Enterprise employee training and development management system based on online cloud platform

Abstract

The invention provides an enterprise employee training and development management system based on an online cloud platform. The enterprise employee training and development management system comprises a user management module used for obtaining user information of employee users; the curriculum library management module is used for establishing a curriculum library to store different classes of training curriculums and managing the stored training curriculums; the course recommendation module is used for calculating a recommendation degree factor of each course in a course library to the employee user in real time according to the user information of the employee user, and generating training course recommendation information for the employee user; and the learning evaluation module is used for evaluating the training effect of the employee user for the training course in the process that the employee user completes the training course, and updating the user information according to the training effect evaluation information. According to the invention, on the basis of a proposed training effect feedback mechanism, the recommended course is adjusted in real time, and the personalized adaptation effect and the intelligent level of employee training are helped to be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of enterprise management, and in particular to an enterprise employee training and development management system based on an online cloud platform. Background Art

[0002] With the expansion of enterprise scale and the development of globalization, the management of enterprise employee training and development has become increasingly complex. Traditional training methods often rely on face-to-face teaching and paper materials, which are not only inefficient but also difficult to meet the personalized needs of different employees. The emergence of online training systems based on cloud platforms provides a new solution for enterprises, enabling employees to receive training anytime and anywhere, and at the same time providing a more flexible and efficient management tool for enterprises.

[0003] Currently, in the actual application of online training systems, there are already some technologies for personalized training of employees, such as: For example, the patent with the patent number CN119599843A discloses an employee intelligent training platform, and specifically discloses recommending several courses and learning paths according to the basic employment information of employees.

[0004] For example, the patent with the patent number CN115082041A discloses a user information management method, and specifically discloses obtaining the satisfaction of users with courses by acquiring the facial expression features and click data of users when browsing different courses, and further constructing a user portrait, and completing the recommendation of courses according to the user portrait.

[0005] However, in the above-mentioned existing technologies, there are still the following problems: Affected by factors such as employees' learning ability and subjective preferences (such as preferences for training methods, styles of keynote speakers, etc.), different employees have different acceptance levels of the same course. However, in the above-mentioned existing technologies, once the personalized courses and learning paths for employees are recommended, it is impossible to accurately track the real-time learning situation of employees, resulting in the situation that even after the personalized recommendation, employees are still not interested in the recommended courses, affecting the effect and quality of enterprise employee training and development management. Summary of the Invention

[0006] In view of the above problems, the present invention aims to provide an enterprise employee training and development management system based on an online cloud platform.

[0007] The object of the present invention is achieved by the following technical solutions: The present invention provides an enterprise employee training and development management system based on an online cloud platform, including a user management module, a course library management module, a course recommendation module, and a learning evaluation module; wherein, The user management module is used to obtain the user information of employee users, where the user information includes the basic information, employee development information, historical training information, and historical training effect information of employee users; The course library management module is used to establish a course library to store training courses of different classifications and manage the stored training courses; The course recommendation module is used to calculate the recommendation degree factors of each course in the course library for employee users in real time according to the user information of employee users, and obtain the recommendation degree factors of employee users for each course classification respectively; according to the recommendation degree factors of employee users for each course classification, select at least 1 course classification with the largest recommendation degree factor, and further randomly select at least 1 series of courses under each selected course classification to form the training course recommendation information for employee users; The learning evaluation module is used to call the test content corresponding to the training course to test the employee user during the process of the employee user completing the training course, and obtain the corresponding course test score; and during the process of the employee user learning the training course, periodically collect the real-time status images of the employee user, and obtain the head pose characteristics and gaze feature information of the employee user according to the obtained real-time status images, and judge the learning status of the current employee user during the training process based on the obtained head pose characteristics and gaze feature information, and obtain the learning status score of the employee user; obtain the comprehensive learning effect score of the employee user for the course according to the obtained course test score and learning status score; feedback the obtained comprehensive learning effect score of the employee user for the current course classification to the course recommendation module; The course recommendation module further includes: according to the comprehensive learning effect score of the employee user for the current course classification, when the comprehensive learning effect score is lower than the preset comprehensive learning effect score standard value, after obtaining the corresponding recommended course classification according to the recommendation degree factors of the employee user for each course classification, re-select another course series different from the current course series from the recommended course classifications, and generate training course recommendation information according to the re-selected course series.

[0008] Preferably, the system further includes a course playback module; The course playback module is used to play the corresponding training course content according to the training course recommendation information, or play the corresponding course content according to the training course information selected by the user.

[0009] Preferably, the user management module specifically includes: For entering the basic information of employee users, where the basic information of employee users includes employee names, positions, and face information; and further matching and obtaining corresponding employee development information according to the position information of employees, and completing the adjustment of employee development information; and recording the information of training courses completed by employee users on the platform and the corresponding training effect evaluation information to obtain the historical training information and historical training effect information of employee users.

