Fatigue reminder methods and devices, electronic devices and readable storage media

Abstract

This invention provides a fatigue reminder method, device, electronic device, and readable storage medium, relating to the field of electronic technology. The method includes: acquiring multiple user features of multiple dimensions corresponding to a target user at multiple preset periods, wherein the multiple user features correspond to different preset periods; inputting the multiple user features into a pre-trained three-branch decision model; and determining the fatigue state of the target user based on the optimal dimension feature using the three-branch decision model, wherein the optimal dimension feature includes the lowest dimension user feature used to determine the fatigue state; and, when the fatigue state is fatigue, providing a fatigue reminder to the target user based on the multiple user features. Therefore, this invention can solve the problems of low accuracy and low recognition speed in related technologies for identifying the fatigue state of end users.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electronics, and in particular to a fatigue reminding method and device, an electronic device, and a readable storage medium. BACKGROUND

[0002] Long-time continuous work in front of a computer can easily cause the worker to have a disorder of psychological function and physiological function, and can cause phenomena such as blurred vision, sore back and waist, slow reaction, and stiff action, and can cause a decline in work function. Rapid and accurate fatigue identification and reminding can not only reduce human errors in work, but also can improve work efficiency.

[0003] At present, the way of judging the fatigue of a worker is mostly a single information source, such as completely using face recognition or being based on an external device. The former is easy to collect images, but existing methods including neural network, template matching, and geometric feature extraction have low recognition accuracy or slow speed. The latter has a high recognition rate, but needs to purchase a device, has high cost, and can cause a psychological burden to a user for long-time wearing. SUMMARY

[0004] Embodiments of the present application provide a fatigue reminding method and device, an electronic device, and a readable storage medium, to solve the problems of low recognition accuracy and low recognition rate of the fatigue state of a terminal user in related technologies.

[0005] To solve the above technical problems, the present application is implemented as follows:

[0006] In a first aspect, the embodiments of the present application provide a fatigue reminding method, which comprises: acquiring a plurality of user features of a plurality of dimensions corresponding to a target user every multiple preset periods, wherein the plurality of user features respectively correspond to different preset periods; inputting the plurality of user features into a three-branch decision model which is pre-trained, and determining a fatigue state corresponding to the target user according to a best dimension feature through the three-branch decision model, wherein the best dimension feature comprises a user feature of a lowest dimension for determining the fatigue state; and in a case where the fatigue state is fatigue, reminding the target user of fatigue according to the plurality of user features.

[0007] Further, the acquiring of the plurality of user features of the plurality of dimensions corresponding to the target user every multiple preset periods, wherein the plurality of user features respectively correspond to different preset periods, comprises: collecting input data of the target user acting on a target terminal and image data of the target user respectively at different preset periods; and performing numerical conversion on the input data and the image data to obtain the plurality of user features.

[0008] Further, the collecting the input data of the target user acting on the target terminal and the image data of the target user in different preset periods respectively comprises: acquiring facial posture image data and facial posture duration of the target user; and acquiring terminal input duration and terminal input frequency of the target user.

[0009] Further, the inputting the plurality of user features into a pre-trained three-way decision model and determining the fatigue state of the target user according to the optimal dimension feature through the three-way decision model comprises: determining the optimal dimension feature according to the plurality of user features in order from low dimension to high dimension; and determining the fatigue state according to the optimal dimension feature.

[0010] Further, the fatigue reminding of the target user according to the plurality of user features in the case that the fatigue state is fatigue comprises: determining a fatigue weighted value of the target user according to the plurality of user features and weights corresponding to the plurality of user features; and fatigue reminding of the target user according to a mode corresponding to the fatigue weighted value.

[0011] Further, after the inputting the plurality of user features into a pre-trained three-way decision model and determining the fatigue state of the target user according to the optimal dimension feature through the three-way decision model, the method further comprises: determining a current optimal dimension feature according to an accuracy and a time value corresponding to the fatigue state, wherein the time value comprises a sum of a first time of acquiring the fatigue state and a second time of collecting the optimal dimension feature; and updating the optimal dimension feature according to the current optimal dimension feature.

