之后，客户服务装置可以基于工作人员类型确认该类型所对应的目标工作人员群组。
步骤702、客户服务装置根据目标工作人员群组的人员属性，确定目标工作人员群组的人员优先级。
其中，人员属性用于表示目标工作人员群组中每个目标工作人员的工作量。人员属性包括工作时长、工作强度以及工作频率中的至少一项。人员优先级用于表示目标工作人员群组中每个目标工作人员的分配次序。
需要说明的是，通过工作人员群组的人员属性，客户服务装置可以确定目标工作人员群组的人员优先级。
示例性的，当目标工作人员的工作时长越长、工作强度越大以及工作频率越高时，客户服务装置确定该目标工作人员的人员优先级越低。反之，当目标工作人员的工作时长越短、工作强度越小以及工作频率越低时，客户服务装置确定该目标工作人员的人员优先级越高。
一种可能的实现方式中，当目标工作人员为服务设备时，客户服务装置可以设置该目标工作人员的人员优先级为高优先级，从而提高服务设备的工作效率。
步骤703、客户服务装置根据目标客户的服务优先级以及目标工作人员群组的人员优先级，为目标客户分配目标工作人员。
示例性的，当目标客户的服务优先级越高时，客户服务装置优先为目标客户分配工作人员。当目标工作人员群组中该目标工作人员的人员优先级越高时，客户服务装置优先将该目标工作人员分配给相应目标客户。
基于上述技术方案，本公开中的客户服务装置能够确定待分配的目标工作人员群组，并且基于目标客户的客户优先级以及目标工作人员群组的人员优先级为目标客户分配目标工作人员。如此一来，客户服务装置能够调节每个工作人员的工作强度，避免由于工作人员分配不均而导致部分工作人员过于劳累的问题。
以下，结合上述步骤401，对客户服务装置确定目标客户的整体图像与局部图像的过程进行具体介绍。
作为本公开的一种可能的实施例，结合图4，如图8所示，上述步骤401还可以通过以下步骤801-步骤804实现：
步骤801、客户服务装置获取待检测图像。
其中，待检测图像包括目标客户。待检测图像为图像采集装置采集的目标区域的图像。图像采集装置可以为一个，也可以为多个。
示例性的，目标区域可以为提供服务的经营场所，例如目标区域为银行、通信营业厅、服装店、酒店等场所。目标区域还可以为展示场所，例如目标区域为博物馆、科技馆、书画展、车展等场所。目标区域还可以为赛事场所，例如目标区域为球赛展厅、车赛展厅等场所。
一种可能的实现方式中，客户服务装置接收一个或多个图像采集装置发送的待检测图像。
步骤802、客户服务装置确定待检测图像中的第一图像区域，以及第一图像区域对应的第一数值和第二数值。
其中，第一数值表示第一图像区域包括整体图像的概率。第二数值表示第一图像区域包括局部图像的概率。第一图像区域为待检测图像中的任一个区域。
示例性的，如图9所示，图9为根据一些实施例提供的待检测图像90的示例图。客户服务装置从待检测图像90中确定第一图像区域(即图9中的任一个虚线框)，以及第一图像区域对应的第一数值和第二数值。
步骤803、在第一数值大于第一阈值的情况下，客户服务装置确定第一图像区域包括目标客户的整体图像。
需要说明的是，第一数值表示第一图像区域包括整体图像的概率，因此第一数值越大，该第一图像区域包括目标客户的整体图像的概率也就越大。
示例性的，以第一阈值为0.9为例，当第一数值大于0.9时，客户服务装置确定第一图像区域包括的图像(如图9中的图像901)为目标客户的整体图像。
步骤804、在第二数值大于第二阈值的情况下，客户服务装置确定第一图像区域包括目标客户的局部图像。
同样的，第二数值越大，该第一图像区域包括目标客户的局部图像的概率也就越大，该第一图像区域包括目标客户的局部图像的概率也就越大。
示例性的，以第二阈值为0.9为例，当第二数值大于0.9时，客户服务装置确定第一图像区域包括的图像(如图9中的图像902)为目标客户的局部图像。
基于上述技术方案，本公开中的客户服务装置能够获取待检测图像并确定待检测图像中的第一图像区域以及对应的第一数值和第二数值。由于第一数值表示第一图像区域包括整体图像的概率，第二数值表示第一图像区域包括局部图像的概率，因此，客户服务装置能够基于所确定的第一图像区域包括整体图像的概率以及第一图像区域包括局部图像的概率确定待检测图像中目标客户的整体图像与局部图像，提高了图像检测的准确性，以便于后续根据整体图像与局部图像为目标客户分配工作人员。
以下，结合上述步骤402，对客户服务装置根据整体图像确定目标客户的群体属性的过程进行具体介绍。
作为本公开的一种可能的实施例，结合图4，如图10所示，上述步骤402还可以通过以下步骤1001-步骤1002实现：
步骤1001、客户服务装置将整体图像输入群体属性预测模型中，确定目标客户的不同群体属性的概率。
其中，群体属性预测模型用于确定目标客户的不同群体属性的概率。
示例性的，群体属性预测模型可以为神经网络模型，例如EfficientNet算法模型、残差网络(ResNet)算法模型。
其中，不同群体属性之间包括互斥群体属性和独立群体属性中的至少一项。
需要说明的是，互斥群体属性是指不具有交集的多个群体对应的属性。目标客户无法同时具有互斥群体属性中的多个群体属性。
例如，互斥群体属性包括工作人员、特殊职业人员以及其他人员。目标 客户可以为工作人员、特殊职业人员以及其他人员中的任一项。
独立群体属性是指互相独立，不受其他群体属性影响的属性。
例如，独立群体属性包括孕妇、残疾人士。目标客户可以为孕妇，也可以为残疾人士，还可以为孕妇以及残疾人士。
一种可能的实现方式中，客户服务装置将整体图像输入群体属性预测模型中，得到不同群体属性的概率因子。
其中，概率因子用于表征目标客户具有该群体属性的可信程度。
客户服务装置根据不同群体属性的概率因子确定目标客户的不同群体属性的概率。
示例性的，在不同群体属性之间包括互斥群体属性的情况下，客户服务装置将互斥群体属性的概率因子输入Softmax函数，得到互斥群体属性的概率。
Softmax函数满足以下公式2：
Figure PCTCN2022078111-appb-000001
其中，p i为互斥群体属性中第i个属性的概率，v i为互斥群体属性中第i个属性的概率因子，v j为互斥群体属性中第j个属性的概率因子。
通过上述公式2，客户服务装置将互斥群体属性的概率因子通过以自然常数e为底数的指数函数转换为非负数，并将第i个互斥群体属性转换后的概率因子作为分子，将互斥群体属性中的每个属性转换后的概率因子之和作为分母，从而得到该第i个互斥群体属性的概率。