[0010] Preferably, the course library management module includes a course library unit and a course classification management unit; where The course library unit is used to enter training course resources and store the training courses in the course library unit; The course classification management unit is used to classify and manage the training courses in the course library unit, including classifying and marking the training courses and marking the course series.

[0011] Preferably, the course recommendation module further includes a push unit; The push unit is used to push the obtained training course recommendation information to the corresponding employee users.

[0012] Preferably, the learning evaluation module includes a test unit; The test unit is used to call the test content corresponding to the training course to test the employee user during the process of the employee user completing the training course, and obtain the corresponding course test score.

[0013] Preferably, the learning evaluation module further includes a status monitoring unit and a comprehensive evaluation unit; The status monitoring unit is used to periodically collect the real-time status images of the employee user during the process of the employee user learning the training course, and analyze the learning status of the employee user based on the trained image analysis model according to the obtained real-time status images to obtain the learning status score of the employee user; The comprehensive evaluation unit is used to obtain the comprehensive learning effect score of the employee user for the course according to the obtained course test score and learning status score, and update the user information according to the obtained comprehensive learning effect score of the employee user for the current course classification; The comprehensive evaluation unit also feeds back the obtained comprehensive learning effect score of the employee user for the current course classification to the generation unit; The generation unit further includes: According to the comprehensive learning effect score of the employee user for the current course classification, when the comprehensive learning effect score is lower than the preset comprehensive learning effect score standard value, after obtaining the corresponding recommended course classification according to the recommendation degree factor of each course classification of the employee user, reselect another course series different from the current course series from the recommended course classification, and generate training course recommendation information according to the reselected course series.

[0014] The beneficial effects of the present invention are as follows: Based on the user information of enterprise employees collected as a foundation, appropriate training courses are screened from the course library for recommendation, thereby realizing the personalized customization of the development and ability training strategies for enterprise employees. Meanwhile, during the process of employees' training and learning, the learning effects of employees are followed up and evaluated in real time. Thus, based on the proposed training effect feedback mechanism, the recommended courses are adjusted in real time to help improve the personalized adaptation effect and intelligent level of employees' training.

[0015] Based on the comprehensive learning effect evaluation method proposed by the present invention, it can be used as a standard to adjust the course recommendation information of employees. Among them, the test scores of training courses can reflect the learning effects of employees and will be reflected in the subsequent course recommendation process. In addition, by further obtaining the learning status scores of employees based on the head pose characteristics and gaze direction characteristics during the learning process of training courses, the learning status of employees towards the courses can be reflected simultaneously. Finally, courses in the same major category are adjusted in a timely manner from different factors (such as different lecturers, different training methods, etc.). By recommending different series of courses in the same major category, it can adapt to the subjective factors of employees' preference for courses (such as different lecturers, different training methods, etc.), further improving the personalization level and adaptability of course recommendation and the training effect of employees. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The present invention is further described with reference to the accompanying drawings. However, the embodiments in the drawings do not constitute any limitation to the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to the following drawings.

[0017] Figure 1 It is a framework structure diagram of an enterprise employee training and development management system based on an online cloud platform shown in an embodiment of the present invention; Figure 2 It is a schematic diagram of the module framework of the course recommendation module and the learning evaluation module shown in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] The present invention is further described in combination with the following application scenarios.

[0019] Refer to Figure 1 As shown in the embodiment, it shows an enterprise employee training and development management system based on an online cloud platform, including a user management module, a course library management module, a course recommendation module, and a learning evaluation module; among them, The user management module is used to obtain the user information of employee users, where the user information includes the basic information, employee development information, historical training information, and historical training effect information of employee users; The course library management module is used to establish a course library to store training courses of different classifications and manage the stored training courses; The course recommendation module is used to calculate the recommendation degree factors of each course in the course library for the employee user in real time according to the user information of the employee user, and obtain the recommendation degree factors of the employee user for each course classification respectively; according to the recommendation degree factors of the employee user for each course classification, select at least 1 course classification with the largest recommendation degree factor, and further randomly select at least 1 series of courses under each selected course classification to form the training course recommendation information for the employee user; The learning evaluation module is used to call the test content corresponding to the training course to test the employee user during the process of the employee user completing the training course, and obtain the corresponding course test score; and during the process of the employee user learning the training course, periodically collect the real-time status images of the employee user, and obtain the head pose characteristics and line-of-sight characteristic information of the employee user according to the obtained real-time status images, and judge the learning status of the current employee user during the training process based on the obtained head pose characteristic information and line-of-sight characteristic information, and obtain the learning status score of the employee user; obtain the comprehensive learning effect score of the employee user for the course according to the obtained course test score and learning status score; feedback the obtained comprehensive learning effect score of the employee user for the current course classification to the course recommendation module; The course recommendation module further includes: according to the comprehensive learning effect score of the employee user for the current course classification, when the comprehensive learning effect score is lower than the preset comprehensive learning effect score standard value, after obtaining the corresponding recommended course classification according to the recommendation degree factors of the employee user for each course classification, reselect another course series different from the current course series from the recommended course classification, and generate the training course recommendation information according to the reselected course series.

