WO2019061660A1

WO2019061660A1 - Recruitment method, electronic device, and readable storage medium

Info

Publication number: WO2019061660A1
Application number: PCT/CN2017/108764
Authority: WO
Inventors: 王健宗; 王晨羽; 马进; 肖京
Original assignee: 平安科技（深圳）有限公司
Priority date: 2017-09-30
Filing date: 2017-10-31
Publication date: 2019-04-04
Also published as: CN107784482A

Abstract

A recruitment method, an electronic device, and a readable storage medium. The method comprises: obtaining face photos of applicants; recognizing the face photos utilizing a pre-determined recognition model to obtain performance levels corresponding to the face photos, such that a recruiter recruits the applicants with reference to the obtained performance levels, the pre-determined recognition model being a depth convolution neural network model obtained in advance by training a preset number of face sample images marked with different performance levels. The method assists in recruitment by providing a new recruitment reference factor, such that the quality of recruitment is improved.

Description

Recruitment method, electronic device and readable storage medium

The present application is based on the priority of the Chinese Patent Application entitled "Recruitment Method, Electronic Device and Readable Storage Medium" filed on September 30, 2017, with the application number of CN 201710916039. The manner of reference is incorporated in the present application.

Technical field

The present application relates to the field of computer technologies, and in particular, to a recruitment method, an electronic device, and a readable storage medium.

Background technique

Most of the existing recruitment methods are traditional written tests, interviews, etc., and there is no “face to face” of candidates to explore potential talents and help recruit technical solutions.

Summary of the invention

The purpose of the present application is to provide a recruitment method, an electronic device, and a readable storage medium, which are intended to assist recruitment by identifying the face of a candidate and predicting corresponding performance.

In order to achieve the above object, a first aspect of the present application provides an electronic device, where the electronic device includes a memory and a processor, where the memory stores a recruitment system executable on the processor, and the recruitment system is The following steps are implemented when the processor is executed:

A1. Obtain a photo of the face of the applicant;

B1, the face photo is identified by using a predetermined recognition model, and the performance level corresponding to the face photo is obtained, so that the recruiter can recruit the candidate according to the obtained performance level; wherein the predetermined The recognition model is a deep convolutional neural network model obtained by training a preset number of face sample images marked with different performance levels in advance.

A second aspect of the present application provides a recruitment method, the recruitment method comprising:

Step 1: Obtain a photo of the face of the candidate;

Step 2: Identifying the face photo by using a predetermined recognition model, and obtaining a performance level corresponding to the face photo, for the recruiter to refer to the obtained performance level to recruit the candidate; wherein, the advance The determined recognition model is a deep convolutional neural network model obtained by training a preset number of face sample pictures marked with different performance levels in advance.

A third aspect of the present application provides a computer readable storage medium storing a recruiting system, the recruiting system being executable by at least one processor to cause the at least one processor to perform the following steps:

A2. Obtain a photo of the face of the applicant;

B2. Identifying the face photo by using a predetermined recognition model, and obtaining a performance level corresponding to the face photo, for the recruiter to refer to the obtained performance level to recruit the candidate; wherein the predetermined The recognition model is a deep convolutional neural network model obtained by training a preset number of face sample images marked with different performance levels in advance.

The recruitment method, system and readable storage medium proposed by the present application identify the face photos of the employed personnel by using a deep convolutional neural network model trained based on a preset number of face sample images marked with different performance levels. According to the recognition result, the performance level corresponding to the candidate is predicted. Through deep learning, find out the relationship between the face and the performance level, and determine the corresponding performance level according to the face of the candidate when recruiting, and can use the predicted performance level of the candidate as the reference factor for recruitment. To provide a new recruitment reference factor to assist recruitment and improve the quality of recruitment.

DRAWINGS

1 is a schematic diagram of an operating environment of a preferred embodiment of the recruitment system 10 of the present application;

FIG. 2 is a schematic flowchart of an embodiment of a recruitment method of the present application.

Detailed ways

In order to make the objects, technical solutions, and advantages of the present application more comprehensible, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It is understood that the specific embodiments described herein are merely illustrative of the application and are not intended to be limiting. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present application without departing from the inventive scope are the scope of the present application.

It should be noted that the descriptions of "first", "second" and the like in the present application are for the purpose of description only, and are not to be construed as indicating or implying their relative importance or implicitly indicating the number of technical features indicated. . Thus, features defining "first" and "second" may include at least one of the features, either explicitly or implicitly. In addition, the technical solutions between the various embodiments may be combined with each other, but must be based on the realization of those skilled in the art, and when the combination of the technical solutions is contradictory or impossible to implement, it should be considered that the combination of the technical solutions does not exist. Nor is it within the scope of protection required by this application.