[0012] In a second aspect, an embodiment of the present application further provides a fatigue reminding device, the device comprising: an acquisition module, configured to acquire a plurality of user features of a plurality of dimensions corresponding to a target user every multiple preset periods, wherein the plurality of user features correspond to different preset periods respectively; a processing module, configured to input the plurality of user features into a pre-trained three-way decision model, and determine a fatigue state of the target user according to an optimal dimension feature through the three-way decision model, wherein the optimal dimension feature comprises a user feature of the lowest dimension used to determine the fatigue state; and a reminding module, configured to fatigue remind the target user according to the plurality of user features in the case that the fatigue state is fatigue.

[0013] Further, the reminding module comprises: a third determination unit configured to determine a fatigue value corresponding to the target user according to the plurality of user features and the weights corresponding to the plurality of user features; and a reminding unit configured to perform fatigue reminding on the target user according to a manner corresponding to the fatigue value.

[0014] Further, the method further comprises: a determination module configured to input the plurality of user features into a pre-trained three-way decision model, and determine a current optimal dimension feature according to an accuracy and a time value corresponding to the fatigue state after determining the fatigue state of the target user according to the optimal dimension feature by the three-way decision model, wherein the time value comprises a sum of a first time of obtaining the fatigue state and a second time of collecting the optimal dimension feature; and an updating module configured to update the optimal dimension feature according to the current optimal dimension feature.

[0015] In a third aspect, an electronic device is additionally provided, which comprises a memory, a processor, and a computer program stored in the memory and capable of running on the processor, and the computer program, when executed by the processor, implements the steps of the fatigue reminding method according to the first aspect.

[0016] In a fourth aspect, a readable storage medium is additionally provided, and the readable storage medium stores a computer program, and the computer program, when executed by a processor, implements the steps of the fatigue reminding method according to the first aspect.

[0017] In the embodiment of the present application, the plurality of user features corresponding to a plurality of dimensions of the target user are obtained every plurality of preset periods, wherein the plurality of user features correspond to different preset periods respectively; the plurality of user features are input into a pre-trained three-way decision model, and the fatigue state of the target user is determined according to an optimal dimension feature by the three-way decision model, wherein the optimal dimension feature comprises a user feature of the lowest dimension used to determine the fatigue state; and the target user is reminded of fatigue according to the plurality of user features in the case that the fatigue state is fatigue. The optimal dimension feature in the plurality of user features obtained at different times by the three-way decision model determines the fatigue state of the target user, which improves the recognition rate of the fatigue state of the target user, and then the target user is reminded of fatigue according to the plurality of user features. The present application solves the problems of low recognition accuracy and low recognition rate of the fatigue state of the terminal user in the related art.

[0018] The above description is only a summary of the technical solutions of the present application. In order to more clearly understand the technical means of the present application, the embodiments of the present application can be implemented according to the content of the description, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application are described below. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating a fatigue reminder method according to an embodiment of the present invention;

[0021] Figure 2 This is a schematic diagram of the structure of a fatigue reminder device according to an embodiment of the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] Example 1

[0024] According to an embodiment of the present invention, a fatigue reminder method is provided, such as... Figure 1 As shown, the method may specifically include the following steps:

[0025] S102, every multiple preset periods, acquire multiple user features corresponding to multiple dimensions of the target user, wherein the multiple user features correspond to different preset periods;

[0026] In practical applications, the target users are the individuals who actually use the target terminal. Data is collected from these users to obtain multiple user characteristics across various dimensions. Due to the methods and order in which the data is collected from the target users, the collection or acquisition time for different dimensions of user characteristics may vary.

[0027] In this embodiment, each user feature corresponds to a preset period, and different user features are obtained from the target user's behavioral data within different preset periods. For example, facial data of the target user and input data applied to the target terminal are obtained within different preset periods.

[0028] In this embodiment, user data corresponding to the target user is collected sequentially at multiple preset cycles, and the user data is numerically converted to obtain multiple user features of multiple dimensions corresponding to the target user.

[0029] In addition, in the embodiment, the feature dimension corresponding to the user feature can be determined according to the length of the preset period. The feature dimension of the user feature with a short preset period is low, and the feature dimension of the user feature with a long preset period is high.

[0030] Optionally, in the embodiment, the plurality of user features corresponding to the target user are acquired every plurality of preset periods, wherein the plurality of user features correspond to different preset periods respectively, including but not limited to: input data of the target user acting on the target terminal and image data of the target user are collected respectively in different preset periods; and the input data and the image data are converted into values to obtain the plurality of user features.