[0020] In the above embodiments of the present invention, by collecting the user information of enterprise employees as a basis, appropriate training courses are screened out from the course library for recommendation, so as to realize the personalized customization of the development and ability training strategies of enterprise employees. At the same time, during the process of employees' training and learning, the learning effect of employees is followed up and evaluated in real time, so as to adjust the recommended courses in real time based on the proposed training effect feedback mechanism, and help improve the personalized adaptation effect and intelligent level of employees' training.

[0021] Based on the comprehensive learning effect evaluation method proposed by the present invention, it can be used as a standard to adjust the course recommendation information of employees. Among them, the test scores of training courses can reflect the learning effect of employees, which will be reflected in the subsequent course recommendation process. In addition, by further obtaining the learning status score of employees based on the head pose characteristics and gaze direction characteristics of employees during the training course learning process, it can simultaneously reflect the learning status of employees towards the course. Finally, by comprehensively considering different factors (such as different lecturers, different training methods, etc.), the courses in the same major category can be adjusted in a timely manner. By recommending different series of courses in the same major category, it can adapt to the subjective factors of employees' preferences for courses (such as different lecturers, different training methods, etc.), further improve the personalization level and adaptability of course recommendations, and improve the effect of employee training.

[0022] Among them, the system of the present invention can be built based on a cloud platform or a cloud server, etc. Employee users can access the system of the present invention through intelligent devices such as computers or mobile terminals, so as to realize online learning of training courses.

[0023] Preferably, the system further includes a course playback module; The course playback module is used to play the corresponding training course content according to the training course recommendation information, or play the corresponding course content according to the training course information selected by the user.

[0024] Employee users can select corresponding courses according to the recommended course information or according to the series of courses they are currently learning. The cloud platform plays the corresponding training course content for employee users to conduct online learning of the corresponding courses.

[0025] Preferably, the user management module specifically includes: It is used to input the basic information of employee users, where the basic information of employee users includes employee name, position, face information, etc.; and further match and obtain the corresponding employee development information according to the position information of employees, and complete the adjustment of employee development information; and record the information of training courses completed by employee users on the platform and the corresponding training effect evaluation information to obtain the historical training information and historical training effect information of employee users.

[0026] Among them, the employee development information includes the development direction information of employees, which can be selected by the employee users themselves or automatically matched according to the positions of employee users. According to the settings of the system, for each course classification in the system, corresponding matching degree information is set for different development directions. For example, for courses oriented to management, the matching degree value for the development direction based on management will relatively increase, but at the same time, for all other courses involved in other classifications, the matching degree values will also show different degrees of values, which can be adjusted according to the actual situation.

[0027] By collecting the basic information of employee users, on the one hand, it is possible to realize the recommendation and recording of initial recommended courses based on the basic information of employees, and at the same time, based on the historical training information and historical training effect evaluation information of employees, record the current learning progress and learning effect of employee users as important reference information for subsequent course recommendation.

[0028] Among them, in the above implementation, the face information of employees is further recorded, which is convenient for subsequent learning effect evaluation. When conducting real-time learning effect evaluation on the state of employee users during the learning process, it serves as the basis for evaluation.

[0029] Preferably, the course library management module includes a course library unit and a course classification management unit; among them, The course library unit is used to input training course resources and store the training courses in the course library unit; The course classification management unit is used to classify and manage the training courses in the course library unit, including classifying and marking the training courses and marking the course series.

[0030] Each classified course contains multiple series of courses, and each series of courses includes at least 1 class hour of training courses.

[0031] Through the establishment of the course library, it is possible to centrally manage and update courses of different classifications. At the same time, when the cloud platform constructs the course library, it specifically incorporates different series of courses for each classified course (for example, explanations of the same classified course by different lecturers and in different forms are called different series of courses). This is convenient for subsequent more personalized and better-effect course recommendation.

[0032] Among them, during the learning process of an employee user for a certain series of courses under a certain classification, the current learning progress of the employee user will be recorded, specifically the current number of class hours. When the employee studies next time, they can continue the current learning progress and continue learning within the same series of courses. Or they can study according to the specified progress based on the selected chapters by the employee.

[0033] The classification of courses can be based on the classification of specific course content, such as management, technology, programming, etc., or be classified according to more specific branch contents such as project management, human resources management, material management, etc.

