Cognitive load prediction method, electronic equipment and storage medium

Abstract

The invention discloses a cognitive load prediction method, electronic equipment and a storage medium, and the method comprises the steps: inputting to-be-recognized data into a cognitive load prediction model which comprises a target learning model and a linear regression model, and obtaining a target cognitive load prediction result, the target learning model is used for obtaining a transition cognitive load prediction result according to the to-be-identified data, and the linear regression model is used for outputting a target cognitive load prediction result according to the transition cognitive load prediction result; the target learning model is obtained through training of a training strategy, the training strategy comprises: obtaining training sample data, the training sample data comprising a plurality of first time nodes, a plurality of first eye movement feature data, a plurality of first differential entropy feature data, a plurality of first power spectrum density data and a plurality of cognitive load truth value labels, a plurality of first eye movement feature data and a plurality of first differential entropy feature data; and training the initial learning model by using the training sample data to obtain a target learning model.

Description

Technical Field

[0001] The present application relates to the field of cognitive neuroscience, and in particular to a cognitive load prediction method, an electronic device, and a storage medium. Background Art

[0002] Cognitive load prediction is a key issue in the field of human-computer interaction. It helps promote cognitive intelligence in machines and enables computers to identify, interpret and respond to the cognitive load level of individuals. Cognitive load refers to the part of the task demand that actually occupies the individual's limited cognitive resources. The change in cognitive load level is a physiological process affected by many factors. Experts in psychology, neuroscience and computer science have been conducting in-depth research on cognitive load prediction for a long time. However, the accuracy of current cognitive load prediction technology is not high. Summary of the invention

[0003] The present application aims to propose a cognitive load prediction method, electronic device and storage medium, which can solve the problem of low accuracy of traditional cognitive load prediction technology.

[0004] In a first aspect, an embodiment of the present application provides a cognitive load prediction method, comprising:

[0005] The data to be recognized is input into the cognitive load prediction model to obtain the target cognitive load prediction result, where:

[0006] The cognitive load prediction model includes a target learning model and a linear regression model, wherein the target learning model is used to obtain a transition cognitive load prediction result according to the to-be-recognized data, and the linear regression model is used to output the target cognitive load prediction result according to the transition cognitive load prediction result;

[0007] The target learning model is obtained by training through a training strategy, and the training strategy includes:

[0008] Acquire training sample data, the training sample data comprising a plurality of first time nodes, a plurality of first eye movement feature data, a plurality of first differential entropy feature data, a plurality of first power spectrum density data, and a plurality of cognitive load truth value labels, wherein the plurality of first eye movement feature data, the plurality of first differential entropy feature data, the plurality of first power spectrum density data, and the plurality of cognitive load truth value labels all correspond to the first time node; the first eye movement feature data is obtained from an eye movement signal, the first differential entropy feature data and the first power spectrum density data are obtained from an electroencephalogram signal, the eye movement signal and the electroencephalogram signal are obtained by monitoring experimental participants in a cognitive load induction experiment, and the cognitive load truth value label represents the true value of the cognitive load of the experimental participant in the cognitive load induction experiment;

[0009] The initial learning model is trained using the training sample data to obtain the target learning model.

[0010] According to some embodiments of the present application, the data to be identified includes multiple second time nodes, multiple second eye movement feature data, multiple second differential entropy feature data, and multiple second power spectrum density data, and the multiple second eye movement feature data, the multiple second differential entropy feature data, and the multiple second power spectrum density data all correspond to the multiple second time nodes; the transition cognitive load prediction result is obtained by the following steps:

[0011] Inputting a plurality of the second time nodes and a plurality of the second eye movement feature data into the target learning model to obtain a first cognitive load prediction result;

[0012] Inputting a plurality of the second time nodes and a plurality of the second differential entropy feature data into the target learning model to obtain a second cognitive load prediction result;

[0013] Inputting a plurality of the second time nodes and a plurality of the second power spectrum density data into the target learning model to obtain a third cognitive load prediction result;

[0014] The first cognitive load prediction result, the second cognitive load prediction result and the third cognitive load prediction result are integrated to obtain a transitional cognitive load prediction result.

[0015] According to some embodiments of the present application, the initial learning model includes an XGBoost model, a CatBoost model, and a random forest model, and the step of training the initial learning model using the training sample data to obtain the target learning model includes:

[0016] The XGBoost model, the CatBoost model and the random forest model are trained respectively using the training sample data to obtain the target learning model.

