WO2023240951A1

WO2023240951A1 - Training method, training apparatus, training device, and storage medium

Info

Publication number: WO2023240951A1
Application number: PCT/CN2022/138186
Authority: WO
Inventors: 张志林; 李胜楠; 杨伟平; 梁栋; 吴景龙
Original assignee: 深圳先进技术研究院
Priority date: 2022-06-13
Filing date: 2022-12-09
Publication date: 2023-12-21
Also published as: CN115171658A

Abstract

A training method, a training apparatus, a training device, and a storage medium, which relate to the technical field of computers. The training method comprises: randomly displaying a perceptual decision-making task, which comprises a visual classification task, an auditory classification task, and a visual-auditory classification task, wherein the visual classification task comprises respectively classifying M first pictures, the auditory classification task comprises respectively classifying N first sounds, and the visual-auditory classification task comprises respectively classifying L visual-auditory stimulation pairs, each visual-auditory stimulation pair comprising a second picture and a second sound corresponding to a target in the second picture; collecting behavior reaction data generated when a user completes the perceptual decision-making task; and determining a training result according to the behavior reaction data, wherein the training result comprises the accuracy of the user completing the perceptual decision-making task. In the training method, training is performed in a multi-channel combination mode by means of visual, auditory and visual-auditory channels, such that the perceptual decision-making capability of an individual can be effectively improved.

Description

Training methods, training devices, training equipment and storage media

Technical field

The present application relates to the field of computer technology, and in particular, to a training method, device, equipment and storage medium.

Background technique

Individuals will experience a certain degree of cognitive decline during the aging process. Alzheimer's disease (AD) is a neurodegenerative disease with an insidious onset and progressive development. Clinical manifestations include comprehensive dementia such as memory impairment, aphasia, apraxia, agnosia, impairment of visuospatial skills, executive dysfunction, and changes in personality and behavior.

For the elderly, Alzheimer's disease will reduce their perceptual decision-making ability, causing them to perform poorly on decision-making tasks that require the investment of perceptual abilities, attention resources, memory and other high-level cognitive abilities.

technical problem

Most of the existing research on the perceptual decision-making of patients with Alzheimer's disease only focuses on a single visual channel, such as training individuals at the visual level. However, this single-channel training is too limited to effectively improve an individual's perceptual decision-making ability.

Technical solutions

In view of this, this application provides training methods, training devices, training equipment and storage media, which can effectively improve an individual's perceptual decision-making ability.

In a first aspect, this application provides a training method, including: randomly displaying a perceptual decision-making task. The perceptual decision-making task includes a visual classification task, an auditory classification task, and a visual and auditory classification task. The visual classification task includes classifying M first pictures respectively. Classification, the auditory classification task includes classifying N first sounds respectively, the audio-visual classification task includes classifying L audio-visual stimulus pairs, each of the audio-visual stimulus pairs includes a second picture and a target in the second picture The corresponding second sound, where M≥2, N≥2, L≥2; collect the behavioral response data generated by the user when completing the perceptual decision-making task; determine the training results based on the behavioral response data, and the training results include the user The accuracy of completing this perceptual decision-making task.

In a possible implementation, the behavioral response data includes classification results corresponding to each classification task in the perceptual decision-making task and reaction times for completing each classification task. The training method also includes: inputting the classification results and reaction times into a preset drift Process in the diffusion model to obtain the drift rate, decision boundary and non-decision time; based on the drift rate, decision boundary and non-decision time, the user's perceptual decision-making ability is evaluated.

In a possible implementation, the training method also includes: determining the user's health status based on the user's perceptual decision-making ability.

In a possible implementation, the training method also includes: obtaining M preset pictures; adjusting the basic attributes of each preset picture to obtain M first pictures; and constructing the visual classification task based on the M first pictures. .

In a possible implementation, the training method further includes: obtaining N preset sounds; adjusting the sound attributes of each preset sound to obtain N first sounds; and constructing an auditory classification task based on the N first sounds.

In a possible implementation, the training method further includes: determining L second pictures among the M first pictures; determining L second sounds among the N first sounds; comparing the L second pictures and L Pair the second sounds to obtain L audio-visual stimulus pairs; construct an audio-visual classification task based on the L audio-visual stimulus pairs.

In a possible implementation, the training method further includes: determining the stimulation intensity corresponding to each of the first pictures and each of the first sounds, and the stimulation intensity is used to reflect each of the first pictures. and the corresponding accuracy rate when each first sound is classified;

Select the picture whose stimulation intensity is the first stimulation intensity and the picture whose stimulation intensity is the second stimulation intensity among the M first pictures;

Select the sound whose stimulation intensity is the first stimulation intensity and the sound whose stimulation intensity is the second stimulation intensity among the N first sounds;

Construct a perceptual decision-making task of the first stimulation intensity based on the picture of the first stimulation intensity and the sound of the first stimulation intensity;

Based on the picture of the second stimulation intensity and the sound of the second stimulation intensity, a perceptual decision-making task of the second stimulation intensity is constructed.

In a second aspect, this application provides a training device, including:

The display unit is used to randomly display perceptual decision-making tasks. The perceptual decision-making tasks include visual classification tasks, auditory classification tasks and visual and auditory classification tasks. The visual classification tasks include classifying the M first pictures respectively, and the auditory classification tasks include classifying the N first pictures. Sounds are classified separately. The audio-visual classification task includes classifying L audio-visual stimulus pairs respectively. Each audio-visual stimulus pair includes a second picture and a second sound corresponding to the target in the second picture, where M≥2, N≥2 ,L≥2;

The collection unit is used to collect behavioral response data generated by users when completing perceptual decision-making tasks;

The determination unit is used to determine the training results based on the behavioral response data, and the training results include the accuracy of the user completing the perceptual decision-making task.

In a third aspect, the present application provides a training device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, any one of the above aspects in the first aspect is implemented. The training method described in one method.

In a fourth aspect, the present application provides a computer-readable storage medium that stores a computer program. When the computer program is executed by a processor, the training method described in any of the above-mentioned first aspects is implemented.