例如，互斥群体属性1的概率因子为0.5，互斥群体属性2的概率因子为2，互斥群体属性3的概率因子为3。转换后的互斥群体属性1的概率因子为1.648721，转换后的互斥群体属性2的概率因子为7.389056，转换后的互斥群体属性3的概率因子为20.085537。因此，得到的互斥群体属性1的概率为5.6612％，互斥群体属性2的概率为25.3716％，互斥群体属性3的概率为68.9672％。
在不同群体属性之间包括独立群体属性的情况下，客户服务装置将独立群体属性的概率因子输入Sigmoid函数，得到独立群体属性的概率。
Sigmoid函数满足以下公式3：
Figure PCTCN2022078111-appb-000002
其中，x为独立群体属性中任一个属性的概率因子，S(x)为该属性的概率。
例如，独立群体属性1的概率因子为3，独立群体属性2的概率因子为2，独立群体属性3的概率因子为4，得到的独立群体属性1的概率为95.2574％，独立群体属性2的概率为88.0797％，独立群体属性3的概率为98.2014％。
步骤1002、客户服务装置根据目标客户的不同群体属性的概率，确定目标客户的群体属性。
一种可能的实现方式中，在不同群体属性之间包括互斥群体属性的情况下，客户服务装置确定目标客户的群体属性包括第一互斥群体属性。
其中，第一互斥群体属性为互斥群体属性中概率最大的群体属性。
需要说明的是，互斥群体属性之间的概率相互影响。互斥群体属性中的其中一个群体属性的概率值越大，互斥群体属性中的其他群体属性的概率值就越小。互斥群体属性的概率之和为100％。
结合上述步骤1001中的示例，互斥群体属性1的概率为5.6612％，互斥群体属性2的概率为25.3716％，互斥群体属性3的概率为68.9672％。客户服务装置确定目标客户的群体属性包括互斥群体属性3。
又一种可能的实现方式中，在不同群体属性之间包括独立群体属性的情况下，客户服务装置确定目标客户的群体属性包括第一独立群体属性。
其中，第一独立群体属性为独立群体属性中概率大于属性概率阈值的群体属性。
需要说明的是，独立群体属性的概率只与该独立群体属性相关，与其他独立群体属性无关。目标客户可以同时具有独立群体属性中的一个或多个群体属性。
结合上述步骤1001中的示例，属性概率阈值设置为90％，独立群体属性1的概率为95.2574％，独立群体属性2的概率为88.0797％，独立群体属性3的概率为98.2014％。客户服务装置确定目标客户的群体属性既包括独立群体属性1又包括独立群体属性3。
基于上述技术方案，本公开中的客户服务装置能够将整体图像输入群体属性预测模型中，确定不同群体属性的概率，从而根据不同群体属性的概率确定目标客户的群体属性，以便于后续为目标客户分配工作人员。此外，由于不同群体属性之间包括互斥群体属性和独立群体属性中的至少一项。其中，互斥群体属性之间的概率相互影响，独立群体属性的概率只与该独立群体属性相关，因此，本公开中客户服务装置基于群体属性通过不同的方式确定对应的群体属性的概率，提高了客户服务装置确定目标客户的群体属性的准确度，从而更好地为目标客户分配工作人员。
以下，结合上述步骤403，对客户服务装置根据局部图像确定目标客户的个体属性的过程进行具体介绍。
作为本公开的一种可能的实施例，结合图4，如图11所示，上述步骤403还可以通过以下步骤1101-步骤1103实现：
步骤1101、客户服务装置将局部图像输入个体属性预测模型中，确定目标客户的每个性别对应的性别概率与每个年龄对应的年龄概率。
其中，个体属性预测模型用于确定目标客户的每个性别对应的性别概率与每个年龄对应的年龄概率。
示例性的，个体属性预测模型可以为神经网络模型，例如ShuffleNet算法模型、残差网络(ResNet)算法模型。
由于性别属性同样为互斥属性，因此，与上述互斥群体属性类似，客户服务装置可以通过Softmax函数确定目标客户的性别，具体实现方式可参考上述相关内容，此处不再赘述。
对于年龄属性，由于其所需要确定的年龄概率较多，本公开中客户服务装置可以通过L1norm对个体属性预测模型中提取的年龄特征进行回归，从而得到目标客户的每个年龄对应的年龄概率。
步骤1102、客户服务装置确定目标客户的性别为性别概率最大的性别。
示例性的，目标客户的性别为男的概率为40％，目标客户的性别为女的概率为60％，客户服务装置确定目标客户的性别为女。
步骤1103、客户服务装置确定目标客户的年龄为年龄概率最大的年龄。
客户服务装置确定目标客户的年龄的方式与上述确定性别的方式类似，此处不再赘述。
基于上述技术方案，本公开中的客户服务装置能够将局部图像输入个体属性预测模型中，确定目标客户的个体属性，例如目标客户的性别、年龄等。同时，针对不同的个体属性，本公开中的客户服务装置采用不同的方式确定对应个体属性的概率，并基于属性概率确定目标客户的个体属性，提高了客户服务装置确定目标客户的个体属性的准确度，以便于后续为目标客户分配工作人员。
以下，结合上述步骤403，对客户服务装置根据局部图像确定目标客户的个体属性的过程进行具体介绍。
作为本公开的又一种可能的实施例，结合图4，如图12所示，上述步骤403还包括以下步骤1201-步骤1202：
步骤1201、客户服务装置根据目标客户的局部图像检测用户数据库中是 否存在目标客户。
其中，用户数据库中预先存储有用户的用户信息。例如用户的局部图像、客户级别、性别、年龄、历史业务等信息。
客户服务装置可以将确定的目标客户的局部图像与用户数据库中每个用户的局部图像相匹配，从而检测用户数据库中是否存在目标客户。
一种可能的实现方式中，客户服务装置将目标客户的局部图像与用户数据库中的每个用户的局部图像输入局部图像识别模型，得到局部图像的局部图像特征。
客户服务装置根据目标客户的局部图像特征与用户数据库中每个用户的局部图像特征确定目标客户与每个用户的图像相似度。
在第一图像相似度大于预设相似度阈值的情况下，客户服务装置确定目标客户为第一图像相似度对应的用户。
又一种可能的实现方式中，用户数据库中可以预先存储每个用户的局部图像的图像特征。
如此一来，客户服务装置可以直接从用户数据库中获取每个用户的局部图像的图像特征，提高了检测效率。