[0017] According to some embodiments of the present application, the step of inputting the plurality of second time nodes and the plurality of second eye movement feature data into the target learning model to obtain a first cognitive load prediction result includes:

[0018] Inputting a plurality of the second time nodes and a plurality of the second eye movement feature data into the XGBoost model to obtain a first intermediate prediction result;

[0019] Inputting a plurality of the second time nodes and a plurality of the second eye movement feature data into the CatBoost model to obtain a second intermediate prediction result;

[0020] Inputting a plurality of the second time nodes and a plurality of the second eye movement feature data into the random forest model to obtain a third intermediate prediction result;

[0021] The first cognitive load prediction result is obtained according to the first intermediate prediction result, the second intermediate prediction result and the third intermediate prediction result.

[0022] According to some embodiments of the present application, the step of inputting the plurality of second time nodes and the plurality of second differential entropy feature data into the target learning model to obtain a second cognitive load prediction result includes:

[0023] Inputting a plurality of the second time nodes and a plurality of the second differential entropy feature data into the XGBoost model to obtain a fourth intermediate prediction result;

[0024] Inputting a plurality of the second time nodes and a plurality of the second differential entropy feature data into the CatBoost model to obtain a fifth intermediate prediction result;

[0025] Inputting a plurality of the second time nodes and a plurality of the second differential entropy feature data into the random forest model to obtain a sixth intermediate prediction result;

[0026] The second cognitive load prediction result is obtained according to the fourth intermediate prediction result, the fifth intermediate prediction result and the sixth intermediate prediction result.

[0027] According to some embodiments of the present application, the step of inputting the plurality of second time nodes and the plurality of second power spectrum density data into the target learning model to obtain a third cognitive load prediction result includes:

[0028] Inputting a plurality of the second time nodes and a plurality of the second power spectrum density data into the XGBoost model to obtain a seventh intermediate prediction result;

[0029] Inputting a plurality of the second time nodes and a plurality of the second power spectrum density data into the CatBoost model to obtain an eighth intermediate prediction result;

[0030] Inputting a plurality of the second time nodes and a plurality of the second power spectrum density data into the random forest model to obtain a ninth intermediate prediction result;

[0031] The third cognitive load prediction result is obtained according to the seventh intermediate prediction result, the eighth intermediate prediction result and the ninth intermediate prediction result.

[0032] According to some embodiments of the present application, the fusing the first cognitive load prediction result, the second cognitive load prediction result, and the third cognitive load prediction result to obtain a transitional cognitive load prediction result includes:

[0033] The first cognitive load prediction result, the second cognitive load prediction result, and the third cognitive load prediction result are concatenated in terms of vector dimension to obtain the transitional cognitive load prediction result.

[0034] According to some embodiments of the present application, the cognitive load truth label is obtained by the following steps:

[0035] Acquiring multiple variables and corresponding multiple response quantities of the cognitive load inducing experiment;

[0036] Calculating partial correlation coefficients between the plurality of response quantities and the plurality of variables to obtain a plurality of target partial correlation coefficients;

[0037] The cognitive load truth label is generated according to a plurality of the target partial correlation coefficients.

[0038] In a second aspect, an embodiment of the present application provides an electronic device, including:

[0039] at least one processor;

[0040] at least one memory for storing at least one program;

[0041] When at least one of the programs is executed by at least one of the processors, the cognitive load prediction method as described above is implemented.

[0042] In a third aspect, an embodiment of the present application provides a computer-readable storage medium, which stores a program executable by a processor, and the program executable by the processor is used to implement the cognitive load prediction method as described above when executed by the processor.

[0043] In the embodiment of the present application, the EEG signal can reflect the electrical activity of the brain in the cognitive process, and the eye movement signal can reflect the processing process of visual information. The combination of the EEG signal and the eye movement signal can reflect the cognitive state from different levels. Therefore, the initial learning model is trained using the first eye movement feature data obtained from the eye movement signal and the first differential entropy feature data and the first power spectrum density data obtained from the EEG signal, so that the obtained target learning model can perform different levels of analysis on the data to be recognized and obtain the transition cognitive load prediction result. The linear regression model performs regression analysis based on the transition cognitive load prediction result and outputs the target cognitive load prediction result. Compared with the traditional cognitive load prediction technology, the accuracy of cognitive load prediction is improved.