The application provides a recruitment system. Please refer to FIG. 1 , which is a schematic diagram of an operating environment of a preferred embodiment of the recruitment system 10 of the present application.

In the present embodiment, the recruitment system 10 is installed and operated in the electronic device 1. The electronic device 1 may include, but is not limited to, a memory 11, a processor 12, and a display 13. Figure 1 shows only the electronic device 1 with components 11-13, but it should be understood that not all illustrated components may be implemented, and more or fewer components may be implemented instead.

The memory 11 comprises at least one type of readable storage medium, which in some embodiments may be an internal storage unit of the electronic device 1, such as a hard disk or memory of the electronic device 1. The memory 11 may also be an external storage device of the electronic device 1 in other embodiments, such as a plug-in hard disk equipped on the electronic device 1, a smart memory card (SMC), and a secure digital device. (Secure Digital, SD) card, flash card, etc. Further, the memory 11 may also include both an internal storage unit of the electronic device 1 and an external storage device. The memory 11 is configured to store application software installed on the electronic device 1 and various types of data, such as program codes of the recruiting system 10 and the like. The memory 11 can also be used to temporarily store data that has been output or is about to be output.

The processor 12, in some embodiments, may be a central processing unit (CPU), a microprocessor or other data processing chip for running program code or processing data stored in the memory 11, for example The recruitment system 10 and the like are executed.

The display 13 in some embodiments may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch sensor, or the like. The display 13 is used to display information processed in the electronic device 1 and a user interface for displaying visualization, such as a face photo of a candidate, a performance level corresponding to a face photo, and the like. The components 11-13 of the electronic device 1 communicate with one another via a system bus.

The recruiting system 10 includes at least one computer readable instructions stored in the memory 11, the at least one computer readable instructions being executable by the processor 12 to implement various embodiments of the present application.

Wherein, when the above recruitment system 10 is executed by the processor 12, the following steps are implemented:

In step S1, a photo of the face of the candidate is obtained.

In this embodiment, the recruiting system 10 receives a face photo of the applicant sent by the user, for example, receiving a photo of the face of the applicant sent by the user through a mobile phone, a tablet computer, a self-service terminal device, and the like, such as receiving the user on the mobile phone or tablet. The face image of the applicant sent by the pre-installed client in the terminal such as the self-service terminal device, or the face image of the candidate sent by the user on the browser system in the terminal such as the mobile phone, the tablet computer, the self-service terminal device, and the like . The face image of the applicant may be various image format types such as JPEG, PNG, and GIF, and is not limited herein.

Optionally, the face photo may be a face photographing frame of a candidate who is first provided by the system, and the applicant takes a picture in the photo frame of the applicant's face and uploads the face of the candidate. The photo is sent to the system so that the photo of the candidate's face received by the system is a uniform specification.

The face photo may also be an image processing of the received face photo after receiving the face photo sent by the applicant, for example, the received face photo may be cropped to be unified. The processing of the pixel size and the like is advantageous for the subsequent recognition process of the face image of the corresponding recruiter.

In step S2, the face image is identified by using a predetermined recognition model, and the performance level corresponding to the face photo is obtained, so that the recruiter can recruit the candidate by referring to the obtained performance level; The determined recognition model is a deep convolutional neural network model obtained by training a preset number of face sample pictures marked with different performance levels in advance.

In this embodiment, after obtaining the face image of the candidate, the pre-trained recognition model is used to identify the face image of the employed person, and the recognition result of the face image of the applicant in the recognition model is identified. The recognition model can be continuously trained, learned, verified, optimized, etc. by manually identifying a plurality of preset number of face sample images marked with different performance levels, so as to train them to accurately identify the corresponding scores of different performance levels. Model. For example, the recognition model may employ a Convolutional Neural Network (CNN) model such as AlexNet, caffeNet or ResNet, and the like. For example, the different performance levels marked can be five levels of “A, B, C, D, E”, or “Excellent, Good, Medium, and Poor”, which can be extracted from the preset employee performance appraisal database. The face sample picture, the preset employee performance appraisal database contains the performance appraisal level of each employee in the historical performance appraisal record in the actual work and the corresponding employee face image, which can be extracted from the preset employee performance appraisal database. The employee face image of the performance level employee is used as the face sample picture, and the performance level corresponding to the employee is marked, and the recognition model can be obtained based on the face sample picture and the performance level of the annotation. After obtaining the face image of the candidate, the trained recognition model can be used to identify the face image of the employed person, and the performance level corresponding to the face image of the candidate is identified, such as “A, B, C, D , E" or "Excellent, Good, Medium, Poor" one of the performance levels.