示例性的，客户服务装置可以通过计算目标客户的局部图像特征与用户数据库中每个用户的局部图像特征的余弦距离确定目标客户与每个用户的图像相似度。
步骤1202、在用户数据库中存在目标客户的情况下，客户服务装置获取目标客户的客户级别。
示例性的，客户级别可以按照多个等级划分，例如1级客户、2级客户、3级客户。客户级别也可以按照客户类别划分，例如VIP客户，非VIP客户。本公开对此不作限定。
基于上述技术方案，本公开中的客户服务装置能够根据目标客户的局部图像从用户数据库检测对应的用户，从而获取目标客户更详细的用户信息以使得后续客户服务装置能更好地为目标客户分配工作人员。
作为本公开的又一种可能的实施例，结合图4，如图12所示，上述方法还包括以下步骤1203-步骤1206：
步骤1203、在用户数据库中存在目标客户的情况下，客户服务装置获取目标客户的基本信息。
示例性的，基本信息可以为性别、年龄、联系方式以及服务相关信息。其中，服务相关信息根据目标区域所对应的场景确定。例如目标区域为银行 场所时，服务相关信息可以为金融领域业务。
步骤1204、客户服务装置向服务终端发送目标客户的基本信息。相应的，服务终端接收客户服务装置发送的目标客户的基本信息。
其中，服务终端为工作人员使用的终端。
客户服务装置通过向分配的工作人员发送目标客户的基本信息，有助于工作人员及时了解目标客户的需求，以便于更好地为目标客户提供服务。
需要说明的是，当分配的工作人员为向目标客户提供服务的设备时，服务终端可以耦合在该设备中。
步骤1205、在目标客户授权允许查询历史业务的情况下，客户服务装置获取目标客户的历史业务。
一种可能的实现方式中，客户服务装置可以向目标客户的用户终端发送业务查询请求。
其中，业务查询请求用于确认目标客户是否授权允许查询历史业务。
在客户服务装置接收到用户终端发送的授权消息的情况下，确认目标客户授权允许查询历史业务。
又一种可能的实现方式中，用户数据库中可以预存用户的授权信息。客户服务装置从授权信息中确认目标客户是否授权允许查询历史业务。
步骤1206、客户服务装置向服务终端发送目标客户的历史业务。
基于上述技术方案，本公开中的客户服务装置能够从用户数据库中进一步获取目标客户的基本信息以及在目标客户授权的情况下获取目标客户的历史业务，并发送至工作人员的服务终端。该方案既保护了目标客户的隐私数据，同时也有助于为目标用户提供更好地服务。
以下，结合上述步骤404，对客户服务装置确定目标客户的服务优先级以及与目标客户对应的工作人员类型的过程进行具体介绍。
作为本公开的一种可能的实施例，结合图4，如图13所示，上述步骤404还包括以下步骤1301-步骤1305：
步骤1301、客户服务装置根据群体属性与个体属性确定目标客户的属性因子。
其中，属性因子用于表示目标客户的属性对服务优先级的影响程度。
需要说明的是，目标客户的属性因子可以包括一个，也可以包括多个。
示例性的，当目标客户的属性因子包括一个时，属性因子满足以下公式4：
Figure PCTCN2022078111-appb-000003
其中，S 1为目标客户的属性因子，w 1为属性因子权重值，
Figure PCTCN2022078111-appb-000004
为目标客户 的属性评估值。
当目标客户的属性因子为多个时，属性因子为多个子属性因子之和。多个子属性因子中每个子属性因子对应的权重值可以为相同取值，也可以为不同取值。子属性因子的确定方式与上述当目标客户的属性因子包括一个时，属性因子的确定方式相同，此处不再赘述。
一种可能的实现方式中，属性因子包括客户级别因子、群体属性因子、年龄因子中的至少一项。也即属性因子的取值为客户级别因子、群体属性因子、年龄因子中的至少一项之和。
下述以属性因子分别包括客户级别因子、群体属性因子、年龄因子中的任一项为例，对本公开进行具体论述。
当属性因子包括客户级别因子时，客户级别因子由目标客户的客户级别评估值与客户级别权重值确定。
示例性的，当目标客户的客户级别按照客户类别划分时，例如目标客户的客户级别为VIP客户时，目标客户的客户级别评估值为1。目标客户的客户级别为非VIP客户时，目标客户的客户级别评估值为0。
当属性因子包括群体属性因子时，群体属性因子由目标客户的群体属性评估值与群体属性权重值确定。
需要说明的是，目标客户的群体属性可以为一个，也可以为多个。当目标客户的群体属性为一个时，实现方式与客户级别因子的确定方式相同，此处不再赘述。
当目标客户的群体属性为多个时，可以分别确定每个群体属性的第一群体属性因子，并将每个群体属性的第一群体属性因子之和作为目标客户的群体属性因子。
其中，每个群体属性对应的群体属性权重值可以为相同取值，也可以为不同取值。客户服务装置确定每个群体属性的第一群体属性因子的方式与上述当目标客户的群体属性为一个时的确定方式相同，此处同样不再赘述。
示例性的，当目标客户的群体属性包括孕妇时，客户服务装置确定目标客户的群体属性因子为孕妇对应的群体属性因子。当目标客户的群体属性包括孕妇、残疾人士时，客户服务装置确定目标客户的群体属性因子为孕妇对应的第一群体属性因子与残疾人士对应的第一群体属性因子之和。
需要说明的是，上述仅以群体属性包括孕妇、残疾人士为例对本公开确定群体属性因子的实现方式进行说明，本公开中的群体属性还可以包括其他属性，具体可根据实际情况设置。
当属性因子包括年龄因子时，年龄因子由目标客户的年龄评估值与年龄权重值确定。
示例性的，目标客户的年龄评估值可以通过分段函数确定。例如图14所示的年龄评估值与年龄的函数图。目标客户的年龄评估值也可以通过曲线函数确定。本公开对此不作限定。
步骤1302、客户服务装置获取目标客户的队列信息，并根据队列信息确定队列因子。
其中，队列因子用于表示目标客户的队列信息对服务优先级的影响程度。
一种可能的实现方式中，目标客户可以通过目标区域中设置的用户登记设备进行预约业务登记，客户服务装置从用户登记设备中获取目标客户的队列信息。目标客户也可以通过用户终端远程预约业务，客户服务装置接收用户终端发送的预约业务信息，并从中确定目标客户的队列信息。
队列因子由目标客户的队列信息评估值与队列信息权重值确定。
示例性的，队列因子满足以下公式5：
Figure PCTCN2022078111-appb-000005
其中，S 2为目标客户的队列因子，w 2为队列信息权重值，
Figure PCTCN2022078111-appb-000006