[0031] Specifically, the input data of the target user acting on the target terminal and the facial data of the target user are collected respectively in different preset periods.

[0032] The input data can be: input data generated by touch operation of the target user on the touchable screen of the target terminal; input data generated by input operation of the target user on the external device of the target terminal, the external device including but not limited to a mouse, a keyboard, a microphone, etc.; input data of the sensor generated by the operation of the target user on the target terminal, the sensor including but not limited to a level meter, a gyroscope, a temperature sensor, etc.

[0033] In the embodiment, the target terminal includes but is not limited to a PC, a mobile terminal, a microcomputer, etc.

[0034] After the input data of the target user acting on the target terminal and the image data of the target user are collected, the input data and the image data are converted into values to obtain the plurality of user features.

[0035] For example, the mouth opening state of the facial posture of the target user is valued, the opening state is 1, and the closing state is 0; the mouth opening state of the eye posture of the target user is valued, the opening state is 1, and the closing state is 0.

[0036] Through the above example, the input data of the target user acting on the target terminal and the image data of the target user are collected in time sequence; the input data and the image data are converted into values to obtain the plurality of user features, so that the plurality of user features of different dimensions at different times can be acquired.

[0037] Optionally, in the embodiment, the input data of the target user acting on the target terminal and the image data of the target user are collected respectively in different preset periods, including but not limited to: facial posture image data and facial posture duration of the target user are acquired; terminal input duration and terminal input frequency of the target user are acquired.

[0038] Specifically, the behavior data of the target user is collected. The behavior data of the target user collected by the target terminal (a notebook computer) is mainly divided into two parts: image data and input data. The image data is facial posture image data obtained by using the camera of the user terminal, and the facial features corresponding to the target user include but are not limited to eye features, mouth features, and face posture features. The input data of the target user is obtained by using the keyboard and mouse of the target terminal, and includes but is not limited to keyboard input time, keyboard input frequency, mouse input time, and mouse input frequency.

[0039] The behavior data of the target user is preprocessed to obtain a series of fatigue feature elements, such as facial posture image data and facial posture duration. The fatigue feature elements include but are not limited to blink frequency, longest interval time between two adjacent blinks, mouth opening height, mouth opening duration, keyboard input time, keyboard input frequency, mouse input time, and mouse input frequency.

[0040] According to the above example, the facial posture image data and the facial posture duration of the target user are obtained, and the terminal input time and the terminal input frequency of the target user are obtained. In the process of using the target terminal by the target user, the user features corresponding to the user are continuously collected to improve the accuracy of recognizing the fatigue state of the user.

[0041] In addition, in this embodiment, the specific way of multi-dimensional expression of the user features of the behavior data of the same time node is as follows:

[0042]

[0043] wherein Xi represents the information of the user features of the target user in different dimensions, X1 represents the information in the low dimension, and Xn represents the information in the high dimension. The value of n is 8, and the high dimension information content is the content containing the low dimension information.

[0044] In this embodiment, the detection period of each user feature is different. For example, the detection period of blinking is relatively short, within 30 seconds, and the detection period of the keyboard is relatively long, within 2 minutes. At a certain time node, there can be 2-dimensional user features, which can be regarded as low-dimensional user features with less dimension; and at a certain time node, there can be 7 or 8-dimensional data, which can be regarded as high-dimensional user features with more dimension.

[0045] According to the above example, a plurality of user features corresponding to the target user in a plurality of dimensions are obtained at different preset periods in chronological order, so as to monitor the user features related to the fatigue state of the target user.

[0046] S104, input multiple user features into a pre-trained three-branch decision model, and determine the fatigue state of the target user based on the optimal dimensional features through the three-branch decision model, wherein the optimal dimensional features include the lowest dimensional user features used to determine the fatigue state.

[0047] In this embodiment, since the acquisition time of different features of the target user is different, it is necessary to determine the fatigue state of the target user based on the best dimension feature among multiple user features in order to achieve rapid identification of the fatigue state of the target user.

[0048] The three-branch decision model in this embodiment can obtain the lowest-dimensional user features that can determine the user's state. In one example, it is assumed that the acquisition time of posture features such as blinking and opening the mouth is earlier than the user features corresponding to the keyboard and mouse input data, and the feature dimension of user features such as blinking and opening the mouth is lower than the feature dimension of the user features corresponding to the input data.