In a fifth aspect, embodiments of the present application provide a computer program product. When the computer program product is run on a processor, it causes the processor to execute the training method described in any of the above-mentioned first aspects.

beneficial effects

The training method provided by this application randomly displays perceptual decision-making tasks to users, and trains users based on the perceptual decision-making tasks. During the training process, the behavioral response data generated by the user when completing the perceptual decision-making task is collected. Based on the behavioral response data, the training results can be determined, such as determining the accuracy of the user in completing the perceptual decision-making task. Since this perceptual decision-making task includes classification tasks on multiple channels of vision, hearing, and visual and auditory, using this perceptual decision-making task to train users can accelerate the user's information storage and encoding in the high-order cognitive process, and can improve The user's reaction speed, in turn, promotes the formation of perceptual decision-making, thereby effectively improving the individual's perceptual decision-making ability.

Description of the drawings

Figure 1 is a schematic flow chart of a training method provided by an exemplary embodiment of the present application;

Figure 2 is a first schematic diagram provided by an embodiment of the present application;

Figure 3 is a first sound diagram provided by an embodiment of the present application;

Figure 4 is a schematic diagram of an audio-visual stimulus pair provided by an embodiment of the present application;

Figure 5 is a specific flow chart of a training method according to another exemplary embodiment of the present application;

Figure 6 is a specific flow chart of a training method according to yet another exemplary embodiment of the present application;

Figure 7 is a schematic diagram of a training device provided by an embodiment of the present application;

Figure 8 is a schematic diagram of a training device provided by another embodiment of the present application.

Embodiments of the invention

In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments These are part of the embodiments of this application, but not all of them. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of this application.

Perceptual decision-making is a continuous hierarchical cognitive operation that converts sensory information into goal-oriented and responds, including encoding and accumulation of decision-making from sensory information (such as information generated by objective things that directly act on sensory organs) Information is applied to make decisions using decision rules, culminating in behavioral responses. For example, the user sees a picture, determines that the content in the picture is an animal, and selects the animal option among the preset options. This entire process is called perceptual decision-making.

In view of this, this application provides a training method, training device, training equipment and storage medium. By randomly presenting perceptual decision-making tasks to users, the users are trained based on the perceptual decision-making tasks. During the training process, the behavioral response data generated by the user when completing the perceptual decision-making task is collected. Based on the behavioral response data, the training results can be determined, such as determining the accuracy of the user in completing the perceptual decision-making task. Since this perceptual decision-making task includes classification tasks on multiple channels of vision, hearing, and visual and auditory, using this perceptual decision-making task to train users can accelerate the user's information storage and encoding in the high-order cognitive process, and can improve The user's reaction speed, in turn, promotes the formation of perceptual decision-making, thereby effectively improving the individual's perceptual decision-making ability.

The technical solution of the present application will be described in detail below with specific examples. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

The embodiment of this application provides training software. The training software can be installed in a training device, which can be a device that can display pictures and have audio playback functions, such as smartphones, tablets, desktop computers, laptops, robots, smart wearables and other devices. The training software provided by this application can not only train users, but also test the user's perceptual decision-making ability before or after training.

Please refer to Figure 1, which is a schematic flow chart of a training method provided by an exemplary embodiment of the present application. The training method as shown in Figure 1 may include: S101~S103, specifically as follows:

S101: Random presentation of perceptual decision-making tasks.

Perceptual decision-making tasks include visual classification tasks, auditory classification tasks, and visual and auditory classification tasks.

The visual classification task includes classifying M first images respectively, M≥2. Among them, M represents the number of the first picture. For example, M can be a positive integer greater than or equal to 2. The first picture may be a picture containing any object. For example, the first picture may be a picture containing faces, a picture containing cars, a picture containing animals, a picture containing plants, a picture containing buildings, a picture containing food, or containing daily necessities. of images, images containing electronic devices, images containing musical instruments, etc. Different types of first pictures can be added according to actual training needs. This is only an illustrative description without limitation.

Please refer to Figure 2. Figure 2 is a first schematic diagram provided by an embodiment of the present application. As shown in Figure 2, Figure 2 shows a first picture in the visual classification task, which is a picture containing a face.

There are no restrictions on the channels through which the first image can be obtained. For example, the first picture may be obtained by taking a photo, may be collected from the Internet, may be obtained by painting, etc.

The auditory classification task involves classifying N first sounds respectively, N≥2. Wherein, N represents the number of first sounds, for example, N can be a positive integer greater than or equal to 2. The first sound may be audio containing any sound. For example, the first sound may be audio containing the sound of a person, audio containing the sound of a car, audio containing the sound of an animal, audio containing the sound of an electronic device, audio containing the sound of an instrument, etc. Different types of first sounds can be added according to actual training needs. This is only an exemplary description and is not limited.

Please refer to Figure 3. Figure 3 is a first sound diagram provided by an embodiment of the present application. As shown in Figure 3, Figure 3 shows a first sound in the auditory classification task. The first sound is an audio containing a character's voice, specifically an audio containing a little girl's voice.

There are no restrictions on the channels through which First Voice can be obtained. For example, the first sound may be obtained by recording, or may be collected from the Internet, etc.

The audio-visual classification task includes classifying L audio-visual stimulus pairs respectively, each audio-visual stimulus pair includes a second picture and a second sound corresponding to the target in the second picture, L≥2. Among them, L represents the number of audio-visual stimulus pairs. For example, L can be a positive integer greater than or equal to 2.

The second picture may be a picture containing any object. For example, the second picture may be a picture containing faces, a picture containing cars, a picture containing animals, a picture containing musical instruments, etc. Correspondingly, the second sound may be audio containing human voices, audio containing car sounds, audio containing animal sounds, audio containing musical instrument sounds, etc.

For example, a picture of a face and the audio of the sound corresponding to the face form an audio-visual stimulus pair, the picture of a car and the audio of the sound corresponding to the car form an audio-visual stimulus pair, and the picture of an animal and the sound corresponding to the animal form an audio-visual stimulus pair. The audio of the instrument forms an audio-visual stimulus pair, and the audio containing the picture of the musical instrument and the audio corresponding to the sound of the instrument forms an audio-visual stimulus pair.