为目标客户的队列信息评估值。
一种可能的实现方式中，客户服务装置可以根据目标客户的队列信息实时更新队列信息评估值，从而确定目标客户的队列因子。
示例性的，队列信息评估值满足以下公式6：
Figure PCTCN2022078111-appb-000007
其中，
Figure PCTCN2022078111-appb-000008
为目标客户的队列信息评估值，N为目标客户的队列信息中的排队次序。
也即是说，目标客户的排队次序越小，即当前排队越靠前，目标客户的队列信息评估值越大。
又一种可能的实现方式中，客户服务装置还可以根据目标客户的队列信息确定目标客户的预计等待时间，并向目标客户的用户终端发送该预计等待时间，以便于目标客户及时了解当前状况。
步骤1303、客户服务装置获取目标客户的预约业务，并根据预约业务确定业务因子。
其中，业务因子用于表示目标客户的预约业务对服务优先级的影响程度。
一种可能的实现方式中，客户服务装置可以从用户登记设备或者目标客 户的用户终端中获取目标客户的预约业务，具体可参考步骤1302中的相关描述，此处不再赘述。
业务因子由目标客户的预约业务评估值与预约业务权重值确定。
示例性的，业务因子满足以下公式7：
Figure PCTCN2022078111-appb-000009
其中，S 3为目标客户的业务因子，w 3为预约业务权重值，
Figure PCTCN2022078111-appb-000010
为目标客户的预约业务评估值。
预约业务评估值与预约业务的等级相关。预约业务的等级越高，预约业务评估值也就越大，相反的，预约业务的等级越低，预约业务评估值也就越小。
步骤1304、客户服务装置根据属性因子、队列因子以及业务因子确定目标客户的服务优先级。
示例性的，服务优先级满足以下公式8：
S＝S 1+S 2+S 3                   公式8
其中，S为目标客户的服务优先级，S 1为目标客户的属性因子，S 2为目标客户的队列因子，S 3为目标客户的业务因子。
步骤1305、客户服务装置根据目标客户的群体属性和目标客户的个体属性，确定与目标客户对应的工作人员类型。
该步骤的具体实现方式可参考上述步骤404-步骤405中的相关描述，此处不再赘述。
基于上述技术方案，本公开中客户服务装置能够分别确定目标客户的属性因子、队列因子以及业务因子，并根据属性因子、队列因子以及业务因子确定目标客户的服务优先级，从而按照服务优先级的优先次序为目标客户分配相应的工作人员。如此一来，客户服务装置可以从目标客户的属性、当前的排队状况以及待办理的业务类型三方面为目标客户分配顺序，使得客户服务装置分配工作人员的过程更加合理，保障了不同目标客户的服务体验。
以下，对客户服务装置确定目标客户的路线轨迹并确定目标客户的关注业务的过程进行具体介绍。
作为本公开的一种可能的实施例，结合图4，如图15所示，上述方法还包括以下步骤1501-步骤1502：
步骤1501、客户服务装置在多个图像采集装置采集的图像中，确定目标客户在不同时刻的多个整体图像。
需要说明的是，由于单个图像采集装置往往无法采集目标区域中每个区 域的图像信息，因此客户服务装置从多个图像采集装置所采集的图像中检测目标客户在不同时刻的多个整体图像，有利于后续更加准确、全面地确定目标客户的路线轨迹。
其中，对于同一图像采集装置所采集的不同时刻的图像信息，客户服务装置可以通过步骤401中的方式确定目标客户的整体图像，此处不再赘述。
对于不同采集装置所采集的图像信息，首先，客户服务装置可以从其中任一个图像采集装置采集的图像信息中确定目标客户的整体图像。之后，客户服务装置可以将该整体图像输入行人重识别模型中，确定该目标客户在多个图像采集装置采集的图像中的整体图像。
示例性的，行人重识别模型可以为多粒度网络(multiple granularity network，MGN)模型。行人重识别模型也可以为用于进行行人检测的模型，本公开对此不作限定。
需要说明的是，对于客户服务装置通过上述方式确定的目标客户在不同时刻确定的多个整体图像，可能会存在一定偏差。因此，客户服务装置还可以根据所检测的目标客户的整体图像，对整体图像的位置进行校正。
一种可能的实现方式中，客户服务装置根据当前时刻目标客户的整体图像的位置信息对下一时刻检测到的目标客户的整体图像的位置信息进行校正，以获取预测的下一时刻的目标客户的整体图像。
示例性的，客户服务装置可以通过SORT算法对目标客户的整体图像进行跟踪。客户服务装置对检测的目标客户的整体图像进行卡尔曼(kalman)滤波，预估目标客户的运动状态，之后通过匈牙利匹配算法进行位置匹配，从而得到校正后的目标客户在不同时刻的多个整体图像。
步骤1502、客户服务装置根据多个整体图像的位置信息，确定目标客户的路线轨迹。
示例性的，如图16所示，客户服务装置可以将多个整体图像映射在目标区域对应的坐标系中，得到每个整体图像对应的坐标。客户服务装置将每个坐标按照时间顺序相连接，得到目标客户的路线轨迹。
一种可能的实现方式中，客户服务装置还可以确定目标物品的图像，通过检测目标物品与目标客户的路线轨迹，确定目标客户是否遗失物品。
示例性的，当确定的目标物品与目标客户的距离大于预设阈值，且持续时长超过一定时长时，客户服务装置检测目标物品与目标客户的路线轨迹的关系。在目标物品与目标客户的距离小于预设阈值且持续时长超过一定时长时，确定该目标物品属于目标客户且该目标客户发生物品遗失。
基于上述技术方案，本公开中客户服务装置能够从多个图像采集设备中获取目标客户在不同时刻下的多个整体图像，从而根据多个整体图像对应的位置信息确定目标客户的路线轨迹，以便于后续根据目标客户的路线轨迹分析目标客户所关注的相关业务，从而为目标客户提供相关业务推荐。同时，客户服务装置还可以根据整体图像的位置信息引导工作人员及时找到待服务的目标客户，从而更快地提供服务。
作为本公开的一种可能的实施例，结合图4，如图15所示，在步骤1502之后，上述方法还包括以下步骤1503-步骤1505：
步骤1503、客户服务装置根据目标客户的路线轨迹确定目标客户在至少一个特定区域内的驻留时长。
其中，至少一个特定区域对应至少一个业务。特定区域可以根据实际情况设置，本公开对此不作限定。