[0049] Then, multiple user features are input into the three-branch decision model. If the best dimension feature is a user feature such as blinking or opening the mouth, then when a user feature such as blinking or opening the mouth is obtained, the fatigue state of the target user can be determined, thus realizing the rapid identification of the fatigue state of the target user.

[0050] The system calculates the conditional probability of classifying employee fatigue into each category from low to high dimensions, and uses three-way decision-making to complete employee fatigue identification.

[0051] In this embodiment, a three-way decision is made based on the low-dimensional features of the target user's user characteristics. If the current dimension of user characteristics can confirm the fatigue state, then the next round of high-dimensional feature extraction and three-way decision-making will not be performed; otherwise, higher-dimensional feature extraction will be performed, and three-way decision-making will be performed based on the higher-dimensional user characteristics until the fatigue state of the target user is determined.

[0052] Before introducing the technical solution of this embodiment, let's first introduce the three decision-making models:

[0053] First, calculate the conditional probability of fatigue classification for target users in each dimension of user characteristics, from low to high dimensions.

[0054] In this embodiment, the conditional probability of classifying the current target user's fatigue state x as j is calculated using a loss function (softmax).

[0055]

[0056] Where l = 1, 2, 3…k, k represents the total number of different dimension categories;

[0057] The three-branch decision model uses α, β, and γ as thresholds for classifying a user into the positive domain (POS), boundary domain (BND), and negative domain (NEG). It accepts the user in the positive domain and rejects them in the negative domain, directly revealing the target user's fatigue state. For the boundary domain, a delayed decision is made to obtain more information at a higher dimension for further application of the three-branch decision model. The expressions for the positive domain (POS), boundary domain (BND), and negative domain (NEG) are as follows:

[0058] POS (α,β) ={x∈U|p(X|[x])≥α}

[0059] BNG (α,β) ={x∈U|β<p(X|[x])<α}

[0060] NEG (α,β) ={x∈U|p(X|[x])≤β}

[0061] The calculation method for αβγ is as follows:

[0062]

[0063]

[0064]

[0065] in, Let X represent the loss functions for acceptance, delay, and rejection respectively when the worker's fatigue state x in the i-th dimension does not belong to category X; These represent the acceptance, delay, and rejection loss functions when the worker's fatigue state x in the i-th dimension belongs to category X, respectively; each loss function is set based on practical experience.

[0066] Next, we set the three decision thresholds in the three-way decision model.

[0067] In this embodiment, the principle for setting the three decision thresholds is to use higher-dimensional judgments only when it is beneficial to the fatigue judgment result. Lower-dimensional judgments will choose larger acceptance thresholds and smaller rejection thresholds, while higher-dimensional judgments will choose smaller acceptance thresholds and larger rejection thresholds. A detailed description follows:

[0068] 0≤β i <α i ≤1, 1≤i≤n, β1≤β2≤...≤β i <α i ≤α2≤α1

[0069] Where i = 1, 2, ..., n-1 represents the sequence from low dimension to high dimension. When dimension i becomes dimension n, it is no longer possible to provide higher-dimensional user features, so the three-way decision becomes a two-way decision, with only acceptance and rejection.

[0070] In this embodiment, the purpose of using a three-way decision-making process is that the acquisition time for each user feature (fatigue information element) is different. For example, the blink cycle is collected within 30 seconds, which can reflect the user's characteristics within 30 seconds; however, the keyboard input frequency requires a longer time to be counted. This is set based on the target user's usage habits and practical experience with the target terminal, and it takes longer to reflect fatigue characteristics. Therefore, if fewer dimensions of user features are acquired and the judgment threshold is reached, a judgment can be made directly; if the three-way decision-making process cannot determine the target user's fatigue state based on the current dimension of user features, it can wait to acquire more dimensions of user features before judging the target user's fatigue state again.

[0071] Therefore, through the above examples, we can improve the real-time performance of fatigue assessment and ensure the accuracy of fatigue assessment for target users.

[0072] S106, when the fatigue state is fatigued, provide fatigue reminders to the target user based on multiple user characteristics.

[0073] In this embodiment, if the target user's fatigue state is determined to be fatigued by the three-branch decision model, the fatigue level of the target user is judged based on multiple user characteristics, and then the target user is given a fatigue reminder based on the fatigue weighted value of the target user.