[0044] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The present application is further described below with reference to the accompanying drawings and embodiments, wherein:

[0046] Figure 1 This is a flow chart of the training strategy in the embodiment of the present application;

[0047] Figure 2 A flowchart of obtaining a transition cognitive load prediction result in an embodiment of the present application;

[0048] Figure 3 A flowchart of obtaining a first cognitive load prediction result in an embodiment of the present application;

[0049] Figure 4 A flowchart of obtaining a second cognitive load prediction result in an embodiment of the present application;

[0050] Figure 5 A flowchart of obtaining a third cognitive load prediction result in an embodiment of the present application;

[0051] Figure 6 A schematic diagram of the structure of the electronic device provided in this application.

[0052] Reference numerals:

[0053] Electronic device 100 , processor 110 , and memory 120 . DETAILED DESCRIPTION

[0054] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and cannot be understood as limiting the present application.

[0055] In the description of the present application, it should be understood that the descriptions involving orientation, such as the orientation or positional relationship indicated as up, down, etc., are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

[0056] In the description of this application, "a plurality" means more than two. If there is a description of "first" or "second", it is only for the purpose of distinguishing technical features, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0057] In the description of this application, unless otherwise clearly defined, terms such as setting, installing, connecting, etc. should be understood in a broad sense, and technicians in the relevant technical field can reasonably determine the specific meanings of the above terms in this application based on the specific content of the technical solution.

[0058] Refer to the following Figures 1 to 6 A cognitive load prediction method, electronic device, and storage medium according to an embodiment of the present application are described.

[0059] The present application embodiment provides a cognitive load prediction method, including:

[0060] Step S100: Input the data to be recognized into the cognitive load prediction model to obtain the target cognitive load prediction result, where:

[0061] The cognitive load prediction model includes a target learning model and a linear regression model. The target learning model is used to obtain a transition cognitive load prediction result according to the data to be identified, and the linear regression model is used to output a target cognitive load prediction result according to the transition cognitive load prediction result.

[0062] The target learning model is trained through the training strategy, such as Figure 1 As shown, the training strategy includes:

[0063] Step S200: acquiring training sample data, the training sample data comprising a plurality of first time nodes, a plurality of first eye movement feature data, a plurality of first differential entropy feature data, a plurality of first power spectrum density data and a plurality of cognitive load truth value labels, the plurality of first eye movement feature data, the plurality of first differential entropy feature data, the plurality of first power spectrum density data and the plurality of cognitive load truth value labels all correspond to the first time node; the first eye movement feature data is obtained from an eye movement signal, the first differential entropy feature data and the first power spectrum density data are obtained from an electroencephalogram signal, the eye movement signal and the electroencephalogram signal are obtained by monitoring experimental participants in a cognitive load induction experiment, and the cognitive load truth value label represents the cognitive load truth value of the experimental participants in the cognitive load induction experiment;

[0064] Step S300: Using training sample data to train the initial learning model to obtain a target learning model.

[0065] In the embodiment of the present application, the EEG signal can reflect the electrical activity of the brain in the cognitive process, and the eye movement signal can reflect the processing process of visual information. The combination of the EEG signal and the eye movement signal can reflect the cognitive state from different levels. Therefore, the initial learning model is trained using the first eye movement feature data obtained from the eye movement signal and the first differential entropy feature data and the first power spectrum density data obtained from the EEG signal, so that the obtained target learning model can perform different levels of analysis on the data to be recognized and obtain the transition cognitive load prediction result. The linear regression model performs regression analysis based on the transition cognitive load prediction result and outputs the target cognitive load prediction result. Compared with the traditional cognitive load prediction technology, the accuracy of cognitive load prediction is improved.

[0066] It can be understood that the cognitive load induction experiment in step S200 can induce continuous changes in the cognitive load level of the experimental participants, and the target learning model obtained by training the initial learning model using the training sample data obtained from the cognitive load induction experiment can continuously predict the cognitive load.