需要说明的是，如图16所示，客户服务装置可以将目标客户位于特定区域内的路线轨迹对应的时间段作为目标客户在该特定区域内的驻留时长。由于每个特定区域对应一个业务，因此，目标客户在特定区域内的驻留时长可以表征目标客户对该特定区域对应业务的关注程度。
步骤1504、客户服务装置确定目标客户的关注业务为驻留时长最长的特定区域对应的业务。
由上述可知，目标客户在特定区域内的驻留时长可以表征目标客户对该特定区域对应业务的关注程度。因此，客户服务装置可以将驻留时长最长的特定区域对应的业务作为目标客户的关注业务，并获取该业务的业务信息。
步骤1505、客户服务装置向服务终端发送关注业务的业务信息。
基于上述技术方案，本公开中客户服务装置能够根据目标客户的路线轨迹分析目标客户所关注的业务，并向所分配的工作人员的服务终端发送该业务的业务信息，以便于工作人员针对性地为目标客户提供服务，提高了目标客户的服务体验。
本公开实施例可以根据上述方法示例对客户服务装置进行功能模块或者功能单元的划分，例如，可以对应各个功能划分各个功能模块或者功能单元，也可以将两个或两个以上的功能集成在一个处理模块中。上述集成的模块既可以采用硬件的形式实现，也可以采用软件功能模块或者功能单元的形式实现。其中，本公开实施例中对模块或者单元的划分是示意性的，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式。
如图17所示，为根据一些实施例提供的一种客户服务装置170的结构示 意图，该装置包括：
处理单元1701，被配置为确定目标客户的整体图像与局部图像。
处理单元1701，还被配置为根据整体图像确定目标客户的群体属性。
其中，群体属性用于表征目标客户所属群体的类型。
处理单元1701，还被配置为根据局部图像确定目标客户的个体属性。
其中，个体属性包括性别、年龄以及客户级别中的至少一个。
处理单元1701，还被配置为根据目标客户的群体属性和目标客户的个体属性，确定目标客户的服务优先级以及与目标客户对应的工作人员类型。
处理单元1701，还被配置为根据目标客户的服务优先级以及与目标客户对应的工作人员类型，为目标客户分配工作人员。
在一些实施例中，通信单元1702，被配置为向服务终端发送目标客户的客户信息和目标客户的当前位置信息；处理单元1701，被配置为根据工作人员的整体图像与目标客户的整体图像确定工作人员与目标客户的第一位置距离；通信单元1702，还被配置为在第一位置距离大于第一预设距离且持续时长大于第一时长的情况下，向服务终端发送服务提示信息。
在一些实施例中，通信单元1702，被配置为接收目标客户的注册信息；注册信息包括目标客户的局部图像以及目标客户的陪同人员的局部图像。
在一些实施例中，处理单元1701，被配置为根据陪同人员的局部图像确定陪同人员的整体图像；处理单元1701，被配置为根据目标客户的整体图像与陪同人员的整体图像确定目标客户与陪同人员的第二位置距离；处理单元1701，被配置为在第二位置距离大于第二预设距离的情况下，根据目标客户的服务优先级以及与目标客户对应的工作人员类型，为目标客户分配工作人员。
在一些实施例中，处理单元1701，被配置为确定目标工作人员群组，目标工作人员群组的类型为目标客户对应的工作人员类型，目标工作人员群组包括多个目标工作人员；处理单元1701，还被配置为根据目标工作人员群组的人员属性，确定目标工作人员群组的人员优先级；处理单元1701，还被配置为根据目标客户的服务优先级以及目标工作人员群组的人员优先级，为目标客户分配目标工作人员。
在一些实施例中，通信单元1702，被配置为获取待检测图像；待检测图像包括目标客户；处理单元1701，被配置为确定待检测图像中的第一图像区域，以及第一图像区域对应的第一数值和第二数值；第一数值表示第一图像区域包括整体图像的概率；第二数值表示第一图像区域包括局部图像的概率； 第一图像区域为待检测图像中的任一个区域；处理单元1701，被配置为在第一数值大于第一阈值的情况下，确定第一图像区域包括目标客户的整体图像；处理单元1701，被配置为在第二数值大于第二阈值的情况下，确定第一图像区域包括目标客户的局部图像。
在一些实施例中，处理单元1701，被配置为将整体图像输入群体属性预测模型中，确定目标客户的不同群体属性的概率；处理单元1701，被配置为根据目标客户的不同群体属性的概率，确定目标客户的群体属性。
在一些实施例中，不同群体属性之间包括互斥群体属性和独立群体属性中的至少一项；处理单元1701，被配置为在不同群体属性之间包括互斥群体属性的情况下，确定目标客户的群体属性包括第一互斥群体属性；第一互斥群体属性为互斥群体属性中概率最大的群体属性；处理单元1701，被配置为在不同群体属性之间包括独立群体属性的情况下，确定目标客户的群体属性包括第一独立群体属性；第一独立群体属性为独立群体属性中概率大于属性概率阈值的群体属性。
在一些实施例中，处理单元1701，被配置为将局部图像输入个体属性预测模型中，确定目标客户的每个性别对应的性别概率与每个年龄对应的年龄概率；处理单元1701，被配置为确定目标客户的性别为性别概率最大的性别；处理单元1701，被配置为确定目标客户的年龄为年龄概率最大的年龄。
在一些实施例中，处理单元1701，被配置为根据目标客户的局部图像检测用户数据库中是否存在目标客户；通信单元1702，被配置为在用户数据库中存在目标客户的情况下，获取目标客户的客户级别。
在一些实施例中，通信单元1702，被配置为在用户数据库中存在目标客户的情况下，获取目标客户的基本信息；通信单元1702，被配置为向服务终端发送目标客户的基本信息；服务终端为工作人员使用的终端；通信单元1702，被配置为在目标客户授权允许查询历史业务的情况下，获取目标客户的历史业务；通信单元1702，被配置为向服务终端发送目标客户的历史业务。
在一些实施例中，处理单元1701，被配置为根据群体属性与个体属性确定目标客户的属性因子；属性因子用于表示目标客户的属性对服务优先级的影响程度；处理单元1701，被配置为获取目标客户的队列信息，并根据队列信息确定队列因子；队列因子用于表示目标客户的队列信息对服务优先级的影响程度；处理单元1701，被配置为获取目标客户的预约业务，并根据预约业务确定业务因子；业务因子用于表示目标客户的预约业务对服务优先级的影响程度；处理单元1701，被配置为根据属性因子、队列因子以及业务因子 确定目标客户的服务优先级。