Please refer to Figure 4, which is a schematic diagram of an audio-visual stimulus pair provided by an embodiment of the present application. As shown in Figure 4, Figure 4 shows an audio-visual stimulus pair in the audio-visual classification task. The audio-visual stimulus pair includes a second picture and a second sound corresponding to the target in the second picture. The second picture is a picture containing a car, and the second sound is an audio containing a car sound, specifically an audio containing a car horn.

There are no restrictions on the channels for obtaining audio-visual stimulation pairs. For example, the second picture in the audio-visual stimulus pair can be selected from the first picture, or it can be re-photographed, or it can be collected from the Internet, or it can be obtained through painting. The second sound in the audio-visual stimulus pair can be selected from the first sound, can be recorded, or can be collected from the Internet.

In one possible implementation, after the user starts the training software installed on the training device, selects a training option, and the perceptual decision-making task starts to be displayed randomly in the display interface of the training device. Among them, when users select training options, they can click manually, operate remotely, or use voice control.

For example, a gaze point is presented in the center of the display interface of the training device. The presentation duration of the gaze point can be set by oneself, for example, it can be set to 2000 ms, and then the visual classification task, the auditory classification task, and the visual and auditory classification tasks are randomly displayed. One way to display it can be to display one task and then another task until all tasks are displayed. For example, the visual classification task is shown first. After all the M first pictures in the visual classification task are shown, the auditory classification task is shown. After the N first sounds in the auditory classification task are shown, the visual and auditory tasks are shown again. Classification task until all L audio-visual stimulus pairs in the audio-visual classification task are displayed.

It is worth noting that there is no limit to the order in which visual classification tasks, auditory classification tasks, and visual and auditory classification tasks are presented. For example, the display order can be a visual classification task, an auditory classification task, a visual and auditory classification task, or it can be a visual classification task, a visual and auditory classification task, an auditory classification task, or it can be a visual and auditory classification task, a visual classification task, an auditory classification task. Wait, there is no limit to this.

For example, another display method may be to intersperse the visual classification task, the auditory classification task, and the audio-visual classification task, that is, the M first pictures, the N first sounds, and the L audio-visual stimulus pairs are interspersed until all The task is displayed. For example, first display several first pictures, then display several audio-visual stimulus pairs, then display several first sounds, then display several first pictures, then display several sounds, and so on until all tasks are displayed.

For another example, first display a first picture, then display a first sound, then display an audio-visual stimulus pair, then display a first sound, then display a first picture, and so on. This is only an illustrative description without limitation.

It is worth mentioning that in order to ensure the validity of the training results, when displaying the first picture in the visual classification task and the second picture in the visual and auditory classification task, these pictures are presented on a uniform color background, and the presented The perspective is the same. For example, they are all presented on gray, white and other backgrounds, and the presented viewing angles are 8°×8°. This is only an illustrative description without limitation.

S102: Collect the user's behavioral response data when completing the perceptual decision-making task.

For example, in the display interface of the training device, in addition to displaying each perceptual decision-making task, options corresponding to each perceptual decision-making task are also displayed. In the process of randomly displaying each perceptual decision-making task, the user makes a choice for each classification task. Operations, the data generated by these selection operations are behavioral response data, and the training device collects these behavioral response data.

For example, the first picture in the visual classification task is displayed in the current display interface, and two options are displayed side by side below, above, right, or left of the first picture. When the first picture is a picture containing a face, the correct choice is to click the left option among the two options displayed side by side; when the first picture is a picture containing a car, the correct choice is to click the right option among the two options displayed side by side. options.

For another example, the current training device plays the first sound in the auditory classification task, and the display interface displays two options side by side. When the first sound is a human voice, the correct choice is to click the left option to display the two options side by side; when the first sound is the sound of a car, the correct choice is to click the right option to display the two options side by side. It is worth mentioning that during the training process, the distance between the user and the training equipment can be set and adjusted by himself. For example, the user is 60 cm away from the display interface and speakers.

During the training process, users make different choices for different perceptual decision-making tasks according to their own abilities, that is, they choose the option that the user thinks is correct. The training device records the choices made by the user for each classification task.

In a possible implementation, the user makes a selection operation for each classification task, which can be achieved through a mouse. For example, for the first picture containing a face, the first sound containing a human voice, the second picture containing a face, and the second sound corresponding to the second picture containing a human voice, the correct choice for these classification tasks is to click the left mouse button. key. For the first picture containing the car, the first sound containing the sound of the car, the second picture containing the car, and the second sound corresponding to the second picture containing the sound of the car, the correct choice for these classification tasks is to click the right mouse button.

For example, what is currently displayed in the display interface is the second picture in the visual and auditory classification task, and the training device plays the second sound corresponding to the second picture. The second picture is a second picture containing a face, the second sound is a second sound containing a character's voice, and the correct selection is to click the left mouse button.

During the training process, users make different choices for different perceptual decision-making tasks according to their own abilities, that is, the user clicks the left or right mouse button. The training device records the choices made by the user for each classification task.

Optionally, in a possible implementation, in order to ensure the effectiveness of training, when displaying the perceptual decision-making task, two adjacent classification tasks can be displayed based on a preset time interval, and the display duration of each classification task is is the default duration. For example, the preset time interval between two adjacent classification tasks can be 1200 to 1500ms, and the display duration of each classification task can be 300ms.

For example, the preset time interval between two adjacent pairs of audio-visual stimuli can be 1200-1500 ms, and the display duration of each pair of audio-visual stimuli can be 300-500 ms. This is only an illustrative description without limitation.

S103: Determine training results based on behavioral response data.

The training results include the accuracy of users completing perceptual decision-making tasks.

For example, the behavioral response data is data generated by the user making a selection operation for each classification task. The behavioral response data corresponding to each classification task is compared with the correct choice corresponding to the task, and the training result is determined based on the comparison result.

For example, one point is scored for each correct choice, and no points are scored for no choice or wrong choice. A score can be obtained based on the user's behavioral response data, and the proportion of this score to the total score (the corresponding score when all tasks are selected correctly) is calculated, and we get The accuracy of users completing perceptual decision-making tasks. This is only an illustrative description without limitation.