[0074] Optionally, in this embodiment, when the fatigue state is fatigued, fatigue reminders are given to the target user based on multiple user characteristics, including but not limited to: determining the fatigue weighting value corresponding to the target user based on multiple user characteristics and the weights corresponding to the multiple user characteristics; and giving fatigue reminders to the target user according to the method corresponding to the fatigue weighting value.

[0075] Specifically, a quantitative analysis of the fatigue state of the target users is conducted, and the fatigue weighting value of the target users is calculated using a weighted fatigue method.

[0076] For example, multiple states such as eye state, mouth state, facial pose state, keyboard input state, and mouse input state are logically combined. The fatigue factor value corresponding to each logical combination of states is obtained.

[0077] Finally, the various state logic combinations and their corresponding fatigue factor values ​​are substituted into the preset fatigue judgment formula to obtain the final fatigue weighted value. The fatigue weighted value judgment formula is as follows:

[0078] S=K1*A1+K2*A2+K3*A3+K4*A4+…+Kn*An

[0079] Where n represents the nth fatigue logic condition, Kn represents 1 if the condition is met under condition n, Kn represents 0 if the condition is not met under condition n, and An represents the fatigue factor value corresponding to the nth fatigue logic condition.

[0080] Optionally, in this embodiment, the fatigue state of the target user is determined based on a fatigue weighted value. Within a preset period, the fatigue state is identified using the aforementioned three-branch decision model. Different fatigue states are determined based on the fatigue weighted value, and fatigue reminders are issued to the target user to alleviate or perform a specified operation to eliminate fatigue.

[0081] 1) No reminder is needed when the target user is determined to be in a normal state.

[0082] 2) When the target user is determined to be slightly fatigued, provide auditory and visual reminders.

[0083] 3) When the target user is determined to be moderately fatigued, the instruction for the specified operation is displayed on the screen of the target terminal. After the target user completes the specified operation, the actual fatigue state of the target user is determined based on the specified operation.

[0084] For example, displaying CAPTCHA questions on the target terminal's screen, requiring the target user to complete the CAPTCHA questions, recording the accuracy and timeliness of the answers, and further assessing the actual fatigue level of the event staff.

[0085] 4) If the target user is determined to be severely fatigued, a reminder and fatigue warning will be issued to the target device corresponding to the target user. The reminder may include, but is not limited to, reminders via voice, video, text, or images.

[0086] For example, when a target user is determined to be severely fatigued, a phone call is made to the user with a voice reminder and fatigue warning, stating, "You are currently in a state of severe fatigue. Please get up and move around to relieve your fatigue." A series of body movement diagrams will appear on the target user's screen, and the user needs to follow the diagrams to continue working.

[0087] Meanwhile, during the activity, the target terminal's camera will capture and detect its movements, and the fatigue warning can only be lifted after the specified actions are completed.

[0088] Through the above example, the fatigue value of the target user is determined based on multiple user characteristics and their corresponding weights; fatigue reminders are then given to the target user according to the fatigue value, thus achieving the matching of the target user's fatigue state with the fatigue reminder method.

[0089] In the specific application scenario of this embodiment, the behavioral data of the target user is collected to obtain the feature representation of each dimension of the state, such as the feature representation of eye state, mouth state, facial pose state, keyboard input state, and mouse input state.

[0090] The indicators of eye condition include, but are not limited to, blink frequency and the longest interval between consecutive blinks. A blink frequency lower than a preset value indicates fatigue and is represented by 1; conversely, a blink frequency higher than a preset value indicates normal function and is represented by 0. A blink frequency lower than a preset value indicates fatigue and is represented by 1; conversely, a blink frequency higher than a preset value indicates normal function and is represented by 0.

[0091] The mouth state indicators include, but are not limited to, mouth opening height and mouth opening duration. A mouth opening height greater than a preset value indicates fatigue, represented by 1; conversely, a mouth opening height less than a preset value indicates normalcy, represented by 0. Similarly, a mouth opening duration less than a preset value indicates fatigue, represented by 1; conversely, a mouth opening duration less than a preset value indicates normalcy, represented by 0.

[0092] The status indicators of keyboard input include, but are not limited to: keyboard input duration and keyboard input frequency. A keyboard input duration below a preset value indicates fatigue and is represented by 1; conversely, it is normal and is represented by 0. Similarly, a keyboard input frequency below a preset value indicates fatigue and is represented by 1; conversely, it is normal and is represented by 0.