[0067] It can be understood that the cognitive load induction experiment in step S200 takes the multi-target tracking task as the task object, and aims to study the mechanism of cognitive load level changes caused by task difficulty differences in the visual attention processing process. The multi-target tracking task includes three stages:

[0068] Phase 1: The prompt phase presents several objects with similar surface features (e.g., airplanes) on the screen, some of which are marked as targets by flashing red, while others are distractors;

[0069] Phase 2: The visual markers from the prompt phase disappear, and all objects begin to move randomly and independently, requiring participants to track the targets marked in the prompt phase;

[0070] Phase 3: After the preset tracking time, the movement of the objects stops and the participants are asked to identify which objects are the targets. Their reaction time and recognition accuracy are recorded simultaneously in the background.

[0071] During the cognitive load induction experiment, the EEG signals and eye movement signals of the participants were collected by the equipment, and then the EEG signals and eye movement signals were pre-processed to remove artifacts and interference as much as possible. The EEG signal preprocessing used the EEGLAB toolbox of MATLAB. First, a 0.1-40Hz bandpass filter was used to eliminate high-frequency noise, and then a 48-52Hz bandpass filter was used to eliminate power frequency interference. After that, the Automatic Artifact Removal (AAR) toolbox was used to automatically eliminate eye artifacts. The EEG signals and eye movement signals were synchronized and aligned through the computer system timestamp.

[0072] It can be understood that the multiple first time nodes in step S200 represent trials of multiple consecutive multi-target tracking tasks, and multiple first eye movement feature data, multiple first differential entropy feature data, multiple first power spectrum density data and multiple cognitive load true value labels are all obtained through multiple consecutive multi-target tracking tasks.

[0073] It can be understood that in step S200, the first power spectrum density data is obtained from the EEG signal, specifically:

[0074] The EEG signal is converted into a frequency domain signal, and the frequency domain signal is divided into four frequency bands: δ (1-3Hz); θ (4-7Hz); α (8-12Hz); β (13-30Hz). The preprocessed EEG signal is converted into a frequency domain signal using Fourier transform, which is specifically converted using the formula below:

[0075]

[0076] Where X(w) is the Fourier transform of the EEG signal, j is the prefix of the imaginary number, and w represents the number of 2π;

[0077] Analyze the signals in different frequency domains to obtain the corresponding frequency domain information, and obtain the power spectrum density data of different frequency bands according to the frequency domain information, which is specifically obtained through the following formula:

[0078]

[0079] Among them, Δt, N and fs represent the sampling period, the number of samples in the time period and the sampling frequency respectively, and S(w) represents the power spectrum density of the corresponding frequency band.

[0080] It is understandable that the differential entropy feature is considered to be one of the most effective EEG features for estimating cognitive states. It is calculated by first defining [x, x~G(μ, σ 2 )] represents a random time series following Gaussian distribution, and f(x|μ,σ 2 ) represents its probability density function. The calculation formula is as follows:

[0081]

[0082] The first differential entropy feature data of the EEG signal is obtained according to the following formula:

[0083]

[0084] Among them, h(x|μ,σ 2 ) is the differential entropy characteristic.

[0085] For eye movement signals, eye movement feature data are exported by Tobii Pro Lab software, including the average gaze point X coordinate of both eyes (gaze point X), the average gaze point Y coordinate of both eyes (gaze point Y), the X-axis gaze direction of both eyes (gaze direction X), the Y-axis gaze direction of both eyes (gaze direction Y), and the average pupil diameter (pupil diameter). Subsequently, the static gaze and saccade in the eye movement signal are analyzed from the perspective of information theory. Specifically, the differential entropy features are calculated for the gaze point X, gaze point Y, gaze direction X and gaze direction Y variables. The obtained differential entropy features are then connected with the pupil diameter features to form a 5-dimensional feature vector as the first eye movement feature data.

[0086] In one embodiment of the present application, the data to be identified in step S100 includes a plurality of second time nodes, a plurality of second eye movement feature data, a plurality of second differential entropy feature data, and a plurality of second power spectrum density data, and the plurality of second eye movement feature data, the plurality of second differential entropy feature data, and the plurality of second power spectrum density data all correspond to the plurality of second time nodes; Figure 2 As shown, the transition cognitive load prediction result in step S100 is obtained through the following steps:

[0087] Step S110: inputting a plurality of second time nodes and a plurality of second eye movement feature data into a target learning model to obtain a first cognitive load prediction result;

[0088] Step S120: inputting a plurality of second time nodes and a plurality of second differential entropy feature data into a target learning model to obtain a second cognitive load prediction result;

[0089] Step S130: inputting a plurality of second time nodes and a plurality of second power spectrum density data into a target learning model to obtain a third cognitive load prediction result;

[0090] Step S140: fusing the first cognitive load prediction result, the second cognitive load prediction result and the third cognitive load prediction result to obtain a transitional cognitive load prediction result.