In this embodiment, the perceptual decision-making task is randomly displayed to the user, and the user is trained based on the perceptual decision-making task. During the training process, the behavioral response data generated by the user when completing the perceptual decision-making task is collected. Based on the behavioral response data, the training results can be determined, such as determining the accuracy of the user in completing the perceptual decision-making task. Since this perceptual decision-making task includes classification tasks on multiple channels of vision, hearing, and visual and auditory, using this perceptual decision-making task to train users can accelerate the user's information storage and encoding in the high-order cognitive process, and can improve The user's reaction speed, in turn, promotes the formation of perceptual decision-making, thereby effectively improving the individual's perceptual decision-making ability.

Please refer to Figure 5. Figure 5 is a specific flow chart of a training method according to another exemplary embodiment of the present application. The training method shown in Figure 5 may include: S201~S205, specifically as follows:

S201: Random presentation of perceptual decision-making tasks.

S202: Collect the user's behavioral response data when completing the perceptual decision-making task.

S203: Determine training results based on behavioral response data.

The above S201 ~ S203 are exactly the same as S101 ~ S103 in the embodiment corresponding to FIG. 1 . For details, refer to the description of S101 ~ S103 in the embodiment corresponding to FIG. 1 , which will not be described again here.

Behavioral response data include the classification results corresponding to each classification task in the perceptual decision-making task and the reaction time to complete each classification task.

For example, the classification results corresponding to each classification task are the selection operations made by the user. For example, the user clicks on the left option of two options displayed side by side, the user clicks on the right option of two options displayed side by side, the user clicks the left button of the mouse, and the user clicks the right button of the mouse.

The reaction time to complete each category task was determined by the time when each category task was initially presented and the time when the user made a selection. For example, for a certain classification task, the timing starts when the classification task is displayed and ends immediately after the user makes a selection. The recorded time is the reaction time corresponding to the classification task.

S204: Input the classification results and reaction time into the preset drift diffusion model for processing, and obtain the drift rate, decision boundary and non-decision time.

The preset drift-diffusion model simulates the decision-making process in the classification task. Each choice of the user is represented as an upper boundary and a lower boundary. The perceptual decision-making process continues to accumulate evidence over time until it reaches one of the two boundaries. , and then trigger corresponding behavioral responses.

The drift rate, decision boundary and non-decision time are different parameters obtained by processing the classification results and reaction time of the drift diffusion model. These different parameters respectively map the cognitive processing behind the behavior of the perceptual decision-making process. Specifically, the drift rate is used to describe the speed at which information is accumulated, the decision boundary is used to describe the response boundary that needs to be reached before a response is made, and the non-decision time is used to describe the time of sensory encoding and motor response.

The distribution of different responses will affect the values of each parameter in the drift-diffusion model. Therefore, the specific parameters of the drift-diffusion model can be calculated under different situations to reflect the user's potential cognitive process in the cross-channel perceptual decision-making process, thereby determining the user's training effect.

In a possible implementation, in order to prevent users from quickly guessing and determining the classification results, which will lead to bias in the training results, before inputting the classification results and reaction times into the preset drift diffusion model for processing, first eliminate those with reaction times smaller than Preset reaction time data. For example, eliminate data with response time less than 300ms.

Optionally, after eliminating data whose reaction time is less than the preset reaction time, the standard deviation can also be calculated based on all remaining reaction times, and then eliminating data whose reaction time exceeds the preset standard deviation range. For example, data whose reaction times exceed plus or minus 2.5 standard deviations are eliminated. This is only an illustrative description without limitation.

The functions applied to the preset drift-diffusion model are as follows:

(1)

In the above formula (1), f(t) is the conditional probability distribution about t. According to Bayes’ theorem, the function can be

It is split into two parts: f(t) prior and f(t) likelihood. The prior refers to the user's subjective guess of the probability distribution without knowing the parameters of the drift-diffusion model, while the likelihood refers to the calculated parameters of the drift-diffusion model when the probability distribution of the behavioral response data is obtained.

Therefore, the focus of the drift-diffusion model is to find the parameter values under likelihood conditions. Due to the complexity of the formula, the parameter values cannot be obtained directly, so the Markov chain Monte Carlo algorithm (Markov chain Monte Carlo algorithm) needs to be used. Chain Monte Carlo, MCMC). The MCMC algorithm can obtain function characteristics through continuous sampling, thereby inferring the parameters of the population through samples. Therefore, the likelihood part in Bayesian is calculated through the MCMC algorithm to estimate the parameter distribution.

For example, the HDDM toolbox of a computer programming language (python) can be used, which provides hierarchical Bayesian parameter estimation of the drift-diffusion model, allowing the drift-diffusion model parameters of each subject to be estimated simultaneously, thereby obtaining the drift rate, decision boundaries and non-decision time.

Optionally, input the classification results and reaction time into the preset drift diffusion model for processing. In addition to the drift rate, decision boundary and non-decision time, the relative starting point, the inter-training variation of the relative starting point, and the drift rate can also be obtained. Parameters such as inter-training variation and inter-training variation of non-decision time.

Among them, the relative starting point is used to describe the starting preference for response selection. Inter-session variation in relative starting points is expressed as a uniformly distributed range of mean relative starting points, describing the distribution of actual starting points for a particular training session. The inter-training variation in drift rate is expressed as the standard deviation of a normal distribution, and the mean is the drift rate, which is used to describe the actual drift rate distribution for a specific training. Inter-training variation in non-decision time is represented by a uniformly distributed range of mean non-decision time and is used to describe the distribution of actual non-decision time in training.

S205: Evaluate the user's perceptual decision-making ability based on the drift rate, decision boundary and non-decision time.

For example, parameters such as drift rate, decision boundary, and non-decision time each correspond to different indicator ranges. For example, the indicator range corresponding to the drift rate can be greater than -5 and less than 5, the indicator range corresponding to the decision boundary can be greater than 0.5 and less than 2, and the indicator range corresponding to the non-decision time can be greater than 0.1 and less than 0.5.

For example, if the user's drift rate, decision boundary, and non-decision time are all within the respective corresponding indicator ranges, the user's perceptual decision-making ability is evaluated to be strong. If two of the user's drift rate, decision boundary, and non-decision time are within their respective corresponding indicator ranges, the user's perceived decision-making ability is assessed to be moderate. If any of the user's drift rate, decision boundary, and non-decision time is within the corresponding indicator range, or if the user's drift rate, decision boundary, and non-decision time are not within the corresponding indicator range, the user's perceived decision-making ability is assessed to be poor. . This is only an illustrative description without limitation.