[0093] The status indicators of mouse input include, but are not limited to, mouse input duration and mouse input frequency. A mouse input duration below a preset value indicates fatigue and is represented by 1; conversely, it is normal and is represented by 0. Similarly, a mouse input frequency below a preset value indicates fatigue and is represented by 1; conversely, it is normal and is represented by 0.

[0094] The fatigue factor table is preset based on the above state logic combination;

[0095] If the eye state, mouth state, keyboard input state, and mouse input state are all 1, then the fatigue factor of this logical condition is preset to 1.

[0096] If the eye status and mouth status are not both 1, further determine the keyboard input status. If the keyboard input status is normal, then the default fatigue state is 0.1 + (eye status value and mouth status value) / 20. Otherwise, the default fatigue factor under this logic condition is 0.3.

[0097] If the eye status and mouth status are not both 1, further determine the mouse input status. If the mouse input status is normal, then the default fatigue state is 0.1 + (eye status value and mouth status value) / 20. Otherwise, the default fatigue factor under this logic condition is 0.2.

[0098] The above logical conditions and their corresponding fatigue factor values ​​are summarized to form a fatigue factor table of preset state logical combinations as shown in Table 1.

[0099] Table 1:

[0100]

[0101] Assuming behavioral data is collected from the target individual during the detection period, and 5 minutes of behavioral data are collected, the collected state data is assumed to be as follows: blinking state value = 1; longest interval between blinks state value = 1; mouth opening height state value = 0; mouth opening duration state value = 1; keyboard input duration state value = 1; keyboard input frequency state value = 0; mouse input duration state value = 1; mouse input frequency state value = 0.

[0102] By substituting various state logic combinations and their corresponding fatigue factor values ​​into the preset fatigue judgment formula, the fatigue weighted value is obtained. The preset fatigue judgment formula is as follows:

[0103] S=K1*A1+K2*A2+K3*A3+K4*A4+…+Kn*An

[0104] Then the fatigue weighted value for the target user can be obtained as follows:

[0105] S=0*1+0*0.6+1*(0.1+3 / 20)+0*0.4+1*(0.1+3 / 20)=0.5

[0106] The fatigue weighting value for the target user is 0.5. By comparing this fatigue weighting value with the preset values ​​for mild, moderate, and severe fatigue, the fatigue state of the target user can be determined.

[0107] Optionally, in this embodiment, after inputting multiple user features into a pre-trained three-branch decision model, and determining the fatigue state of the target user based on the optimal dimension feature through the three-branch decision model, the process further includes, but is not limited to: determining the current optimal dimension feature based on the accuracy and time value corresponding to the fatigue state, wherein the time value includes the sum of the first time of acquiring the fatigue state and the second time of collecting the optimal dimension feature; and updating the optimal dimension feature based on the current optimal dimension feature.

[0108] Specifically, based on the time of acquiring user characteristics and the accuracy of the target user's fatigue state, the best dimension feature that can be used for fatigue state judgment is obtained, and the user characteristics of this dimension are used as the best dimension feature for the next judgment.

[0109] Q = w*A + T*(1-w)

[0110] Where Q represents the optimal dimensional feature, A represents the accuracy of fatigue state identification, w represents the weight value, and T represents the sum of the first time required to determine the fatigue state of the target user through the three-branch decision model and the second time required to extract multi-dimensional user features.

[0111] The above example demonstrates how calculating the optimal dimensional features can save time and improve the efficiency of identifying the fatigue state of target users.

[0112] Through this invention, multiple user features of multiple dimensions corresponding to a target user are acquired at multiple preset periods, wherein each user feature corresponds to a different preset period. These user features are input into a pre-trained three-branch decision model. The three-branch decision model then determines the fatigue state of the target user based on the optimal dimensional feature, where the optimal dimensional feature includes the lowest-dimensional user feature used to determine the fatigue state. When the fatigue state is identified as fatigue, a fatigue reminder is issued to the target user based on the multiple user features. By using the three-branch decision model to determine the target user's fatigue state from the optimal dimensional feature among the multiple user features acquired at different times, the recognition rate of the target user's fatigue state is improved. Then, a corresponding fatigue reminder is issued to the user based on the multiple user features. This invention solves the problems of low accuracy and low recognition rate in related technologies for identifying the fatigue state of end users.