[0091] In this embodiment, the target learning model is used to analyze multiple second eye movement feature data, multiple second differential entropy feature data, and multiple second power spectrum density data, so that according to the multiple second eye movement feature data, the multiple second differential entropy feature data, and the multiple second power spectrum density data, the first cognitive load prediction result, the second cognitive load prediction result, and the third cognitive load prediction result corresponding to the second time node are output respectively, and the first cognitive load prediction result, the second cognitive load prediction result, and the third cognitive load prediction result are integrated to obtain the transitional cognitive load prediction result, which can accurately reflect the cognitive load level.

[0092] In one embodiment of the present application, the initial learning model includes an XGBoost model, a CatBoost model, and a random forest model, and the initial learning model is trained using training sample data to obtain a target learning model, including:

[0093] The training sample data are used to train the XGBoost model, CatBoost model and random forest model respectively to obtain the target learning model.

[0094] In this embodiment, the XGBoost model is an algorithm based on a gradient-enhanced decision tree, which uses second-order derivatives to improve the accuracy of the loss function, regularization terms to prevent tree overfitting, and block storage for parallel computing, making it efficient, flexible, and lightweight. The XGBoost model has been widely used in fields such as data mining and feature engineering. The CatBoost model uses a forgetting tree as a base learner, has fewer parameters, supports categorical variables and higher accuracy, and solves the problem of effective processing of categorical features. CatBoost solves problems such as gradient bias and prediction transfer, thereby improving the accuracy and generalization ability of the target learning model. The random forest model is based on a decision tree as a base learner, introduces random attribute selection during the decision tree training process, and enhances the generalization ability of the target learning model.

[0095] The target learning model is obtained by training the XGBoost model, CatBoost model and random forest model using the training sample data, so that the target learning model can integrate learning, has strong generalization ability and high precision, and improves the prediction ability of cognitive load level.

[0096] It can be understood that since the target learning model integrates the XGBoost model, the CatBoost model and the random forest model, the XGBoost model, the CatBoost model and the random forest model are used to predict the data to be identified respectively to obtain multiple prediction results, and then the multiple prediction results are regressed and analyzed through the linear regression model to obtain accurate target cognitive load prediction results.

[0097] In one embodiment of the present application, the step S110 of “inputting a plurality of second time nodes and a plurality of second eye movement feature data into a target learning model to obtain a first cognitive load prediction result” is further described as follows: Figure 3 As shown, step S110 includes:

[0098] Step S111: inputting a plurality of second time nodes and a plurality of second eye movement feature data into an XGBoost model to obtain a first intermediate prediction result;

[0099] Step S112: inputting the plurality of second time nodes and the plurality of second eye movement feature data into a CatBoost model to obtain a second intermediate prediction result;

[0100] Step S113: inputting the plurality of second time nodes and the plurality of second eye movement feature data into the random forest model to obtain a third intermediate prediction result;

[0101] Step S114: obtaining a first cognitive load prediction result according to the first intermediate prediction result, the second intermediate prediction result and the third intermediate prediction result.

[0102] In this embodiment, the second eye movement feature data is analyzed by using the XGBoost model, CatBoost model and random forest model in the target learning model, respectively, and the XGBoost model, CatBoost model and random forest model respectively output the first intermediate prediction result, the second intermediate prediction result and the third intermediate prediction result, and the first intermediate prediction result, the second intermediate prediction result and the third intermediate prediction result are concatenated to obtain the first cognitive load prediction result. For example, if the first intermediate prediction result is 0.6, the second intermediate prediction result is 0.7, and the third intermediate prediction result is 0.8, then the first cognitive load prediction result is [0.6, 0.7, 0.8].