在一些实施例中，处理单元1701，被配置为在多个图像采集装置采集的图像中，确定目标客户在不同时刻的多个整体图像；处理单元1701，被配置为根据多个整体图像的位置信息，确定目标客户的路线轨迹。
在一些实施例中，处理单元1701，被配置为根据目标客户的路线轨迹确定目标客户在至少一个特定区域内的驻留时长；至少一个特定区域对应至少一个业务；处理单元1701，被配置为确定目标客户的关注业务为驻留时长最长的特定区域对应的业务；通信单元1702，被配置为向服务终端发送关注业务的业务信息。
在通过硬件实现时，本公开实施例中的通信单元1702可以集成在通信接口上，处理单元1701可以集成在处理器上。具体实现方式如图18所示。
图18示出了上述实施例中所涉及的客户服务装置的又一种可能的结构示意图。该客户服务装置180包括：处理器1802和通信接口1803。处理器1802被配置为对客户服务装置180的动作进行控制管理，例如，执行上述处理单元1701执行的步骤，和/或被配置为执行本文所描述的技术的其它过程。通信接口1803被配置为支持客户服务装置180与其他网络实体的通信，例如，执行上述通信单元1702执行的步骤。客户服务装置180还可以包括存储器1801和总线1804，存储器1801被配置为存储客户服务装置180的程序代码和数据。
其中，存储器1801可以是客户服务装置180中的存储器等，该存储器可以包括易失性存储器，例如随机存取存储器；该存储器也可以包括非易失性存储器，例如只读存储器，快闪存储器，硬盘或固态硬盘；该存储器还可以包括上述种类的存储器的组合。
上述处理器1802可以是实现或执行结合本公开公开内容所描述的各种示例性的逻辑方框，模块和电路。该处理器可以是中央处理器，通用处理器，数字信号处理器，专用集成电路，现场可编程门阵列或者其他可编程逻辑器件、晶体管逻辑器件、硬件部件或者其任意组合。其可以实现或执行结合本公开公开内容所描述的各种示例性的逻辑方框，模块和电路。该处理器也可以是实现计算功能的组合，例如包含一个或多个微处理器组合，DSP和微处理器的组合等。
总线1804可以是扩展工业标准结构(Extended Industry Standard Architecture，EISA)总线等。总线1804可以分为地址总线、数据总线、控制总线等。为便于表示，图18中仅用一条粗线表示，但并不表示仅有一根总线或一种类型的总线。
图18中的客户服务装置180还可以为芯片。该芯片包括一个或两个以上(包括两个)处理器1802和通信接口1803。
在一些实施例中，该芯片还包括存储器1801，存储器1801可以包括只读存储器和随机存取存储器，并向处理器1802提供操作指令和数据。存储器1801的一部分还可以包括非易失性随机存取存储器(non-volatile random access memory，NVRAM)。
在一些实施方式中，存储器1801存储了如下的元素，执行模块或者数据结构，或者他们的子集，或者他们的扩展集。
在本公开实施例中，通过调用存储器1801存储的操作指令(该操作指令可存储在操作系统中)，执行相应的操作。
通过以上的实施方式的描述，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，仅以上述各功能模块的划分进行举例说明，实际应用中，可以根据需要而将上述功能分配由不同的功能模块完成，即将装置的内部结构划分成不同的功能模块，以完成以上描述的全部或者部分功能。上述描述的系统，装置和单元的具体工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。
本公开的一些实施例提供了一种计算机可读存储介质(例如，非暂态计算机可读存储介质)，该计算机可读存储介质中存储有计算机程序指令，计算机程序指令在计算机(例如，客户服务装置)上运行时，使得计算机执行如上述实施例中任一实施例所述的客户服务方法。
示例性的，上述计算机可读存储介质可以包括，但不限于：磁存储器件(例如，硬盘、软盘或磁带等)，光盘(例如，CD(Compact Disk，压缩盘)、DVD(Digital Versatile Disk，数字通用盘)等)，智能卡和闪存器件(例如，EPROM(Erasable Programmable Read-Only Memory，可擦写可编程只读存储器)、卡、棒或钥匙驱动器等)。本公开描述的各种计算机可读存储介质可代表用于存储信息的一个或多个设备和/或其它机器可读存储介质。术语“机器可读存储介质”可包括但不限于，无线信道和能够存储、包含和/或承载指令和/或数据的各种其它介质。
本公开的一些实施例还提供了一种计算机程序产品，例如该计算机程序产品存储在非瞬时性的计算机可读存储介质上。该计算机程序产品包括计算机程序指令，在计算机(例如，客户服务装置)上执行该计算机程序指令时，该计算机程序指令使计算机执行如上述实施例所述的客户服务方法。
本公开的一些实施例还提供了一种计算机程序。当该计算机程序在计算 机(例如，客户服务装置)上执行时，该计算机程序使计算机执行如上述实施例所述的客户服务方法。
上述计算机可读存储介质、计算机程序产品及计算机程序的有益效果和上述一些实施例所述的客户服务方法的有益效果相同，此处不再赘述。
在本公开所提供的几个实施例中，应该理解到，所揭露的系统、设备和方法，可以通过其它的方式实现。例如，以上所描述的设备实施例仅仅是示意性的，例如，所述单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。另一点，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口，设备或单元的间接耦合或通信连接，可以是电性，机械或其它的形式。
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施例方案的目的。
另外，在本公开各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。
以上所述，仅为本公开的具体实施方式，但本公开的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本公开揭露的技术范围内，想到变化或替换，都应涵盖在本公开的保护范围之内。因此，本公开的保护范围应以所述权利要求的保护范围为准。