In this implementation, the user's classification results and reaction time are processed through a preset drift diffusion model to obtain the drift rate, decision boundary and non-decision time. Through parameters such as drift rate, decision boundary and non-decision time, the user's potential cognitive process in the process of cross-channel perception decision-making can be accurately reflected. Then the drift rate, decision boundary and non-decision time can be analyzed accurately to accurately evaluate The user’s perceptual decision-making ability.

Please refer to Figure 6. Figure 6 is a specific flow chart of a training method according to yet another exemplary embodiment of the present application. The training method shown in Figure 6 may include: S301~S306. It is worth noting that S301 ~ S305 in this embodiment are exactly the same as S201 ~ S205 in the embodiment corresponding to Figure 5. For details, refer to the description of S201 ~ S205 in the embodiment corresponding to Figure 5. There are no differences in this embodiment. Again. S306 details are as follows:

S306: Determine the user's health status based on the user's perceptual decision-making ability.

For example, some conditions can reduce a user's perceptual decision-making ability. For example, in older adults, Alzheimer's disease can reduce their perceptual decision-making abilities. Obtain the perceptual decision-making ability of healthy users and use the perceptual decision-making ability of healthy users as the benchmark. Compare the user's perceptual decision-making ability obtained during this training with the perceptual decision-making ability of healthy users, and determine the health status of the user in this training based on the comparison results.

For example, if a healthy user has strong perceptual decision-making ability, but the user's perceptual decision-making ability obtained during this training process is poor, the health state of the user in this training is determined to be an unhealthy state. Specifically, it can be determined that the user of this training is a patient with Alzheimer's disease. This is only an illustrative description without limitation.

In this embodiment, by comparing the user's perceptual decision-making ability in this training with the perceptual decision-making ability of a healthy user, the user's health status can be accurately determined. For example, it is beneficial to accurately and timely detect Alzheimer's disease patients so that Alzheimer's disease patients can be treated as early as possible.

Optionally, in a possible implementation, before randomly displaying the perceptual decision-making task, the training method provided by this application may also include: obtaining M preset pictures; adjusting the basic attributes of each preset picture to obtain M first pictures; construct a visual classification task based on M first pictures.

For example, the default picture refers to the original first picture. Basic attributes can include the spatial frequency, contrast, brightness, pixels, size, clarity, format, etc. of the image. For example, get several preset images, half of which contain faces and the other half of which contain cars. Adjust the spatial frequency, contrast, brightness and pixels of these images to be consistent. For example, you can adjust the pixels to 670 × 670 pixels.

After adjusting the spatial frequency, contrast, brightness and pixels to be consistent, use preset software (such as Matlab software) to adjust the clarity of each picture through the signal-to-noise ratio. For example, the clarity of each picture is adjusted to 8 different levels of 30%, 32.5%, 35%, 37.5%, 40%, 42.5%, 45%, and 50% through the signal-to-noise ratio.

After the above adjustment, M first pictures are obtained, for example, 240 first pictures are obtained. According to the specific picture content contained in each first picture, set the correct option for each first picture. For example, the correct choice corresponding to the first picture is to click the option on the left of the two options to be displayed side by side, or the first picture corresponds to The correct choice is to click the right option of the two options displayed side by side, or the correct choice corresponding to the first picture is to click the left mouse button, or the correct choice corresponding to the first picture is to click the right mouse button, etc. According to each first picture and the correct option corresponding to each first picture, a visual classification task is constructed.

In this implementation, each first picture obtained has its basic attributes adjusted, which effectively avoids training bias caused by differences in the basic attributes of each picture and ensures that the basic attributes of each picture will not affect the user's choice. , thus improving the accuracy of training results.

Optionally, in a possible implementation, before randomly displaying the perceptual decision-making task, the training method provided by this application may also include: obtaining N preset sounds; adjusting the sound attributes of each preset sound to obtain N First sounds; construct an auditory classification task based on N first sounds.

Illustratively, the preset sound refers to the original first sound. Sound attributes can include frequency, pitch, loudness, timbre, etc. of the sound. For example, get several preset sounds, half of which are human sounds and the other half are car sounds. Adjust the loudness and frequency of these sounds to be consistent. For example, preset software (such as Matlab software) is used to normalize these sounds, and the loudness and frequency of the processed sounds are adjusted to be consistent. Then use speech synthesis software to embed these processed sounds into white noise of different loudness to obtain the first sounds with different signal-to-noise ratios.

For example, the loudness of the processed sound can be reduced to 50%, and speech synthesis software can be used to embed the loudness-adjusted sounds into eight white noises of different loudnesses, resulting in signal-to-noise ratios of 12.5% and 25 respectively. %, 37.5%, 50%, 62.5%, 75%, 87.5% and 100% of multiple first voices. The loudness of these first sounds is consistent, for example, the loudness of these first sounds is 60dB.

Adjust the sharpness of each picture via signal-to-noise ratio. For example, the clarity of each picture is adjusted to 8 different levels of 30%, 32.5%, 35%, 37.5%, 40%, 42.5%, 45%, and 50% through the signal-to-noise ratio.

After the above adjustment, N first sounds are obtained, for example, 240 first sounds are obtained. According to the specific sound content of each first sound, the correct option is set for each first sound. For example, the correct choice for the first sound is to click on the left option of the two options to display side by side, or for the first sound to The correct choice is to click the right option of the two options displayed side by side, or the correct choice for the first sound is to click the left mouse button, or the correct choice for the first sound is to click the right mouse button, etc. According to each first sound and the correct option corresponding to each first sound, an auditory classification task is constructed.

In this implementation, each first sound obtained has adjusted sound attributes, which effectively avoids training deviations caused by differences in the attributes of each sound and ensures that the sound attributes of each sound will not affect the user's choice. This improves the accuracy of training results.

Optionally, in a possible implementation, before randomly presenting the perceptual decision-making task, the training method provided by this application may also include: determining L second pictures among the M first pictures; Determine L second sounds among the sounds; pair L second pictures and L second sounds to obtain L audio-visual stimulus pairs; construct an audio-visual classification task based on the L audio-visual stimulus pairs.