[0103] In one embodiment of the present application, the step S120 of “inputting a plurality of second time nodes and a plurality of second differential entropy feature data into the target learning model to obtain a second cognitive load prediction result” is further described as follows: Figure 4 As shown, step S120 includes:

[0104] Step S121: inputting a plurality of second time nodes and a plurality of second differential entropy feature data into the XGBoost model to obtain a fourth intermediate prediction result;

[0105] Step S122: inputting the plurality of second time nodes and the plurality of second differential entropy feature data into the CatBoost model to obtain a fifth intermediate prediction result;

[0106] Step S123: inputting the plurality of second time nodes and the plurality of second differential entropy feature data into the random forest model to obtain a sixth intermediate prediction result;

[0107] Step S124: obtaining a second cognitive load prediction result according to the fourth intermediate prediction result, the fifth intermediate prediction result and the sixth intermediate prediction result.

[0108] In this embodiment, the second differential entropy feature data are analyzed respectively using the XGBoost model, CatBoost model and random forest model in the target learning model, and the XGBoost model, CatBoost model and random forest model respectively output the fourth intermediate prediction result, the fifth intermediate prediction result and the sixth intermediate prediction result, and the fourth intermediate prediction result, the fifth intermediate prediction result and the sixth intermediate prediction result are spliced ​​to obtain the second cognitive load prediction result corresponding to multiple second time nodes.

[0109] In one embodiment of the present application, in step S130, “inputting a plurality of second time nodes and a plurality of second power spectrum density data into the target learning model to obtain a third cognitive load prediction result”, for example Figure 5 As shown, including:

[0110] Step S131: inputting a plurality of second time nodes and a plurality of second power spectrum density data into the XGBoost model to obtain a seventh intermediate prediction result;

[0111] Step S132: inputting the plurality of second time nodes and the plurality of second power spectrum density data into the CatBoost model to obtain an eighth intermediate prediction result;

[0112] Step S133: inputting the plurality of second time nodes and the plurality of second power spectrum density data into the random forest model to obtain a ninth intermediate prediction result;

[0113] Step S134: obtaining a third cognitive load prediction result according to the seventh intermediate prediction result, the eighth intermediate prediction result and the ninth intermediate prediction result.

[0114] In this embodiment, the second power spectral density data is analyzed respectively using the XGBoost model, CatBoost model and random forest model in the target learning model, and the XGBoost model, CatBoost model and random forest model respectively output the seventh intermediate prediction result, the eighth intermediate prediction result and the ninth intermediate prediction result, and the seventh intermediate prediction result, the eighth intermediate prediction result and the ninth intermediate prediction result are spliced ​​to obtain the third cognitive load prediction result corresponding to multiple second time nodes.

[0115] In one embodiment of the present application, the step S140 of “fusing the first cognitive load prediction result, the second cognitive load prediction result, and the third cognitive load prediction result to obtain a transitional cognitive load prediction result” includes:

[0116] Step S141: concatenating the first cognitive load prediction result, the second cognitive load prediction result and the third cognitive load prediction result in terms of vector dimensions to obtain a transitional cognitive load prediction result.

[0117] In one embodiment of the present application, the cognitive load truth label is obtained through the following steps:

[0118] Obtain multiple variables and corresponding multiple response quantities of cognitive load induction experiments;

[0119] Calculate the partial correlation coefficients between multiple responses and multiple variables to obtain multiple target partial correlation coefficients;

[0120] Generate cognitive load ground truth labels based on multiple target partial correlation coefficients.

[0121] In this embodiment, the change in the response amount of the cognitive load induced experiment is caused by the change of the variable. By calculating the partial correlation coefficients between multiple response amounts and multiple variables, multiple target partial correlation coefficients are obtained, and the cognitive load true value label is generated according to the multiple target partial correlation coefficients.

[0122] For example, in the multi-target tracking task, the variables related to the difficulty of the task include the number of target objects t, the total number of objects o, the horizontal moving speed of the objects h, and the vertical moving speed of the objects v. In order to consider the impact of different variables on the performance of the participants, the number of target objects C correctly identified by the experimental participants was introduced in the response stage of the cognitive load truth label generation process.

[0123] The specific steps of generating labels involve calculating the partial correlation coefficients between the correct identification number of the target and various variables. The partial correlation coefficient calculation formula is as follows:

[0124]

[0125] In the formula, r 12 represents the correlation coefficient between x1 and x2, r 13 represents the correlation coefficient between x1 and x3, r 23 represents the correlation coefficient between x2 and x3, where the correlation coefficient is calculated using the Pearson correlation coefficient. After expansion, x3, x4, ..., x p After the influence of the above factors, the correlation coefficient between x1 and x2 is:

[0126]

[0127] By substituting C, t, o, h, and v into the formula, the partial correlation coefficient r of C relative to t, o, h, and v can be calculated. t 、r o 、r h 、r v The weight coefficient wt of the number of target objects in the cognitive load truth label is calculated by the following formula:

[0128]

[0129] Among them, m represents the corresponding experimental participant.