Claims (19)

  1. 一种客户服务方法，包括：
    确定目标客户的整体图像与局部图像；
    根据所述整体图像确定所述目标客户的群体属性；所述群体属性用于表征所述目标客户所属群体的类型；
    根据所述局部图像确定所述目标客户的个体属性；所述个体属性包括性别、年龄以及客户级别中的至少一个；
    根据所述目标客户的群体属性和所述目标客户的个体属性，确定所述目标客户的服务优先级以及与所述目标客户对应的工作人员类型；
    根据所述目标客户的服务优先级以及与所述目标客户对应的工作人员类型，为所述目标客户分配工作人员。
  2. 根据权利要求1所述的方法，还包括：
    向服务终端发送所述目标客户的客户信息和所述目标客户的当前位置信息；所述服务终端为所述工作人员使用的终端；
    根据所述工作人员的整体图像与所述目标客户的整体图像确定所述工作人员与所述目标客户的第一位置距离；
    在所述第一位置距离大于第一预设距离且持续时长大于第一时长的情况下，向所述服务终端发送服务提示信息。
  3. 根据权利要求1或2所述的方法，还包括：
    接收所述目标客户的注册信息；所述注册信息包括所述目标客户的局部图像以及所述目标客户的陪同人员的局部图像。
  4. 根据权利要求3所述的方法，其中，所述根据所述目标客户的服务优先级以及与所述目标客户对应的工作人员类型，为所述目标客户分配工作人员，包括：
    根据所述陪同人员的局部图像确定所述陪同人员的整体图像；
    根据所述目标客户的整体图像与所述陪同人员的整体图像确定所述目标客户与所述陪同人员的第二位置距离；
    在所述第二位置距离大于第二预设距离的情况下，根据所述目标客户的服务优先级以及与所述目标客户对应的工作人员类型，为所述目标客户分配工作人员。
  5. 根据权利要求1-4任一项所述的方法，其中，所述根据所述目标客户的服务优先级以及与所述目标客户对应的工作人员类型，为所述目标客户分配工作人员，包括：
    确定目标工作人员群组，所述目标工作人员群组的类型为所述目标客户 对应的工作人员类型，所述目标工作人员群组包括多个目标工作人员；
    根据所述目标工作人员群组的人员属性，确定所述目标工作人员群组的人员优先级；
    根据所述目标客户的服务优先级以及所述目标工作人员群组的人员优先级，为所述目标客户分配所述目标工作人员。
  6. 根据权利要求1-5任一项所述的方法，其中，所述确定目标客户的整体图像与局部图像，包括：
    获取待检测图像；所述待检测图像包括所述目标客户；
    确定所述待检测图像中的第一图像区域，以及所述第一图像区域对应的第一数值和第二数值；所述第一数值表示所述第一图像区域包括所述整体图像的概率；所述第二数值表示所述第一图像区域包括所述局部图像的概率；所述第一图像区域为所述待检测图像中的任一个区域；
    在所述第一数值大于第一阈值的情况下，确定所述第一图像区域包括所述目标客户的整体图像；
    在所述第二数值大于第二阈值的情况下，确定所述第一图像区域包括所述目标客户的局部图像。
  7. 根据权利要求1-6任一项所述的方法，其中，所述根据所述整体图像确定所述目标客户的群体属性，包括：
    将所述整体图像输入群体属性预测模型中，确定所述目标客户的不同群体属性的概率；
    根据所述目标客户的不同群体属性的概率，确定所述目标客户的群体属性。
  8. 根据权利要求7所述的方法，其中，所述不同群体属性之间包括互斥群体属性和独立群体属性中的至少一项；所述根据所述目标客户的不同群体属性的概率，确定所述目标客户的群体属性，包括：
    在所述不同群体属性之间包括互斥群体属性的情况下，确定所述目标客户的群体属性包括第一互斥群体属性；所述第一互斥群体属性为所述互斥群体属性中概率最大的群体属性；
    在所述不同群体属性之间包括独立群体属性的情况下，确定所述目标客户的群体属性包括第一独立群体属性；所述第一独立群体属性为所述独立群体属性中概率大于属性概率阈值的群体属性。
  9. 根据权利要求1-8任一项所述的方法，其中，所述根据所述局部图像确定所述目标客户的个体属性，包括：
    将所述局部图像输入个体属性预测模型中，确定所述目标客户的每个性别对应的性别概率与每个年龄对应的年龄概率；
    确定所述目标客户的性别为性别概率最大的性别；
    确定所述目标客户的年龄为年龄概率最大的年龄。
  10. 根据权利要求1-9任一项所述的方法，其中，所述根据所述局部图像确定所述目标客户的个体属性，包括：
    根据所述目标客户的局部图像检测用户数据库中是否存在所述目标客户；
    在所述用户数据库中存在所述目标客户的情况下，获取所述目标客户的客户级别。
  11. 根据权利要求10所述的方法，还包括：
    在所述用户数据库中存在所述目标客户的情况下，获取所述目标客户的基本信息；
    向服务终端发送所述目标客户的基本信息；所述服务终端为所述工作人员使用的终端；
    在所述目标客户授权允许查询历史业务的情况下，获取所述目标客户的历史业务；
    向所述服务终端发送所述目标客户的历史业务。
  12. 根据权利要求1-11任一项所述的方法，其中，所述根据所述目标客户的群体属性和所述目标客户的个体属性，确定所述目标客户的服务优先级，包括：
    根据所述群体属性与所述个体属性确定所述目标客户的属性因子；所述属性因子用于表示所述目标客户的属性对所述服务优先级的影响程度；
    获取所述目标客户的队列信息，并根据所述队列信息确定队列因子；所述队列因子用于表示所述目标客户的队列信息对所述服务优先级的影响程度；
    获取所述目标客户的预约业务，并根据所述预约业务确定业务因子；所述业务因子用于表示所述目标客户的预约业务对所述服务优先级的影响程度；
    根据所述属性因子、所述队列因子以及所述业务因子确定所述目标客户的服务优先级。
  13. 根据权利要求1-12任一项所述的方法，还包括：
    在多个图像采集装置采集的图像中，确定所述目标客户在不同时刻的多 个整体图像；
    根据所述多个整体图像的位置信息，确定所述目标客户的路线轨迹。
  14. 根据权利要求13所述的方法，还包括：
    根据所述目标客户的路线轨迹确定所述目标客户在至少一个特定区域内的驻留时长；所述至少一个特定区域对应至少一个业务；
    确定所述目标客户的关注业务为所述驻留时长最长的特定区域对应的业务；
    向服务终端发送所述关注业务的业务信息；所述服务终端为所述工作人员使用的终端。
  15. 一种客户服务装置，包括：
    处理单元，被配置为确定目标客户的整体图像与局部图像；
    所述处理单元，还被配置为根据所述整体图像确定所述目标客户的群体属性；所述群体属性用于表征所述目标客户所属群体的类型；
    所述处理单元，还被配置为根据所述局部图像确定所述目标客户的个体属性；所述个体属性包括性别、年龄以及客户级别中的至少一个；
    所述处理单元，还被配置为根据所述目标客户的群体属性和所述目标客户的个体属性，确定所述目标客户的服务优先级以及与所述目标客户对应的工作人员类型；
    所述处理单元，还被配置为根据所述目标客户的服务优先级以及与所述目标客户对应的工作人员类型，为所述目标客户分配工作人员。
  16. 一种客户服务装置，包括：处理器和通信接口；所述通信接口和所述处理器耦合，所述处理器用于运行计算机程序或指令，以实现如权利要求1-14任一项中所述的客户服务方法。
  17. 一种客户服务系统，包括客户服务装置和至少一个图像采集装置，所述图像采集装置用于采集图像信息，所述客户服务装置用于执行如权利要求1-14任一项中所述的客户服务方法。
  18. 一种非暂态计算机可读存储介质，其中，所述非暂态计算机可读存储介质中存储有指令，当计算机执行所述指令时，所述计算机执行上述权利要求1-14任一项中所述的客户服务方法。
  19. 一种计算机程序产品，所述计算机程序产品包括计算机程序指令，在计算机上执行所述计算机程序指令时，所述计算机程序指令使计算机执行如权利要求1-14任一项中所述的客户服务方法。
PCT/CN2022/078111 2022-02-25 2022-02-25 客户服务方法、装置、系统及存储介质 WO2023159525A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN202280000320.XA CN116964610A (zh) 2022-02-25 2022-02-25 客户服务方法、装置、系统及存储介质
PCT/CN2022/078111 WO2023159525A1 (zh) 2022-02-25 2022-02-25 客户服务方法、装置、系统及存储介质