[0034] Preferably, referring to Figure 2 , the course recommendation module includes a recommendation analysis unit and a generation unit; among them, The recommendation analysis unit is used to calculate the recommendation degree factor of each course in the course library for the employee user in real time according to the user information of the employee user. The recommendation degree factor calculation function adopted is: ; In the formula, RL(a,b) represents the employee user a recommendation factor for course classification b ; T(a,b) represents the total learning duration of the employee user a in the course classification b ; T(a,c) represents the total learning duration of the employee user a in the course classification c , where the variable c= 1,2,…N , N represents the total number of course classifications, P(a,b) represents the test score of the employee user a in the course classification b ; P(a,c) represents the test score of the employee user a in the course classification c ; D(a,b) represents the development fit information of the course classification b for the employee user a ; D(a,c) represents the development fit information of the course classification c for the employee user a ; TL(a,b) represents the continuous learning duration of the employee user a in the course classification b ; θTL represents the set continuous duration threshold, k represents the set decay adjustment factor, ω T represents the set duration weight factor, ω P represents the set score weight factor, D(a,b) represents the set development weight factor; Obtain the recommendation factors of the employee user for each course classification respectively; The generation unit is used to select at least 1 course classification with the largest recommendation factor according to the recommendation factors of the employee user for each course classification, and further randomly select at least 1 series of courses under each selected course classification to form the training course recommendation information for the employee user.

[0035] In the above embodiments of the present invention, when making personalized course recommendations for employee users based on the course recommendation module, first, based on the development information and learning information of the employees, the suitability and attention degree of the employees to different classified courses are judged, so as to serve as the basis for recommending the corresponding classified courses. At the same time, adding the test scores as the basis can further judge the degree of content mastery of the employee users for the courses, so as to reflect the talents, interest points or basic mastery degree of the employee users in turn as the basis; and by introducing the continuous learning duration to inhibit the recommendation factor, the idea of scientific training can be combined to improve the balanced development of the employee users, and avoid the lack of all-round development due to the employee users only learning single knowledge for a long time. Through the above ideas and the proposed specific recommendation factor calculation function, the scientific development and personalized training effects for employees can be realized, effectively improving the intelligent level and recommendation effect of training course recommendations.

[0036] When recommending a certain major category of courses to employees, a random method can be used to first recommend one series of courses under this major category to the employee users, and at the same time, during the process of the employees learning this series of courses, the learning effects of the employees are evaluated to further collect more characteristic data for course recommendations.

[0037] Among them, when the user information of the employee users (including the basic information, employee development information, historical training information and historical training effect information of the employee users) is updated, the course recommendation module further synchronously updates the current training course recommendation information according to the updated user information, so that the course recommendation information can conform to the current learning situation and demand situation of the employee users in real time, which helps to improve the effect and pertinence of course recommendations.

[0038] Preferably, the course recommendation module further includes a push unit; The push unit is used to push the obtained training course recommendation information to the corresponding employee users.

[0039] Based on the course recommendation list obtained according to the real-time feedback information, it can be timely pushed to the corresponding employee users to assist the employee users in timely adjusting the selection of training courses.

[0040] Preferably, the learning evaluation module includes a test unit; The test unit is used to call the test content corresponding to the training course to test the employee users during the process of the employee users completing the training course, and obtain the corresponding course test scores.

[0041] Among them, the learning evaluation method includes testing the content of the training course, and testing the degree of the employee users' mastery of the training course content through questionnaire testing to give an intuitive feedback on the effect of the training course.

[0042] Considering that using the traditional single test score as the consideration for employees' learning effect of training courses has deficiencies in objectivity (for example, the course difficulty setting is unreasonable, and employees can complete the test without participating in the training; for example, employees do not attend classes seriously but complete the test by guessing or other cheating methods), it is prone to deviation and cannot truly grasp the learning effect of employees on training courses.

[0043] Preferably, the learning evaluation module further includes a status monitoring unit and a comprehensive evaluation unit; The status monitoring unit is used to periodically collect the real-time status images of the employee user during the learning process of the training course, and analyze the learning status of the employee user based on the obtained real-time status images using the trained image analysis model to obtain the learning status score of the employee user; The comprehensive evaluation unit is used to obtain the comprehensive learning effect score of the employee user for the course according to the obtained course test score and learning status score, and the comprehensive learning effect scoring function used is: ; Wherein, EL(a,b) represents the comprehensive learning effect score of the employee user a for the current course classification b ; P(a,b) represents the test score of the employee user a in the course classification b ; S(a,b) represents the learning status score of the employee user a in the course classification b , and the learning status score can be obtained based on the average or minimum score of the learning status scores of the employee user at multiple moments obtained according to the image analysis model during the learning process of the training course; ω EP represents the set test score weight factor, ω ES represents the learning status weight factor; Update the user information according to the obtained comprehensive learning effect score of the employee user for the current course classification.

[0044] The real-time status image is obtained by the system initiating image acquisition to the learning terminal of the employee user, and the learning terminal real-time collects the real-time status image of the employee user in front of the current interaction screen, and the status monitoring unit receives the real-time status image transmitted back by the learning terminal.

[0045] Based on the above embodiments, the present invention further introduces a periodic image sampling method for employee users in training courses to record the status of employees during training, and evaluates the learning effect of employees based on the training status of employees in online training courses. By combining the learning status in the training course and the after-class test scores, a comprehensive feedback on the learning effect of employees is provided, which helps to improve the accuracy and intelligent level of the evaluation of the learning effect of employees for specific courses.