[0130] Similarly, the weight coefficient w of the total number of objects in the cognitive load truth label can be calculated by the above formula: o , the weight coefficient w of the object's horizontal moving speed in the cognitive load truth label h , the weight coefficient w of the object's vertical moving speed in the cognitive load truth label v .

[0131] The following formula is used to normalize t, o, h, and v and limit them to the range of [0,1]:

[0132]

[0133] The true value label of the cognitive load of the experimental participants is calculated by the following formula:

[0134] Y (m) =w t (m) ×t+w o (m) ×o+w h (m) ×h+m v (m) ×v,

[0135] By traversing all experimental participants, a matrix of cognitive load truth labels can be obtained.

[0136] In addition, an embodiment of the present application further discloses an electronic device 100, such as Figure 6 As shown, including:

[0137] at least one processor 110;

[0138] At least one memory 120, used to store at least one program;

[0139] When at least one program is executed by at least one processor 110 , the cognitive load prediction method as described above is implemented.

[0140] The electronic device 100 provided in the embodiment of the present application can achieve Figures 1 to 5 The various processes implemented by the method embodiment achieve the same beneficial effects, and to avoid repetition, they will not be described again here.

[0141] In addition, an embodiment of the present application further discloses a computer-readable storage medium, which stores a program executable by a processor. When the program executable by the processor is executed by the processor, it is used to implement the cognitive load prediction method as described above.

[0142] The computer-readable storage medium provided in the embodiment of the present application can realize Figures 1 to 5 The various processes implemented by the method embodiment achieve the same beneficial effects, and to avoid repetition, they will not be described again here.

[0143] It will be appreciated by those skilled in the art that all or some of the steps and systems in the disclosed method above may be implemented as software, firmware, hardware and appropriate combinations thereof. Some physical components or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor or a microprocessor, or may be implemented as hardware, or may be implemented as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or a non-transitory medium) and a communication medium (or a temporary medium). As known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, disk storage or other magnetic storage devices, or any other medium that may be used to store desired information and may be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically embodies computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0144] The embodiments of the present application are described in detail above in conjunction with the accompanying drawings, but the present application is not limited to the above embodiments, and various changes can be made within the knowledge scope of ordinary technicians in the relevant technical field without departing from the purpose of the present application.

Claims

1. A cognitive load prediction method, characterized in that: include: The data to be recognized is input into the cognitive load prediction model to obtain the target cognitive load prediction result, where: The cognitive load prediction model includes a target learning model and a linear regression model, wherein the target learning model is used to obtain a transition cognitive load prediction result according to the to-be-recognized data, and the linear regression model is used to output the target cognitive load prediction result according to the transition cognitive load prediction result; The target learning model is obtained by training through a training strategy, and the training strategy includes: Acquire training sample data, the training sample data comprising a plurality of first time nodes, a plurality of first eye movement feature data, a plurality of first differential entropy feature data, a plurality of first power spectrum density data, and a plurality of cognitive load truth value labels, wherein the plurality of first eye movement feature data, the plurality of first differential entropy feature data, the plurality of first power spectrum density data, and the plurality of cognitive load truth value labels all correspond to the first time node; the first eye movement feature data is obtained from an eye movement signal, the first differential entropy feature data and the first power spectrum density data are obtained from an electroencephalogram signal, the eye movement signal and the electroencephalogram signal are obtained by monitoring experimental participants in a cognitive load induction experiment, and the cognitive load truth value label represents the true value of the cognitive load of the experimental participant in the cognitive load induction experiment; The initial learning model is trained using the training sample data to obtain the target learning model.