In an optional implementation manner, the training process of the predetermined recognition model is as follows:

A. Prepare a corresponding preset number of face sample images for each preset performance level (such as "A, B, C, D, E" or "Y, L, ZHONG, D", etc.) for each sample. The picture is labeled with the corresponding performance level. For example, the employee face image of each employee of different performance levels may be extracted from the preset employee performance appraisal database as a face sample picture, and the performance level corresponding to the employee is marked.

B. Perform image preprocessing on each face sample image to obtain a training picture to be trained by the model. The model training is performed by performing image preprocessing such as cropping and flipping on each face sample image, so that each face sample image is processed into a standard face sample image with uniform specifications and the face is in the middle, which can effectively improve the model. The authenticity and accuracy of the training.

C. Divide all training pictures into a first scale (for example, 75%) of the training set, a second ratio (eg, 25%) of the validation set;

D. training the predetermined recognition model by using the training set;

E. Using the verification set to verify the accuracy of the training recognition model, if the accuracy rate is greater than or equal to the preset accuracy rate (for example, 95%), the training ends, or if the accuracy rate is less than the preset accuracy rate, increase The number of face sample pictures corresponding to each performance level and the above steps B, C, D, and E are re-executed until the accuracy of the trained recognition model is greater than or equal to the preset accuracy rate.

After using the trained recognition model to identify the performance level corresponding to the face image of the candidate, the recruiter can refer to the obtained performance level to recruit the candidate. For example, different scores are set according to the level of performance, and the higher the performance level of the applicants identified by the trained recognition model, the higher the recruitment score, or the performance level is excellent or A can be added to the recruitment score. Points to assist recruitment.

Compared with the prior art, the present embodiment identifies a face image of a hired person by using a deep convolutional neural network model obtained by training a preset number of face sample pictures marked with different performance levels, and according to the recognition result. Predict the performance level of the candidate. Through deep learning, find out the relationship between the face and the performance level, and determine the corresponding performance level according to the face of the candidate when recruiting, and can use the predicted performance level of the candidate as the reference factor for recruitment. To provide a new recruitment reference factor to assist recruitment and improve the quality of recruitment.

In an optional embodiment, on the basis of the foregoing embodiment of FIG. 1 , when the recruiting system 10 is executed by the processor 12 to implement the step S2, the method further includes:

Find the employee face photos corresponding to different performance levels from the preset database. For example, the employee face images of employees with different performance levels can be extracted from the preset employee performance appraisal database. Further, the facial features such as eyebrows, eyes, ears, nose, and/or mouth in each employee's face photo may be extracted, and the number of preset classification features of each organ in the photo of the facial features corresponding to different performance levels may be counted. According to the statistical results, the performance classification characteristics of each organ corresponding to different performance levels are determined. That is, the characteristics of each organ of the facial features in the employee's face photos of employees of different performance levels are statistically summarized. The specific features that can be referred to for each organ of the facial features include, but are not limited to, the curvature of the upper lip and the inner corner of the two eyes. The distance, the angle formed by the connection of the tip of the nose with the two corners of the mouth, the shape of the eyebrows, the type of the ear, and the like. According to these reference features, some statistical analysis can be done. For example, among the reference features corresponding to each organ in the photos of the facial features corresponding to different performance levels, which of the feature values in the same class have a smaller variance, between different classes. The variance is large. It is also possible to put certain category reference features (such as the type of nose) into the second classifier for training. If a higher accuracy is obtained, the reference feature has a value as a classification reference.

Here, in the eyes of the facial features, the preset classification feature, that is, the corresponding reference feature, is an example of the distance between the two inner corners of the eye. The employee's face photo corresponding to each performance level is counted as the eye of the facial features. The number of distributions of the distance parameters between the two inner corners of the eye, such as the number of distributions of the distance parameters between the two inner corners of the eyes in each photo of the number of employee face photos corresponding to the statistical performance level A, may select a number of employee faces The distance parameter between the two inner corners of the upper eye in the photo (but a specific parameter value, or a smaller parameter value range such as 50mm-51mm) is the most distributed eye performance classification feature corresponding to the performance level A. According to this, the eye performance classification characteristics corresponding to each performance level B, C, D, and E can be obtained.