[0046] Preferably, the comprehensive evaluation unit further feeds back the comprehensive learning effect score of the employee user for the current course classification to the generation unit; The generation unit further includes: According to the comprehensive learning effect score of the employee user for the current course classification, when the comprehensive learning effect score is lower than the preset comprehensive learning effect score standard value, after obtaining the corresponding recommended course classification according to the recommendation factor of each course classification of the employee user, another course series different from the current course series is reselected from the recommended course classification, and training course recommendation information is generated according to the reselected course series.

[0047] The push unit further pushes the re-obtained training course recommendation information to the corresponding employee user.

[0048] Based on the feedback of the comprehensive learning effect evaluation, it can be used as a standard to adjust the course recommendation information of employees. Among them, the test scores of training courses can reflect the learning effect of employees, which will be reflected in the subsequent course recommendation process. In addition, based on the comprehensive learning effect score, it can simultaneously reflect the learning status of employees for courses (whether they are interested in courses) and their grades, and adjust the courses in the same major category in a timely manner from different factors (such as different lecturers, different training methods, etc.). By recommending different courses in the same major category, it can adapt to the subjective factors of employees' preference for courses (such as different lecturers, different training methods, etc.), further improve the personalization level and adaptability of course recommendations, and improve the effect of employee training.

[0049] It should be noted that the different series of courses referred to in the above factual manner are specifically embodied as the explanations of the same course content by different lecturers, or for the same course content, but different teaching methods are adopted, which are called different series of courses under the same major category of course content.

[0050] In an exemplary scenario, when the instructor of a certain series of courses has a male bass voice, the degree of preference for this voice type among different employees is different. If the employee user does not like this voice pitch, it is easy to appear listless or distracted during the training course. In this case, another course of the same content by a female instructor is recommended. The voice pitch of the explanation is slightly higher, which is more likely to arouse the interest of employees; on the contrary, such a high-pitched voice may also make some employees more prone to fatigue. Therefore, in this case, through the method of learning state evaluation, it is possible to evaluate the learning state of employees during the training course, so as to adjust the course recommendation content, which helps to improve the effect and personalization level of course recommendation.

[0051] Among them, for the evaluation of the learning state of employees during training, it is possible to collect the face state of employees during the training course as a basis for judgment.

[0052] Preferably, the state monitoring unit specifically includes: During the process of employees learning the training course, periodically obtain the real-time state images of the employee users; Preprocess the obtained real-time state images to improve the quality of the real-time state images; Detect the face area according to the preprocessed real-time state images; Based on the detected face area, use the head pose estimation model based on Hopenet to extract the head pose feature information of the employee users HZ ; where the head pose feature information includes the yaw angle, pitch angle and roll angle of the head; Use the implementation estimation model based on L2CS-Net to extract the gaze feature information of the employee users YZ , where the gaze feature information includes the yaw angle and pitch angle of the gaze direction; Based on the obtained head pose feature information and gaze feature information, judge the learning state of the current employee user during the training process, and the learning state scoring function used is: ; Among them, S(a,b) represents the learning state score of the employee user a in the course classification b , ZP represents the total learning state score of the employee user a during the learning process of the course classification b , the variable n = 1, 2, …, ZP corresponds to the n th capture moment; fsh(n, HZ, HZT) represents the head pose judgment function. When based on then The head pose feature information of the employee user obtained at a capture moment HZ within the preset head pose standard range HZT then, obtain fsh(n, HZ, HZT) = 1 otherwise fsh(n, HZ, HZT) = 0 ; fsh(n, YZ, YZT) denote the line-of-sight feature judgment function. When based on the n line-of-sight feature information of the employee user obtained at the YZ capture moment within the preset line-of-sight standard range YZT then, obtain fsh(n, YZ, YZT) = 1 otherwise fsh(n, YZ, YZT) = 0 ; ω hz and ω yz are respectively the preset weight factors.

[0053] In the above embodiments of the present invention, based on image processing technology, it is possible to perform face recognition of employees based on the real-time status images collected during the online training courses of employees. Based on the obtained face information and combined with the image processing model based on deep learning, it is possible to accurately identify the head pose and eye focus area of the employees as the evaluation criteria for the learning status of employees during the training courses. The learning status of employees is quantified through the learning status scoring function, which accurately reflects the degree of concentration or seriousness of the employee user towards the training courses, thereby reflecting the degree of interest of the employee in the course content.

[0054] In addition, when the corresponding employee user is not detected / recognized based on the real-time status image, it is determined that the learning status score at the current moment is 0.

[0055] Among them, considering that when employees are learning online based on a computer or a mobile terminal, the environments they are in are uneven. Therefore, for the real-time status images captured from the employees' terminals, the image quality is likely to be uneven, which affects the effect of subsequent further learning status evaluation based on the real-time status images. Therefore, the status monitoring unit of the present invention first preprocesses the obtained real-time status images to improve the image quality, which helps to improve the clarity of the images and ensure the adaptability and reliability of subsequent evaluation of the learning status of employee users based on the image processing model.