2. The cognitive load prediction method according to claim 1, characterized in that: The data to be identified includes a plurality of second time nodes, a plurality of second eye movement feature data, a plurality of second differential entropy feature data, and a plurality of second power spectrum density data, wherein the plurality of second eye movement feature data, the plurality of second differential entropy feature data, and the plurality of second power spectrum density data all correspond to the plurality of second time nodes; the transition cognitive load prediction result is obtained by the following steps: Inputting a plurality of the second time nodes and a plurality of the second eye movement feature data into the target learning model to obtain a first cognitive load prediction result; Inputting a plurality of the second time nodes and a plurality of the second differential entropy feature data into the target learning model to obtain a second cognitive load prediction result; Inputting a plurality of the second time nodes and a plurality of the second power spectrum density data into the target learning model to obtain a third cognitive load prediction result; The first cognitive load prediction result, the second cognitive load prediction result and the third cognitive load prediction result are integrated to obtain a transitional cognitive load prediction result.

3. The cognitive load prediction method according to claim 2, characterized in that: The initial learning model includes an XGBoost model, a CatBoost model, and a random forest model, and the training of the initial learning model using the training sample data to obtain the target learning model includes: The XGBoost model, the CatBoost model and the random forest model are trained respectively using the training sample data to obtain the target learning model.

4. The cognitive load prediction method according to claim 3, characterized in that: The step of inputting the plurality of second time nodes and the plurality of second eye movement feature data into the target learning model to obtain a first cognitive load prediction result includes: Inputting a plurality of the second time nodes and a plurality of the second eye movement feature data into the XGBoost model to obtain a first intermediate prediction result; Inputting a plurality of the second time nodes and a plurality of the second eye movement feature data into the CatBoost model to obtain a second intermediate prediction result; Inputting a plurality of the second time nodes and a plurality of the second eye movement feature data into the random forest model to obtain a third intermediate prediction result; The first cognitive load prediction result is obtained according to the first intermediate prediction result, the second intermediate prediction result and the third intermediate prediction result.

5. The cognitive load prediction method according to claim 3, characterized in that: The step of inputting the plurality of second time nodes and the plurality of second differential entropy feature data into the target learning model to obtain a second cognitive load prediction result includes: Inputting a plurality of the second time nodes and a plurality of the second differential entropy feature data into the XGBoost model to obtain a fourth intermediate prediction result; Inputting a plurality of the second time nodes and a plurality of the second differential entropy feature data into the CatBoost model to obtain a fifth intermediate prediction result; Inputting a plurality of the second time nodes and a plurality of the second differential entropy feature data into the random forest model to obtain a sixth intermediate prediction result; The second cognitive load prediction result is obtained according to the fourth intermediate prediction result, the fifth intermediate prediction result and the sixth intermediate prediction result.

6. The cognitive load prediction method according to claim 3, characterized in that: The step of inputting the plurality of second time nodes and the plurality of second power spectrum density data into the target learning model to obtain a third cognitive load prediction result includes: Inputting a plurality of the second time nodes and a plurality of the second power spectrum density data into the XGBoost model to obtain a seventh intermediate prediction result; Inputting a plurality of the second time nodes and a plurality of the second power spectrum density data into the CatBoost model to obtain an eighth intermediate prediction result; Inputting a plurality of the second time nodes and a plurality of the second power spectrum density data into the random forest model to obtain a ninth intermediate prediction result; The third cognitive load prediction result is obtained according to the seventh intermediate prediction result, the eighth intermediate prediction result and the ninth intermediate prediction result.

7. The cognitive load prediction method according to claim 3, characterized in that: The fusing the first cognitive load prediction result, the second cognitive load prediction result and the third cognitive load prediction result to obtain a transitional cognitive load prediction result includes: The first cognitive load prediction result, the second cognitive load prediction result, and the third cognitive load prediction result are concatenated in terms of vector dimension to obtain the transitional cognitive load prediction result.

8. The cognitive load prediction method according to claim 1, characterized in that: The cognitive load truth label is obtained through the following steps: Acquiring multiple variables and corresponding multiple response quantities of the cognitive load inducing experiment; Calculating partial correlation coefficients between the plurality of response quantities and the plurality of variables to obtain a plurality of target partial correlation coefficients; The cognitive load truth label is generated according to a plurality of the target partial correlation coefficients.

9. An electronic device, characterized in that: include: at least one processor; at least one memory for storing at least one program; When at least one of the programs is executed by at least one of the processors, the cognitive load prediction method according to any one of claims 1 to 8 is implemented.

10. A computer-readable storage medium, characterized in that: A program executable by a processor is stored therein, and when the program executable by the processor is executed by the processor, it is used to implement the cognitive load prediction method according to any one of claims 1 to 8.

Images