For example, the second picture in the audio-visual stimulus pair may be selected from the first picture, and the second sound in the audio-visual stimulus pair may be selected from the first sound. For example, L first pictures are selected from M first pictures, and these L first pictures are determined as L second pictures. Select L first sounds from the N first sounds, and determine these L first sounds as L second sounds.

It can be understood that since the second sound is the sound corresponding to the target in the second picture, in order to improve the speed of constructing the visual and auditory classification task, when determining the second picture, select the sound corresponding to the target in the second picture. picture. For example, a certain first picture is selected as a picture containing a car among M first pictures, and the sound corresponding to the car happens to be among the N first sounds. The selected first picture is determined as the second picture, and the N first pictures are selected as the second pictures. The sound corresponding to the first sound of the car is determined as the second sound. This is only an illustrative description without limitation.

Pair the selected L second pictures and L second sounds to obtain L audio-visual stimulus pairs. According to each audio-visual stimulus pair, set the correct option for each audio-visual stimulus pair. For example, the correct choice for the audio-visual stimulus pair is to click the left option of the two options to display side by side, or the correct choice for the audio-visual stimulus pair is to click to display side by side. The right option among the two options, or the correct choice of the audio-visual stimulus pair is to click the left mouse button, or the correct choice of the audio-visual stimulus pair is to click the right mouse button, etc. Based on each audio-visual stimulus pair and the correct option for each audio-visual stimulus pair, an audio-visual classification task is constructed.

In this embodiment, the second picture and the second sound in each audio-visual stimulus pair are selected from the first picture and the first sound. Since the basic attributes of the first picture and the sound attributes of the first sound have been adjusted, it is equivalent to the basic attributes of the second picture and the sound attributes of the second sound in each audio-visual stimulus pair being adjusted. It effectively avoids training bias caused by differences in the basic attributes and sound attributes of each picture, ensuring that the basic attributes and sound attributes of each picture will not affect the user's choice, thus improving the accuracy of the training results.

Optionally, in a possible implementation, in order to improve the accuracy of the training results, pre-training may also be included before formal training. Specifically, the training method provided by this application may also include: S401~S405.

S401: Determine the stimulation intensity corresponding to each first picture and each first sound.

S402: Select the picture whose stimulation intensity is the first stimulation intensity and the picture whose stimulation intensity is the second stimulation intensity among the M first pictures.

S403: Select the sound whose stimulation intensity is the first stimulation intensity and the sound whose stimulation intensity is the second stimulation intensity from the N first sounds.

The stimulus intensity is used to reflect the corresponding accuracy of each first picture and each first sound when they are classified.

For example, M first pictures are displayed in the display interface of the training device, the user makes a selection operation on each first picture, and the selection operation corresponding to each first picture is performed with the correct selection corresponding to the first picture. Compare. One point is scored for each correct choice. No points are scored for no selection or wrong selection. A score can be obtained based on all the selection operations of the user this time, and the proportion of this score to the total score (the corresponding score when all the first pictures are selected correctly) is calculated. , to obtain the accuracy of the user's pre-training.

Based on the accuracy rate, the stimulation intensity for the user to select the correct first picture is determined. When the accuracy rate is the first threshold, the user selects the stimulation intensity of the correct first picture as the first stimulation intensity; when the accuracy rate is the second threshold, the user selects the stimulation intensity of the correct first picture as the second stimulation intensity. Wherein, the first threshold is greater than the second threshold, and the first stimulation intensity is higher than the second stimulation intensity. For example, the first threshold is 90%, the second threshold is 70%, the first stimulation intensity is high intensity, and the second stimulation intensity is low intensity.

For example, the accuracy rate of this pre-training is 90%, and this time the user selects the correct stimulation intensity of the first picture as the first stimulation intensity, that is, high intensity. Or, the accuracy rate of this pre-training is 70%, and this time the user selects the correct stimulation intensity of the first picture as the second stimulation intensity, that is, low intensity. This is only an illustrative description without limitation.

For example, N first sounds are displayed in the display interface of the training device, the user makes a selection operation for each first sound, and the selection operation corresponding to each first sound is performed with the correct selection corresponding to the first sound. Compare. One point is scored for each correct choice. No points are scored for no choice or wrong choice. A score can be obtained based on all the selection operations of the user this time, and the proportion of this score to the total score (the corresponding score when all first voice selections are correct) is calculated. , to obtain the accuracy of the user's pre-training.

Based on the accuracy rate, the stimulation intensity of the first sound selected correctly by the user is determined. When the accuracy rate is the first threshold, the user selects the correct stimulation intensity of the first sound as the first stimulation intensity; when the accuracy rate is the second threshold, the user selects the correct stimulation intensity of the first sound as the second stimulation intensity. Wherein, the first threshold is greater than the second threshold, and the first stimulation intensity is higher than the second stimulation intensity. For example, the first threshold is 90%, the second threshold is 70%, the first stimulation intensity is high intensity, and the second stimulation intensity is low intensity.

For example, the accuracy rate of this pre-training is 90%, and this time the user selects the correct stimulation intensity of the first sound as the first stimulation intensity, that is, high intensity. Or, the accuracy rate of this pre-training is 70%, and this time the user selects the correct stimulation intensity of the first sound as the second stimulation intensity, that is, low intensity. This is only an illustrative description without limitation.

S404: Construct a perceptual decision-making task of the first stimulus intensity based on the picture of the first stimulus intensity and the sound of the first stimulus intensity.

The perceptual decision-making tasks of the first stimulus intensity include the visual classification task of the first stimulus intensity, the auditory classification task of the first stimulus intensity, and the visual and auditory classification task of the first stimulus intensity.

It can be understood that the process of constructing the visual classification task of the first stimulus intensity, the auditory classification task of the first stimulus intensity, and the visual and auditory classification task of the first stimulus intensity is different from the above-mentioned construction of the visual classification task, the auditory classification task and the visual and auditory classification. The task process is similar.