[0056] Preferably, in the status monitoring unit, preprocessing the obtained real-time status images to improve the quality of the real-time status images specifically includes: According to the obtained real-time status images, divide the images into M sub-regions, where M∈[4,100] and convert the real-time status images to the Lab color space, and respectively extract the luminance channel sub-images of the status images XL, color channel sub - graph Xa and color channel sub - graph Xb ; For each sub - region, calculate the image quality factor of each sub - region respectively. The image quality factor calculation function used is: ; where Qua(i) represents the perceived quality evaluation value of the i th sub - region, σ(XL(i)) represents the standard deviation of the luminance channel values of each pixel in the i th sub - region in the luminance channel sub - graph; κ represents the set sensitivity adjustment factor, where κ ∈ [1.5, 2] ; μ(XL(i)) represents the average luminance channel value of each pixel in the i th sub - region in the luminance channel sub - graph; pζ represents the anti - zero factor, where pζ ∈ [0.01, 0.1] ; E hl (XL(i)) represents the average horizontal direction energy value of the i th sub - region in the luminance channel sub - graph, obtained by statistically calculating the horizontal direction energy of the spectrogram obtained after performing Fourier transform on the i th sub - region in the luminance channel sub - graph; E lh (XL(i)) represents the average vertical direction energy value of the i th sub - region in the luminance channel sub - graph, obtained by statistically calculating the vertical direction energy of the spectrogram obtained after performing Fourier transform on the i th sub - region in the luminance channel sub - graph; pξ represents the anti - zero factor, where pξ ∈ [0.0001,0.001] ; ω L1 and ω L2 represent the set weight factors respectively, skewness(Xa(i)) and skewness(Xb (i)) represent the skewness of the a th sub - region in the color channel sub - graph b and color channel sub - graph i respectively; kurtosis(Xa(i)) and kurtosis(Xb(i)) represent the kurtosis of the a th sub - region in the color channel sub - graph b and color channel sub - graph i respectively; β represents the set magnification adjustment factor, where β∈[1.1,1.5] ; ρRepresents the set coupling coefficient, where ρ ∈ [0.8, 0.9] ; Mark the sub-regions according to the obtained image quality factor. When Qua(i) ≤ QT1 it is the case, mark the sub-region as a low-dark region PA ; when Qua(i) ≥ QT2 it is the case, mark the sub-region as a high-brightness region PB ; mark the remaining regions as normal regions PC ; According to the marked types of the sub-regions, perform adaptive enhancement processing on each sub-region respectively. The enhancement processing function adopted is: ; In the formula, L ' (i) Represents the luminance channel value of the i th sub-region after enhancement processing; L(i) Represents the luminance channel value of the i th sub-region, φ(i) = PA, PB, PC respectively represent that the i th sub-region is marked as a low-dark region, a high-brightness region, and a normal region; |∇L(i)| Represents the average gradient value of the luminance channel values of each pixel point in the i th sub-region; G(L(i)) Represents the contrast of the co-occurrence matrix of the luminance channel values of the i th sub-region; atanh Represents the inverse hyperbolic tangent function, θ Represents the set adjustment amplitude factor, where θ ∈ [0.9, 1.1] ; Represents the exposure suppression factor, where ， The variable (x, y) represents a pixel point that belongs to the i th sub-region and whose luminance channel value is greater than or equal to 85, Num((x, y) ∈ φ(i) and L(x, y) ≥ 85) represents the qualified pixel points (x, y) the total number of, L(x, y) represents the luminance channel value of the pixel point (x, y) , pv represents the anti-zero factor, pv ∈ [0.00001,0.0001] ; erf represents the error function, represents the Laplacian sharpening function; LT represents the preset luminance standard value, where LT ∈ [65, 75] ; After completing the adaptive enhancement processing of each sub-region respectively, according to the luminance channel sub-map XL ' and the color channel sub-mapXa , the color channel sub - image Xb is reconstructed to obtain the pre - processed real - time status image.

[0057] Further, the face region is detected based on the pre - processed real - time status image.

[0058] In one scenario, set QT1 = 0.32 , QT2 = 1.85 .

[0059] In the above - mentioned embodiments of the present invention, a technical solution for pre - processing the real - time status image captured during the training course learning of employee users is proposed to improve the image quality. During the pre - processing process, first, based on the idea of image block processing, the real - time status image is divided into multiple sub - regions, and the image is converted to the Lab color space as the basis for analysis and processing. For each sub - region, a calculation function of an image quality factor is first proposed to evaluate the contrast (able to distinguish texture features and noise features), brightness energy, and chromaticity of the image based on each sub - region, so as to quantify the local quality of the image region. And the image is divided into regions according to the quantified image quality factor; for the low - dark regions lacking brightness energy, based on the idea of brightness restoration, texture stretching is performed on the low - brightness regions to specifically improve the contrast of the image while suppressing extremely low - brightness noise, which helps to improve the signal - to - noise ratio and brightness restoration level of the low - brightness regions. For the high - bright regions with large brightness energy, based on the idea of progressive enhancement, while improving the contrast of the high - bright regions, the detailed information contained therein is retained to the greatest extent, so as to optimize the local sharpness. For the general regions, based on the idea of overall adjustment, the overall sharpness of the image is adjusted to improve the overall sharpness level of the image. After finally completing the local enhancement processing of each sub - region, the image is reconstructed to obtain the pre - processed real - time status image for further subsequent learning status evaluation and analysis processing. Through the above - proposed image pre - processing technical solution, the local quality of the image can be accurately evaluated, and at the same time, adaptive enhancement adjustment is performed for the high - bright and low - dark regions. While removing image noise, the contrast of the extreme brightness regions is improved, thereby enhancing the sharpness and feature representation level of key detailed information.