[0113] Example 2

[0114] This invention provides a detailed description of a fatigue reminder device according to an embodiment of the present invention.

[0115] Reference Figure 2 The diagram shows a structural schematic of a fatigue reminder device according to an embodiment of the present invention.

[0116] The fatigue reminder device of this invention includes: an acquisition module 20, a processing module 22, and a reminder module 24.

[0117] The functions of each module and the interaction between them are described in detail below.

[0118] The acquisition module 20 is used to acquire multiple user features of a target user in multiple dimensions at multiple preset periods, wherein the multiple user features correspond to different preset periods.

[0119] Processing module 22 is used to input the multiple user features into a pre-trained three-branch decision model, and to determine the fatigue state of the target user based on the optimal dimension features through the three-branch decision model, wherein the optimal dimension features include the lowest dimension user features used to determine the fatigue state.

[0120] The reminder module 24 is used to remind the target user of fatigue based on the multiple user characteristics when the fatigue state is fatigued.

[0121] Optionally, in this embodiment, the reminder module 24 includes:

[0122] The third determining unit is used to determine the fatigue value corresponding to the target user based on the plurality of user features and the weights corresponding to the plurality of user features;

[0123] The reminder unit is used to remind the target user of fatigue according to the fatigue value.

[0124] Optionally, in this embodiment, the device further includes:

[0125] The determination module is used to determine the current best dimension feature based on the accuracy and time value corresponding to the fatigue state after the multiple user features are input into a pre-trained three-branch decision model and the fatigue state corresponding to the target user is determined by the three-branch decision model based on the best dimension feature. The time value includes the sum of the first time of acquiring the fatigue state and the second time of acquiring the best dimension feature.

[0126] The update module is used to update the best dimension feature based on the current best dimension feature.

[0127] Furthermore, in this embodiment of the invention, multiple user features corresponding to multiple dimensions of the target user are acquired at multiple preset periods, wherein each user feature corresponds to a different preset period; the multiple user features are input into a pre-trained three-branch decision model; and the fatigue state of the target user is determined by the three-branch decision model based on the optimal dimension feature, wherein the optimal dimension feature includes the lowest dimension user feature used to determine the fatigue state; when the fatigue state is fatigue, fatigue reminders are given to the target user based on the multiple user features. By using the three-branch decision model to determine the fatigue state of the target user from the optimal dimension feature among the multiple user features acquired at different times, the recognition rate of the target user's fatigue state is improved. Then, fatigue reminders are given to the user accordingly based on the multiple user features. This invention solves the problems of low accuracy and low recognition rate of the fatigue state of end users in related technologies.

[0128] Example 3

[0129] Preferably, embodiments of the present invention also provide an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the fatigue reminder method steps described above.

[0130] Optionally, in this embodiment, the memory is configured to store program code for performing the following steps:

[0131] S1, acquire multiple user features of multiple dimensions corresponding to the target user at multiple preset periods, wherein the multiple user features correspond to different preset periods;

[0132] S2, input the multiple user features into a pre-trained three-branch decision model, and determine the fatigue state of the target user based on the optimal dimension features through the three-branch decision model, wherein the optimal dimension features include the lowest dimension user features used to determine the fatigue state.

[0133] S3, when the fatigue state is fatigue, provide fatigue reminders to the target user based on the multiple user characteristics.

[0134] Optionally, specific examples in this embodiment can refer to the examples described in Embodiment 1 above, and will not be repeated here.

[0135] Example 4

[0136] Embodiments of the present invention also provide a readable storage medium. Optionally, in this embodiment, the readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the fatigue reminder method as described in Embodiment 1.

[0137] Optionally, in this embodiment, the readable storage medium is configured to store program code for performing the following steps:

[0138] S1, acquire multiple user features of multiple dimensions corresponding to the target user at multiple preset periods, wherein the multiple user features correspond to different preset periods;

[0139] S2, input the multiple user features into a pre-trained three-branch decision model, and determine the fatigue state of the target user based on the optimal dimension features through the three-branch decision model, wherein the optimal dimension features include the lowest dimension user features used to determine the fatigue state.

[0140] S3, when the fatigue state is fatigue, provide fatigue reminders to the target user based on the multiple user characteristics.

[0141] Optionally, the readable storage medium is also configured to store program code for performing the steps included in the method of Embodiment 1 above, which will not be described again in this embodiment.