The difference is that the above-mentioned construction of the visual classification task, the auditory classification task and the visual and auditory classification task are based on the first picture, the first sound, the second picture and the second sound. In this embodiment, they are based on the first stimulation intensity. The picture and sound construct of the first stimulus intensity are obtained. The specific process can be referred to the above-mentioned process of constructing a visual classification task, an auditory classification task, and a visual and auditory classification task, and will not be described again here.

For example, the constructed visual classification task of the first stimulus intensity contains 50 pictures of the first stimulus intensity, the auditory classification task of the first stimulus intensity contains 50 sounds of the first stimulus intensity, and the visual and auditory classification task of the first stimulus intensity The classification task included 50 pairs of audio-visual stimuli.

S405: Construct a perceptual decision-making task of the second stimulus intensity based on the picture of the second stimulus intensity and the sound of the second stimulus intensity.

The perceptual decision-making tasks of the second stimulus intensity include the visual classification task of the second stimulus intensity, the auditory classification task of the second stimulus intensity, and the visual and auditory classification task of the second stimulus intensity.

It can be understood that the process of constructing the visual classification task of the second stimulus intensity, the auditory classification task of the second stimulus intensity, and the visual and auditory classification task of the second stimulus intensity is different from the above-mentioned construction of the visual classification task, the auditory classification task, and the visual and auditory classification. The task process is similar.

The difference is that the above-mentioned construction of the visual classification task, the auditory classification task and the visual and auditory classification task are based on the first picture, the first sound, the second picture and the second sound. In this embodiment, they are based on the second stimulation intensity. The picture and the sound of the second stimulus intensity are constructed. The specific process can be referred to the above-mentioned process of constructing a visual classification task, an auditory classification task, and a visual and auditory classification task, and will not be described again here.

For example, the constructed visual classification task of the second stimulus intensity contains 50 pictures of the second stimulus intensity, the auditory classification task of the second stimulus intensity contains 50 sounds of the second stimulus intensity, and the visual and auditory task of the second stimulus intensity The classification task included 50 pairs of audio-visual stimuli.

For example, different users (such as normal elderly people and Alzheimer's disease patients) are trained using the constructed perceptual decision-making task of the first stimulation intensity and the perceptual decision-making task of the second stimulation intensity. Collect behavioral response data generated by the user when completing the perceptual decision-making task of the first stimulus intensity and the perceptual decision-making task of the second stimulus intensity; determine the target training results based on the behavioral response data, and the target training results include the perceptual decision-making of the first stimulus intensity Correctness on a perceptual decision-making task of task and secondary stimulus intensity.

In this embodiment, perceptual decision-making tasks with different stimulation intensities are constructed. Perceptual decision-making tasks with different stimulation intensities can be used for training for different users, and the perceptual decision-making abilities of different users can be improved in a targeted manner.

Optionally, in a possible implementation, the training method provided by this application may also include: adjusting the difficulty of the perceptual decision-making task according to the training results, thereby more effectively improving the user's perceptual decision-making ability.

For example, when the user's accuracy in completing a perceptual decision-making task is greater than the preset accuracy rate, it proves that the user's current training effect is good and the difficulty of the perceptual decision-making task can be increased. For example, you can gradually increase the types of pictures and sounds in the perceptual decision-making task, shorten the preset time interval between two adjacent classification tasks, increase the options corresponding to each classification task, etc.

For another example, when the user's accuracy in completing the perceptual decision-making task is less than or equal to the preset accuracy rate, it proves that the user's current training effect is not good, and the difficulty of the perceptual decision-making task can be reduced. For example, the types of pictures and sounds in perceptual decision-making tasks can be reduced, and the preset time interval between two adjacent classification tasks can be increased.

Optionally, in a possible implementation, the training method provided in this application can also use a competitive model to study the impact of cross-channels on patients with Alzheimer's disease. Individuals respond faster when visual and auditory information are presented simultaneously compared to single-channel information (such as visual information or auditory information). This phenomenon is called the redundant signal effect (RSE).

RSE can be explained by statistical facilitation, that is, individuals respond to the single-channel stimulus (visual stimulus, auditory stimulus) that reaches the sensory threshold first among multi-sensory channel stimuli (visual and auditory stimuli), resulting in dual-channel information even though Reactions to stimuli can be accelerated without integration occurring. Through the multi-channel combination training in this application, the individual can first reach the sensory threshold in multi-sensory channel stimulation (visual and auditory stimulation), thereby improving the individual's perceptual decision-making ability.

Figure 7 is a schematic diagram of a training device provided by an embodiment of the present application. As shown in Figure 7, the training device provided by this embodiment includes:

The display unit 510 is used to randomly display perceptual decision-making tasks. The perceptual decision-making tasks include visual classification tasks, auditory classification tasks, and visual and auditory classification tasks. The visual classification tasks include classifying M first pictures respectively. The auditory classification tasks The task includes classifying N first sounds respectively, and the audio-visual classification task includes classifying L audio-visual stimulus pairs, each of the audio-visual stimulus pairs including a second picture and a target corresponding to the second picture. The second sound, where M≥2, N≥2, L≥2;

The collection unit 520 is used to collect behavioral response data generated by the user when completing the perceptual decision-making task;

The determining unit 530 is configured to determine training results according to the behavioral response data, where the training results include the accuracy of the user completing the perceptual decision-making task.

Optionally, the behavioral response data includes classification results corresponding to each classification task in the perceptual decision-making task and reaction times for completing each classification task.

Optionally, the training device also includes:

An evaluation unit is used to input the classification result and the reaction time into a preset drift diffusion model for processing, and obtain the drift rate, decision boundary and non-decision time; according to the drift rate, the decision boundary and the The non-decision time is used to evaluate the user's perceived decision-making ability.

Optionally, the training device also includes:

A state determination unit is configured to determine the health state of the user based on the user's perceptual decision-making ability.

Optionally, the training device also includes:

The first construction unit is used to obtain M preset pictures; adjust the basic attributes of each preset picture to obtain M first pictures; and construct the visual classification task based on the M first pictures.

Optionally, the training device also includes:

The second construction unit is used to obtain N preset sounds; adjust the sound attributes of each preset sound to obtain N first sounds; and construct the auditory classification task based on the N first sounds.