[0060] It should be noted that in each embodiment of the present invention, each functional unit / module can be integrated into one processing unit / module, or each unit / module can exist physically alone, or two or more units / module can be integrated into one unit / module. The above - integrated unit / module can be implemented in the form of hardware or in the form of software functional unit / module.

[0061] Through the description of the above embodiments, those skilled in the art can clearly understand that the embodiments described herein can be implemented in hardware, software, firmware, middleware, code, or any appropriate combination thereof. For hardware implementation, the processor can be implemented in one or more of the following units: application specific integrated circuit (ASIC), digital signal processor (DSP), digital signal processing device (DSPD), programmable logic device (PLD), field programmable gate array (FPGA), processor, controller, microcontroller, microprocessor, other electronic units designed to implement the functions described herein, or a combination thereof. For software implementation, part or all of the processes of the embodiments can be completed by instructing the relevant hardware through a computer program. When implemented, the above program can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium. The computer-readable medium includes computer storage media and communication media, where the communication media includes any medium that facilitates the transfer of a computer program from one place to another. The storage media can be any available medium that can be accessed by a computer. The computer-readable medium can include, but is not limited to, RAM, ROM, EEPROM, CD-ROM or other optical disc storage, magnetic disk storage media or other magnetic storage devices, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer.

[0062] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the protection scope of the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the essence and scope of the technical solutions of the present invention.

Claims

1. An enterprise employee training and development management system based on an online cloud platform, characterized in that, It includes a user management module, a course library management module, a course recommendation module, and a learning assessment module; among them, The user management module is used to obtain the user information of employee users, where the user information includes the basic information, employee development information, historical training information, and historical training effect information of employee users; The course library management module is used to establish a course library to store training courses of different classifications and manage the stored training courses; The course recommendation module is used to calculate the recommendation factor of each course in the course library for employee users in real time according to the user information of employee users, and obtain the recommendation factor of employee users for each course classification respectively; according to the recommendation factor of employee users for each course classification, select at least 1 course classification with the largest recommendation factor, and further randomly select at least 1 series of courses under each selected course classification to form training course recommendation information for employee users; The learning assessment module is used to call the test content corresponding to the training course to test the employee user during the process of the employee user completing the training course, and obtain the corresponding course test score; and during the process of the employee user learning the training course, periodically collect the real-time status images of the employee user, and obtain the head pose characteristics and gaze feature information of the employee user according to the obtained real-time status images, and judge the learning status of the current employee user during the training process based on the obtained head pose characteristics and gaze feature information to obtain the learning status score of the employee user; obtain the comprehensive learning effect score of the employee user for the course according to the obtained course test score and learning status score; feedback the obtained comprehensive learning effect score of the employee user for the current course classification to the course recommendation module; The course recommendation module further includes: according to the comprehensive learning effect score of the employee user for the current course classification, when the comprehensive learning effect score is lower than the preset comprehensive learning effect score standard value, after obtaining the corresponding recommended course classification according to the recommendation factor of the employee user for each course classification, re-select another course series different from the current course series from the recommended course classification, and generate training course recommendation information according to the re-selected course series.

2. The enterprise employee training and development management system based on an online cloud platform according to claim 1, wherein It also includes a course playback module; The course playback module is used to play the corresponding training course content according to the training course recommendation information, or play the corresponding course content according to the training course information selected by the user.

3. An enterprise employee training and development management system based on an online cloud platform according to claim 1, characterized in that, The user management module specifically includes: It is used to input the basic information of employee users, where the basic information of employee users includes employee name, position, and face information; and further match and obtain the corresponding employee development information according to the position information of the employee, and complete the adjustment of the employee development information; and record the information of the training courses completed by the employee user on the platform and the corresponding training effect evaluation information to obtain the historical training information and historical training effect information of the employee user.

4. An enterprise employee training and development management system based on an online cloud platform according to claim 3, characterized in that, The course library management module includes a course library unit and a course classification management unit; among them, The course library unit is used to input training course resources and store the training courses in the course library unit; The course classification management unit is used to classify and manage the training courses in the course library unit, including classifying and marking the training courses and marking the course series.