After determining the performance classification characteristics of each organ corresponding to different performance levels, the identification models corresponding to the respective organs may be separately trained for each organ, for example, in an optional implementation manner, the training process of the recognition model corresponding to each organ as follows:

A. Prepare a corresponding preset number of images of each organ sample in the five senses for each performance level (such as "A, B, C, D, E" or "Excellent, Good, Medium, Poor", etc.) for each organ The sample picture marks a corresponding performance level; wherein each organ sample picture is a picture containing each organ performance classification feature corresponding to each different performance level. For example, for the eye organ, a picture of the eye performance classification feature corresponding to each performance level may be prepared. For example, the eye performance classification feature corresponding to the performance level A is “the distance parameter between the two inner corners is between 50 mm and 51 mm”, then A sample picture with a preset number of features corresponding to the "distance between two inner corners of the eye between 50mm and 51mm" can be prepared for performance level A.

B. Dividing the organ sample picture corresponding to each organ into a training set of a first ratio (for example, 75%) and a verification set of a second ratio (for example, 25%);

C. Training the identification model corresponding to each organ by using the training set;

D. Using the verification set to verify the accuracy of the trained recognition model, if the accuracy of the recognition model corresponding to all organs is greater than or equal to the preset accuracy rate (for example, 95%), the training ends, or if there is an organ corresponding If the accuracy of the recognition model is less than the preset accuracy rate, the number of organ sample images corresponding to different performance levels of the organ is increased and steps B, C, and D are re-executed. That is to say, the accuracy of the recognition model corresponding to all organs can reach a certain requirement (for example, 95%), and the model training process can be completed.

After training the recognition model corresponding to each organ, the trained recognition model can be used to identify the face image of the employed person. First, extract the facial features in the face image of the applicant, and then use the identification model corresponding to each of the trained organs to identify the performance level corresponding to each organ in the extracted facial features, and then identify each of the facial features of the applicant. The performance level corresponding to the organ. Finally, according to the preset different organ weights, and the corresponding performance levels of the identified organs, the performance level corresponding to the face image of the candidate is calculated. Specifically, in an optional implementation manner, the formula for calculating the performance level M corresponding to the face photo is:

M=a*M1+b*M2+c*M3+d*M4+e*M5

Where a is the weighting coefficient of the performance level M1 corresponding to the eyebrow, b is the weighting coefficient of the performance level M2 corresponding to the eye, c is the weighting coefficient of the performance level M3 corresponding to the ear, and d is the weighting coefficient of the performance level M4 corresponding to the nose, e is the weighting factor of the performance level M5 corresponding to the mouth.

For example, the weight coefficients a, b, c, d, e of different organs may be preset, such as the weight coefficient a of the eyebrows is 0.1, the weight coefficient b of the eye is 0.3, the weight coefficient c of the ear is 0.3, and the weight coefficient of the nose d is 0.1 and the weight coefficient d of the mouth is 0.2. At the same time, the corresponding indicator values are assigned to different performance levels. For example, the performance level corresponds to an indicator value of 5, the performance level “good” corresponds to an indicator value of 4, and the performance level “middle” corresponds to an indicator value of 2. The performance level "poor" corresponds to an indicator value of 1. Then, if the trained recognition model is used to identify the face image of the employed person, the performance level of the eyebrows in the facial features of the candidate is “good”, the performance level of the eye is “good”, and the performance level of the ear is “medium”. "The performance level of the nose is "poor", and the performance level of the mouth is "excellent", then the performance level corresponding to the face image of the candidate is calculated to be M=0.1*4+0.3*4+0.3*2+ 0.1*1+0.2*5=3.3, then the performance grade M scores of different candidates can be calculated according to this, sorted according to the level of the score to assist recruitment. It is also possible to pre-set the correspondence between the performance level M score and the performance level, and classify them according to the performance level of the candidate to different performance levels.

As shown in FIG. 2, FIG. 2 is a schematic flowchart of an embodiment of a recruitment method according to the present application. The recruitment method includes the following steps:

In step S10, a photo of the face of the candidate is obtained.

In this embodiment, the recruiting system receives the face photo of the applicant sent by the user, for example, receiving a photo of the face of the applicant sent by the user through a mobile phone, a tablet computer, a self-service terminal device, and the like, such as receiving the user on the mobile phone, the tablet computer, A photo of the candidate's face sent by the pre-installed client in the terminal such as the self-service terminal device, or a photo of the candidate's face sent by the user on the browser system in the terminal such as a mobile phone, a tablet computer, or a self-service terminal device. The face image of the applicant may be various image format types such as JPEG, PNG, and GIF, and is not limited herein.