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2022/078111 WO2023159525A1 (zh) 2022-02-25 2022-02-25 客户服务方法、装置、系统及存储介质

Publications (1)

Publication Number Publication Date
WO2023159525A1 true WO2023159525A1 (zh) 2023-08-31

Family

ID=87764321

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/078111 WO2023159525A1 (zh) 2022-02-25 2022-02-25 客户服务方法、装置、系统及存储介质

Country Status (2)

Country Link
CN (1) CN116964610A (zh)
WO (1) WO2023159525A1 (zh)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101901461A (zh) * 2010-07-16 2010-12-01 北京天诚盛业科技有限公司 具有智能人脸识别的银行vip客户服务系统和方法
JP2017146652A (ja) * 2016-02-15 2017-08-24 日本電気株式会社 情報処理装置、情報処理システム、制御方法、及びプログラム
CN108764932A (zh) * 2018-04-09 2018-11-06 国网山东省电力公司 基于人脸识别的电力营业厅智能分级调度方法和系统
US20210342750A1 (en) * 2019-08-14 2021-11-04 Nec Corporation Information processing apparatus, information processing method, and storage medium

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101901461A (zh) * 2010-07-16 2010-12-01 北京天诚盛业科技有限公司 具有智能人脸识别的银行vip客户服务系统和方法
JP2017146652A (ja) * 2016-02-15 2017-08-24 日本電気株式会社 情報処理装置、情報処理システム、制御方法、及びプログラム
CN108764932A (zh) * 2018-04-09 2018-11-06 国网山东省电力公司 基于人脸识别的电力营业厅智能分级调度方法和系统
US20210342750A1 (en) * 2019-08-14 2021-11-04 Nec Corporation Information processing apparatus, information processing method, and storage medium

Also Published As

Publication number Publication date
CN116964610A (zh) 2023-10-27

Similar Documents

Publication Publication Date Title
US11868926B2 (en) Systems and methods for managing public place in smart city
US20180181834A1 (en) Method and apparatus for security inspection
JP2020504358A (ja) 画像ベースの車両損害評価方法、装置、およびシステム、ならびに電子デバイス
US11232327B2 (en) Smart video surveillance system using a neural network engine
WO2021043074A1 (zh) 基于图像识别的城市宠物活动轨迹监测方法及相关设备
US9147128B1 (en) Machine learning enhanced facial recognition
JP7303384B2 (ja) 乗降者数計数システム及び乗降者数計数装置
US20200401791A1 (en) Methods and Systems for Dynamic and Incremental Face Recognition
CN112183380B (zh) 基于人脸识别的客流量分析方法和系统、电子设备
CN113962326A (zh) 聚类方法、装置、设备及计算机存储介质
CN114022841A (zh) 人员监控识别方法、装置、电子设备及可读存储介质
WO2023159525A1 (zh) 客户服务方法、装置、系统及存储介质
JP6894395B2 (ja) 情報取得装置及び情報集計システム並びに情報集計装置
CN116453226A (zh) 基于人工智能的人体姿态识别方法、装置及相关设备
CN113537148B (zh) 人体动作识别方法及装置、可读存储介质、电子设备
CN114998782A (zh) 面审视频的场景分类方法、装置、电子设备及存储介质
CN115222109A (zh) 信息处理方法、装置、介质及设备
CN114882420A (zh) 接待人数统计方法、装置、电子设备及可读存储介质
CN113435975A (zh) 轮椅租赁处理方法、装置及相关设备
CN113792569A (zh) 对象识别方法、装置、电子设备及可读介质
CN111275035A (zh) 一种识别背景信息的方法及系统
JP7328401B2 (ja) 入出場マッチングシステムの概念
CN113360688B (zh) 信息库的构建方法、装置及系统
US11494942B2 (en) Information processing device, information processing method, program, recording medium, and camera system
JP6421087B2 (ja) 周辺状況推定支援装置、方法およびプログラム

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 202280000320.X

Country of ref document: CN

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

Ref document number: 22927808

Country of ref document: EP

Kind code of ref document: A1