[0142] Optionally, in this embodiment, the aforementioned readable storage medium may include, but is not limited to, various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0143] Optionally, specific examples in this embodiment can refer to the examples described in Embodiment 1 above, and will not be repeated here.

[0144] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0145] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0146] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of the present invention.

[0147] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0148] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0149] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0150] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0151] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0152] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0153] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A fatigue reminding method characterized by, The method comprises: acquiring a plurality of user features of a plurality of dimensions corresponding to a target user every plurality of preset periods, wherein the plurality of user features correspond to different preset periods respectively; inputting the plurality of user features into a pre-trained three-way decision model, and determining a fatigue state corresponding to the target user according to optimal dimension features through the three-way decision model, wherein the optimal dimension features include user features of the lowest dimension for determining the fatigue state; in a case where the fatigue state is fatigue, reminding the target user of fatigue according to the plurality of user features; the acquiring of the plurality of user features of a plurality of dimensions corresponding to a target user every plurality of preset periods comprises: collecting input data of the target user acting on a target terminal and image data of the target user respectively at different preset periods; performing numerical conversion on the input data and the image data to obtain the plurality of user features; the collecting of the input data of the target user acting on a target terminal and the image data of the target user respectively at different preset periods comprises: acquiring facial posture image data and facial posture duration of the target user; acquiring terminal input duration and terminal input frequency of the target user; the reminding of the target user of fatigue according to the plurality of user features in a case where the fatigue state is fatigue comprises: determining a fatigue weighted value corresponding to the target user according to the plurality of user features and weights corresponding to the plurality of user features; reminding the target user of fatigue according to a mode corresponding to the fatigue weighted value.

2. The method of claim 1, wherein, after the inputting of the plurality of user features into a pre-trained three-way decision model, and the determining of a fatigue state corresponding to the target user according to optimal dimension features through the three-way decision model, the method further comprises: determining a current optimal dimension feature according to an accuracy and a time value corresponding to the fatigue state, wherein the time value includes a sum of a first time of acquiring the fatigue state and a second time of collecting the optimal dimension feature; updating the optimal dimension feature according to the current optimal dimension feature.

3. A fatigue reminding apparatus characterized by comprising: The device comprises: an acquisition module configured to acquire a plurality of user features of a plurality of dimensions corresponding to a target user every plurality of preset periods, wherein the plurality of user features correspond to different preset periods respectively; a processing module configured to input the plurality of user features into a pre-trained three-way decision model, and determine a fatigue state corresponding to the target user according to optimal dimension features through the three-way decision model, wherein the optimal dimension features include user features of the lowest dimension for determining the fatigue state; a reminding module configured to remind the target user of fatigue according to the plurality of user features in a case where the fatigue state is fatigue; the acquisition of the plurality of user features of a plurality of dimensions corresponding to a target user every plurality of preset periods comprises: Collect input data of the target user acting on a target terminal and image data of the target user at different preset periods respectively; Perform numerical conversion on the input data and the image data to obtain the plurality of user features; The collecting of the input data of the target user acting on a target terminal and the image data of the target user at different preset periods respectively comprises: Obtaining facial posture image data and facial posture duration of the target user; Obtaining terminal input duration and terminal input frequency of the target user; The fatigue reminding of the target user according to the plurality of user features in the case that the fatigue state is fatigue comprises: Determining a fatigue weighted value corresponding to the target user according to the plurality of user features and weights corresponding to the plurality of user features; Fatigue reminding of the target user according to a manner corresponding to the fatigue weighted value.

4. The apparatus of claim 3, wherein, Further comprising: A determining module configured to, after the inputting of the plurality of user features into a pre-trained three-way decision model and the determination of the fatigue state of the target user corresponding to the target user according to the best dimension feature by the three-way decision model, determine a current best dimension feature according to an accuracy corresponding to the fatigue state and a time value, wherein the time value comprises a sum of a first time of obtaining the fatigue state and a second time of collecting the best dimension feature; An updating module configured to update the best dimension feature according to the current best dimension feature.

5. An electronic device, comprising: Comprise: A memory, a processor and a computer program stored on the memory and executable on the processor, wherein the computer program is executed by the processor to implement the steps of the fatigue reminding method according to claim 1 or 2.

6. A readable storage medium characterized by, The computer program is stored on the readable storage medium and is executed by the processor to implement the steps of the fatigue reminding method according to claim 1 or 2.

Images