The face photo may also be an image processing of the received face photo after receiving the face photo sent by the applicant, for example, the received face photo may be cropped to a uniform pixel size, etc., to facilitate subsequent The identification process of the face image of the applicant is more precise.

Step S20, identifying the face photo by using a predetermined recognition model, Obtaining a performance level corresponding to the face photo for the recruiter to refer to the obtained performance level to recruit the candidate; wherein the predetermined recognition model is a pre-determined number of preset people with different performance levels A deep convolutional neural network model obtained by training a face sample picture.

C. Divide all training pictures into a training set of a first ratio (for example, 75%) and a verification set of a second ratio (for example, 25%);

D. training the predetermined recognition model by using the training set;

E. verifying the accuracy of the trained recognition model by using the verification set, if the accuracy is large At or equal to the preset accuracy rate (for example, 95%), the training ends, or, if the accuracy rate is less than the preset accuracy rate, increase the number of face sample pictures corresponding to each performance level and re-execute the above steps B, C, D, E, until the accuracy of the training recognition model is greater than or equal to the preset accuracy.

In an optional embodiment, based on the foregoing embodiment, the method further includes:

Here, in the eyes of the facial features, the preset classification feature, that is, the corresponding reference feature, is an example of the distance between the two inner corners of the eye. The employee's face photo corresponding to each performance level is counted as the eye of the facial features. The number of distribution parameters of the distance between the two inner corners of the eye, such as the number of photos of the employee's face photo corresponding to the statistical performance level A The number of distance parameters between the two inner corners of the eye in the film, the distance parameter between the two inner corners of the upper eye in the number of employees' faces may be selected (but a specific parameter value, or a smaller one) The parameter value range is 50mm-51mm. The eye distribution performance feature corresponding to the performance level A is the most distributed. According to this, the eye performance classification characteristics corresponding to each performance level B, C, D, E can be obtained.

M=a*M1+b*M2+c*M3+d*M4+e*M5

Moreover, the present application also provides a computer readable storage medium storing a recruitment system, the recruitment system being executable by at least one processor to cause the at least one processor to perform the implementation as described above The steps of the recruitment method in the example, the specific implementation processes of the steps S10, S20, etc. of the recruitment method are as described above, and are not described herein again.

It is to be understood that the term "comprises", "comprising", or any other variants thereof, is intended to encompass a non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes those elements. It also includes other elements that are not explicitly listed, or elements that are inherent to such a process, method, article, or device. An element that is defined by the phrase "comprising a ..." does not exclude the presence of additional equivalent elements in the process, method, item, or device that comprises the element.

Through the description of the above embodiments, those skilled in the art can clearly understand that the foregoing embodiment method can be implemented by means of software plus a necessary general hardware platform, and can also be implemented by hardware, but in many cases, the former is A better implementation. Based on such understanding, the technical solution of the present application, which is essential or contributes to the prior art, may be embodied in the form of a software product stored in a storage medium (such as ROM/RAM, disk, The optical disc includes a number of instructions for causing a terminal device (which may be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to perform the methods described in various embodiments of the present application.

The preferred embodiments of the present application have been described above with reference to the drawings, and are not intended to limit the scope of the application. The serial numbers of the embodiments of the present application are merely for the description, and do not represent the advantages and disadvantages of the embodiments. In addition, although the logical order is shown in the flowchart, in some cases, The steps shown or described may be performed in an order different than that herein.

A person skilled in the art can implement the present application in various variants without departing from the scope and spirit of the present application. For example, the features of one embodiment can be used in another embodiment to obtain another embodiment. Any modifications, equivalent substitutions and improvements made within the technical concept of the application should be within the scope of the application.

Claims

An electronic device, comprising: a memory, a processor, wherein the memory stores a recruitment system operable on the processor, and the recruitment system is implemented by the processor as follows step:

A1. Obtain a photo of the face of the applicant;

B1, the face photo is identified by using a predetermined recognition model, and the performance level corresponding to the face photo is obtained, so that the recruiter can recruit the candidate according to the obtained performance level; wherein the predetermined The recognition model is a deep convolutional neural network model obtained by training a preset number of face sample images marked with different performance levels in advance.
The electronic device according to claim 1, wherein when the recruiting system is executed by the processor to implement the step B1, the method further comprises:

Find the employee face photos corresponding to different performance levels from the preset database, extract the facial features in each employee's face photos, and count the number of preset classification features of each organ in the photos of the facial features corresponding to different performance levels. And determining the performance classification characteristics of each organ corresponding to different performance levels according to the statistical results;