5. An enterprise employee training and development management system based on an online cloud platform according to claim 3, characterized in that, The course recommendation module includes a recommendation analysis unit and a generation unit; among them, The recommendation analysis unit is used to calculate the recommendation factor of each course in the course library for the employee user in real time according to the user information of the employee user. The recommendation factor calculation function adopted is: ; Wherein, RL(a,b) represents the employee user a recommendation factor for course classification b ; T(a,b) represents the total learning duration of the employee user a in the course classification b ; T(a,c) represents the total learning duration of the employee user a in the course classification c where the variable c=1, 2,…N , N represents the total number of course classifications, P(a,b) represents the test score of the employee user a in the course classification b ; P(a,c) represents the test score of the employee user a in the course classification c ; D(a,b) represents the development fit information of the course classification b for the employee user a ; D(a,c) represents the development fit information of the course classification c for the employee user a ; TL(a,b) represents the continuous learning duration of the employee user a in the course classification b ; θTL represents the set continuous duration threshold, k represents the set decay adjustment factor, ω T represents the set duration weight factor, ω P represents the set score weight factor, D(a,b) represents the set development weight factor; Obtain the recommendation factors of the employee user for each course classification respectively; The generation unit is used to select at least 1 course classification with the largest recommendation factor according to the recommendation factors of the employee user for each course classification, and further randomly select at least 1 series of courses under each selected course classification respectively to form the training course recommendation information for the employee user.

6. The enterprise employee training and development management system based on an online cloud platform according to claim 5, wherein, The course recommendation module also includes a push unit; The push unit is used to push the obtained training course recommendation information to the corresponding employee user.

7. An enterprise employee training and development management system based on an online cloud platform according to claim 1, characterized in that, The learning evaluation module includes a test unit, a status monitoring unit and a comprehensive evaluation unit; among them,; The test unit is used to call the test content corresponding to the training course to test the employee user during the process of the employee user completing the training course, and obtain the corresponding course test score; The status monitoring unit is used to periodically collect the real-time status images of the employee user during the process of the employee user learning the training course, and analyze the learning status of the employee user based on the trained image analysis model according to the obtained real-time status images, and obtain the learning status score of the employee user; The comprehensive evaluation unit is used to obtain the comprehensive learning effect score of the employee user for the course according to the obtained course test score and learning status score. The comprehensive learning effect score function adopted is: ; Among them, EL(a,b) represents the comprehensive learning effect score of the employee user a for the current course classification b ; P(a,b) represents the test score of the employee user a in the course classification b ; S(a,b) represents the learning status score of the employee user a in the course classification b , where the learning status score can be obtained based on the average or minimum score of the learning status scores of the employee user at multiple moments obtained by the image analysis model during the employee's training course learning process; ω EP represents the set test score weight factor, ω ES represents the learning status weight factor; update the user information according to the obtained comprehensive learning effect score of the employee user for the current course classification; The comprehensive evaluation unit also feeds back the comprehensive learning effect score of the employee user for the current course classification obtained to the generation unit; The generation unit further includes: According to the comprehensive learning effect score of the employee user for the current course classification, when the comprehensive learning effect score is lower than the preset comprehensive learning effect score standard value, after obtaining the corresponding recommended course classification according to the recommendation factors of the employee user for each course classification, reselect another course series different from the current course series from the recommended course classification, and generate training course recommendation information according to the reselected course series.

8. An enterprise employee training and development management system based on an online cloud platform according to claim 7, characterized in that, The status monitoring unit specifically includes: Periodically obtain the real-time status images of the employee user during the process of the employee user learning the training course; Preprocess the obtained real-time status images to improve the quality of the real-time status images; Detect the face area according to the preprocessed real-time status images; Based on the detected face region, a head pose estimation model based on Hopenet is used to extract the head pose feature information of the employee user HZ ; where the head pose feature information includes the yaw angle, pitch angle, and roll angle of the head; An implementation estimation model based on L2CS-Net is used to extract the line-of-sight feature information of employee users YZ , where the line-of-sight feature information includes the yaw angle and pitch angle of the line-of-sight direction; Judge the learning status of the current employee user during the training process based on the obtained head pose feature information and gaze feature information. The learning status score function adopted is: ; Among them, S(a,b) represents the learning status score of an employee user a in the course classification b ; ZP represents the total learning status score of an employee user a in the learning process of the course classification b . The variable n = 1, 2, …, ZP corresponds to the n th capture moment; fsh(n,HZ, HZT) represents the head pose judgment function. When the head pose feature information of the employee user obtained based on the n th capture moment HZ is within the preset head pose standard range HZT , fsh(n, HZ, HZT) = 1 is obtained, otherwise fsh(n, HZ, HZT) = 0 ; fsh(n, YZ, YZT) represents the line-of-sight feature judgment function. When the line-of-sight feature information of the employee user obtained based on the n th capture moment YZ is within the preset line-of-sight standard range YZT , fsh(n, YZ, YZT) = 1 is obtained, otherwise fsh(n, YZ, YZT) = 0 ; ω hz and ω yz are the preset weight factors respectively.

Images