Wherein the facial features include at least one of an eyebrow, an eye, an ear, a nose and a mouth; the predetermined classification feature is a shape of an eyebrow, a distance between two inner corners of the eye, a type of the ear, a tip of the nose and two corners of the mouth Connect the angle formed or the curvature of the upper lip.
The electronic device of claim 1, wherein the training process of the predetermined recognition model is as follows:

A. Prepare corresponding preset number of images of each organ in the five senses for each performance level, and mark the corresponding performance level for each organ sample image; wherein each organ sample image is a performance of each organ corresponding to each different performance level. a picture of the classification feature;

B. Dividing the image of the organ sample corresponding to each organ into a training set of the first ratio and a verification set of the second ratio;

C. Training the identification model corresponding to each organ by using the training set;

D. Using the verification set to verify the accuracy of the training recognition model, if the accuracy of the recognition model corresponding to all organs is greater than or equal to the preset accuracy rate, the training ends, or if the accuracy of the identification model corresponding to the organ is If it is less than the preset accuracy rate, the number of organ sample pictures corresponding to different performance levels of the organ is increased and the above steps B, C, and D are re-executed.
The electronic device according to claim 2, wherein said said advance The training process for the identified recognition model is as follows:

A. Prepare corresponding preset number of images of each organ in the five senses for each performance level, and mark the corresponding performance level for each organ sample image; wherein each organ sample image is a performance of each organ corresponding to each different performance level. a picture of the classification feature;

B. Dividing the image of the organ sample corresponding to each organ into a training set of the first ratio and a verification set of the second ratio;

C. Training the identification model corresponding to each organ by using the training set;

D. Using the verification set to verify the accuracy of the training recognition model, if the accuracy of the recognition model corresponding to all organs is greater than or equal to the preset accuracy rate, the training ends, or if the accuracy of the identification model corresponding to the organ is If it is less than the preset accuracy rate, the number of organ sample pictures corresponding to different performance levels of the organ is increased and the above steps B, C, and D are re-executed.
The electronic device according to claim 3, wherein when the recruiting system is executed by the processor to implement the step B1, the method further comprises:

Extracting the facial features in the photo of the face;

Identifying the performance level corresponding to each organ in the extracted facial features by using the trained recognition model of each organ;

The formula for calculating the performance level M corresponding to the face photo according to the preset different organ weights and the corresponding performance levels of the identified organs is:

M=a*M1+b*M2+c*M3+d*M4+e*M5

Where a is the weighting coefficient of the performance level M1 corresponding to the eyebrow, b is the weighting coefficient of the performance level M2 corresponding to the eye, c is the weighting coefficient of the performance level M3 corresponding to the ear, and d is the weighting coefficient of the performance level M4 corresponding to the nose, e is the weighting factor of the performance level M5 corresponding to the mouth.
The electronic device according to claim 4, wherein when the recruitment system is executed by the processor to implement the step B1, the method further comprises:

Extracting the facial features in the photo of the face;

Identifying the performance level corresponding to each organ in the extracted facial features by using the trained recognition model of each organ;

The formula for calculating the performance level M corresponding to the face photo according to the preset different organ weights and the corresponding performance levels of the identified organs is:

M=a*M1+b*M2+c*M3+d*M4+e*M5

Where a is the weighting coefficient of the performance level M1 corresponding to the eyebrow, b is the weighting coefficient of the performance level M2 corresponding to the eye, c is the weighting coefficient of the performance level M3 corresponding to the ear, and d is the weighting coefficient of the performance level M4 corresponding to the nose, e is the performance of the mouth The weighting factor of level M5.
The electronic device according to claim 1, wherein the training process of the predetermined recognition model is as follows:

A. Prepare a corresponding preset number of face sample pictures for each performance level, and mark a corresponding performance level for each face sample picture;

B. Perform image preprocessing on each face sample image to obtain a training picture to be trained by the model;

C. Divide all training pictures into a training set of a first ratio and a verification set of a second ratio;

D. training the predetermined recognition model by using the training set;

E. Using the verification set to verify the accuracy of the training recognition model, if the accuracy rate is greater than or equal to the preset accuracy rate, the training ends, or if the accuracy rate is less than the preset accuracy rate, the person corresponding to each performance level is increased. The number of face sample pictures and re-execute steps B, C, D, E above.
A recruitment method, characterized in that the recruitment method comprises:

Step 1: Obtain a photo of the face of the candidate;

Step 2: Identifying the face photo by using a predetermined recognition model, and obtaining a performance level corresponding to the face photo, for the recruiter to refer to the obtained performance level to recruit the candidate; wherein, the advance The determined recognition model is a deep convolutional neural network model obtained by training a preset number of face sample pictures marked with different performance levels in advance.
The hiring method according to claim 8, wherein the method further comprises:

Find the employee face photos corresponding to different performance levels from the preset database, extract the facial features in each employee's face photos, and count the number of preset classification features of each organ in the photos of the facial features corresponding to different performance levels. And determining the performance classification characteristics of each organ corresponding to different performance levels according to the statistical results;

Wherein the facial features include at least one of an eyebrow, an eye, an ear, a nose and a mouth; the predetermined classification feature is a shape of an eyebrow, a distance between two inner corners of the eye, a type of the ear, a tip of the nose and two corners of the mouth Connect the angle formed or the curvature of the upper lip.
The recruitment method according to claim 8, wherein the training process of the predetermined recognition model is as follows:

A. Prepare corresponding preset number of images of each organ in the five senses for each performance level, and mark the corresponding performance level for each organ sample image; wherein each organ sample image is a performance of each organ corresponding to each different performance level. a picture of the classification feature;

B. Dividing the image of the organ sample corresponding to each organ into a training set of the first ratio and a verification set of the second ratio;

C. Training the identification model corresponding to each organ by using the training set;

D. Using the verification set to verify the accuracy of the training recognition model, if the accuracy of the recognition model corresponding to all organs is greater than or equal to the preset accuracy rate, the training ends, or if the accuracy of the identification model corresponding to the organ is If it is less than the preset accuracy rate, the number of organ sample pictures corresponding to different performance levels of the organ is increased and the above steps B, C, and D are re-executed.
The recruitment method according to claim 9, wherein the training process of the predetermined recognition model is as follows:

A. Prepare corresponding preset number of images of each organ in the five senses for each performance level, and mark the corresponding performance level for each organ sample image; wherein each organ sample image is a performance of each organ corresponding to each different performance level. a picture of the classification feature;

B. Dividing the image of the organ sample corresponding to each organ into a training set of the first ratio and a verification set of the second ratio;

C. Training the identification model corresponding to each organ by using the training set;

D. Using the verification set to verify the accuracy of the training recognition model, if the accuracy of the recognition model corresponding to all organs is greater than or equal to the preset accuracy rate, the training ends, or if the accuracy of the identification model corresponding to the organ is If it is less than the preset accuracy rate, the number of organ sample pictures corresponding to different performance levels of the organ is increased and the above steps B, C, and D are re-executed.
The recruitment method of claim 10, wherein the step two comprises:

Extracting the facial features in the photo of the face;

Identifying the performance level corresponding to each organ in the extracted facial features by using the trained recognition model of each organ;

The formula for calculating the performance level M corresponding to the face photo according to the preset different organ weights and the corresponding performance levels of the identified organs is:

M=a*M1+b*M2+c*M3+d*M4+e*M5

Where a is the weighting coefficient of the performance level M1 corresponding to the eyebrow, b is the weighting coefficient of the performance level M2 corresponding to the eye, c is the weighting coefficient of the performance level M3 corresponding to the ear, and d is the weighting coefficient of the performance level M4 corresponding to the nose, e is the weighting factor of the performance level M5 corresponding to the mouth.
The recruitment method according to claim 11, wherein said step two include:

Extracting the facial features in the photo of the face;

Identifying the performance level corresponding to each organ in the extracted facial features by using the trained recognition model of each organ;

The formula for calculating the performance level M corresponding to the face photo according to the preset different organ weights and the corresponding performance levels of the identified organs is:

M=a*M1+b*M2+c*M3+d*M4+e*M5

Where a is the weighting coefficient of the performance level M1 corresponding to the eyebrow, b is the weighting coefficient of the performance level M2 corresponding to the eye, c is the weighting coefficient of the performance level M3 corresponding to the ear, and d is the weighting coefficient of the performance level M4 corresponding to the nose, e is the weighting factor of the performance level M5 corresponding to the mouth.
The recruitment method according to claim 8, wherein the training process of the predetermined recognition model is as follows:

A. Prepare a corresponding preset number of face sample pictures for each performance level, and mark a corresponding performance level for each face sample picture;

B. Perform image preprocessing on each face sample image to obtain a training picture to be trained by the model;

C. Divide all training pictures into a training set of a first ratio and a verification set of a second ratio;

D. training the predetermined recognition model by using the training set;

E. Using the verification set to verify the accuracy of the training recognition model, if the accuracy rate is greater than or equal to the preset accuracy rate, the training ends, or if the accuracy rate is less than the preset accuracy rate, the person corresponding to each performance level is increased. The number of face sample pictures and re-execute steps B, C, D, E above.
A computer readable storage medium, wherein the computer readable storage medium stores a recruitment system, the recruitment system being executable by at least one processor to cause the at least one processor to perform the following steps:

A2. Obtain a photo of the face of the applicant;

B2. Identifying the face photo by using a predetermined recognition model, and obtaining a performance level corresponding to the face photo, for the recruiter to refer to the obtained performance level to recruit the candidate; wherein the predetermined The recognition model is a deep convolutional neural network model obtained by training a preset number of face sample images marked with different performance levels in advance.
The computer readable storage medium according to claim 15, wherein when the recruiting system is executed by the processor to implement the step B2, the method further comprises:

Find out the photo of the employee's face corresponding to each performance level from the preset database. Taking out the facial features in each employee's face photo, counting the number of preset classification features of each organ in the photos of the facial features corresponding to different performance levels, and determining the performance classification characteristics of each organ corresponding to different performance levels according to the statistical results;

Wherein the facial features include at least one of an eyebrow, an eye, an ear, a nose and a mouth; the predetermined classification feature is a shape of an eyebrow, a distance between two inner corners of the eye, a type of the ear, a tip of the nose and two corners of the mouth Connect the angle formed or the curvature of the upper lip.
The computer readable storage medium of claim 15, wherein the training process of the predetermined recognition model is as follows:

A. Prepare corresponding preset number of images of each organ in the five senses for each performance level, and mark the corresponding performance level for each organ sample image; wherein each organ sample image is a performance of each organ corresponding to each different performance level. a picture of the classification feature;

B. Dividing the image of the organ sample corresponding to each organ into a training set of the first ratio and a verification set of the second ratio;

C. Training the identification model corresponding to each organ by using the training set;

D. Using the verification set to verify the accuracy of the training recognition model, if the accuracy of the recognition model corresponding to all organs is greater than or equal to the preset accuracy rate, the training ends, or if the accuracy of the identification model corresponding to the organ is If it is less than the preset accuracy rate, the number of organ sample pictures corresponding to different performance levels of the organ is increased and the above steps B, C, and D are re-executed.
The computer readable storage medium of claim 16, wherein the training process of the predetermined recognition model is as follows:

A. Prepare corresponding preset number of images of each organ in the five senses for each performance level, and mark the corresponding performance level for each organ sample image; wherein each organ sample image is a performance of each organ corresponding to each different performance level. a picture of the classification feature;

B. Dividing the image of the organ sample corresponding to each organ into a training set of the first ratio and a verification set of the second ratio;

C. Training the identification model corresponding to each organ by using the training set;

D. Using the verification set to verify the accuracy of the training recognition model, if the accuracy of the recognition model corresponding to all organs is greater than or equal to the preset accuracy rate, the training ends, or if the accuracy of the identification model corresponding to the organ is If it is less than the preset accuracy rate, the number of organ sample pictures corresponding to different performance levels of the organ is increased and the above steps B, C, and D are re-executed.
The computer readable storage medium of claim 17 wherein: When the recruiting system is executed by the processor to implement the step B2, the method further includes:

Extracting the facial features in the photo of the face;

Identifying the performance level corresponding to each organ in the extracted facial features by using the trained recognition model of each organ;

The formula for calculating the performance level M corresponding to the face photo according to the preset different organ weights and the corresponding performance levels of the identified organs is:

M=a*M1+b*M2+c*M3+d*M4+e*M5

Where a is the weighting coefficient of the performance level M1 corresponding to the eyebrow, b is the weighting coefficient of the performance level M2 corresponding to the eye, c is the weighting coefficient of the performance level M3 corresponding to the ear, and d is the weighting coefficient of the performance level M4 corresponding to the nose, e is the weighting factor of the performance level M5 corresponding to the mouth.
The computer readable storage medium according to claim 18, wherein when the recruiting system is executed by the processor to implement the step B2, the method further comprises:

Extracting the facial features in the photo of the face;

Identifying the performance level corresponding to each organ in the extracted facial features by using the trained recognition model of each organ;

The formula for calculating the performance level M corresponding to the face photo according to the preset different organ weights and the corresponding performance levels of the identified organs is:

M=a*M1+b*M2+c*M3+d*M4+e*M5

Where a is the weighting coefficient of the performance level M1 corresponding to the eyebrow, b is the weighting coefficient of the performance level M2 corresponding to the eye, c is the weighting coefficient of the performance level M3 corresponding to the ear, and d is the weighting coefficient of the performance level M4 corresponding to the nose, e is the weighting factor of the performance level M5 corresponding to the mouth.