Optionally, the training device also includes:

A third construction unit, configured to determine L second pictures among the M first pictures; determine L second sounds among the N first sounds; and combine the L second pictures and the L second sounds are paired to obtain the L audio-visual stimulus pairs; the audio-visual classification task is constructed based on the L audio-visual stimulus pairs.

Optionally, the training device also includes:

A third construction unit is used to determine the stimulation intensity corresponding to each of the first pictures and each of the first sounds. The stimulation intensity is used to reflect each of the first pictures and each of the first sounds. The corresponding accuracy rate when the sounds are classified; select the picture with the stimulation intensity as the first stimulation intensity and the picture with the stimulation intensity as the second stimulation intensity among the M first pictures; select from the N first sounds The stimulation intensity is the sound of the first stimulation intensity and the stimulation intensity is the sound of the second stimulation intensity; according to the picture of the first stimulation intensity and the sound of the first stimulation intensity, a perceptual decision-making task of the first stimulation intensity is constructed; according to The picture of the second stimulation intensity and the sound of the second stimulation intensity construct a perceptual decision-making task of the second stimulation intensity.

Please refer to FIG. 8 , which is a schematic diagram of a training device provided by another embodiment of the present application. As shown in FIG. 8 , the training device 6 of this embodiment includes: a processor 60 , a memory 61 and a computer program 62 stored in the memory 61 and executable on the processor 60 . When the processor 60 executes the computer program 62, it implements the steps in each of the above training method embodiments, such as S101 to S103 shown in FIG. 1 . Alternatively, when the processor 60 executes the computer program 62, it implements the functions of each unit in the above embodiments, such as the functions of units 510 to 530 shown in FIG. 7 .

Exemplarily, the computer program 62 may be divided into one or more units, and the one or more units are stored in the memory 61 and executed by the processor 60 to complete the present application. The one or more units may be a series of computer instruction segments capable of completing specific functions. The instruction segments are used to describe the execution process of the computer program 62 in the training device 6 . For example, the computer program 62 can be divided into a display unit, a collection unit and a determination unit, and the specific functions of each unit are as described above.

The training device may include, but is not limited to, a processor 60 and a memory 61 . Those skilled in the art can understand that FIG. 8 is only an example of the training device 6 and does not constitute a limitation of the device. It may include more or fewer components than shown in the figure, or combine certain components, or different components, such as The training equipment may also include input and output equipment, network access equipment, buses, etc.

The processor 60 may be a central processing unit (Central Processing Unit). Processing Unit (CPU), or other general-purpose processors, digital signal processors (Digital Signal Processor (DSP), Application Specific Integrated Circuit (Application Specific Integrated Circuit (ASIC), off-the-shelf programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc.

The memory 61 may be an internal storage unit of the training device, such as a hard disk or memory of the device. The memory 61 may also be an external storage terminal of the training device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), or a secure digital card equipped on the training device. Digital, SD) card, flash memory card (Flash Card) etc. Further, the memory 61 may also include both an internal storage unit of the device and an external storage terminal. The memory 61 is used to store the computer instructions and other programs and data required by the terminal. The memory 61 can also be used to temporarily store data that has been output or is to be output.

Embodiments of the present application also provide a computer storage medium. The computer storage medium may be non-volatile or volatile. The computer storage medium stores a computer program. When the computer program is executed by a processor, the above trainings are implemented. Steps in method embodiments.

This application also provides a computer program product. When the computer program product is run on a training device, it causes the device to perform the steps in each of the above training method embodiments.

Embodiments of the present application also provide a chip or integrated circuit. The chip or integrated circuit includes: a processor, configured to call and run a computer program from a memory, so that the training equipment installed with the chip or integrated circuit performs each of the above trainings. Steps in method embodiments.

Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above functional units and modules is used as an example. In actual applications, the above functions can be allocated to different functional units and modules according to needs. Module completion means dividing the internal structure of the device into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiment can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be hardware-based. It can also be implemented in the form of software functional units. In addition, the specific names of each functional unit and module are only for the convenience of distinguishing each other and are not used to limit the scope of protection of the present application. For the specific working processes of the units and modules in the above system, please refer to the corresponding processes in the foregoing method embodiments, and will not be described again here.

In the above embodiments, each embodiment is described with its own emphasis. For parts that are not detailed or documented in a certain embodiment, please refer to the relevant descriptions of other embodiments.

Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented with electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Skilled artisans may implement the described functionality using different methods for each specific application, but such implementations should not be considered beyond the scope of this application.

The above-described embodiments are only used to illustrate the technical solutions of the present application, but not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still implement the above-mentioned implementations. The technical solutions described in the examples are modified, or some of the technical features are equivalently replaced; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit of the technical solutions in the embodiments of the application, and shall be included in this application. within the scope of protection applied for.

Claims

A training method, characterized by including:

Perceptual decision-making tasks are randomly displayed. The perceptual decision-making tasks include visual classification tasks, auditory classification tasks and audio-visual classification tasks. The visual classification tasks include classifying the M first pictures respectively. The auditory classification tasks include classifying the Nth pictures. A sound is classified separately, and the audio-visual classification task includes classifying L audio-visual stimulus pairs respectively, each of the audio-visual stimulus pairs includes a second picture and a second sound corresponding to the target in the second picture, wherein, M≥2, N≥2, L≥2;

Collect behavioral response data generated by users when completing the perceptual decision-making task;

Training results are determined based on the behavioral response data, and the training results include the accuracy of the user completing the perceptual decision-making task.
The training method according to claim 1, wherein the behavioral response data includes classification results corresponding to each classification task in the perceptual decision-making task and reaction times for completing each classification task, and the training method further includes :

Input the classification results and the reaction time into the preset drift diffusion model for processing, and obtain the drift rate, decision boundary and non-decision time;

The user's perceived decision-making ability is evaluated based on the drift rate, the decision boundary, and the non-decision time.
The training method according to claim 2, wherein after evaluating the user's perceptual decision-making ability based on the drift rate, the decision boundary and the non-decision time, the training method further includes:

The user's health status is determined based on the user's perceptual decision-making ability.
The training method according to claim 1, characterized in that before the random presentation of the perceptual decision-making task, the training method further includes:

Get M preset pictures;

Adjust the basic attributes of each preset picture to obtain M first pictures;

The visual classification task is constructed based on the M first pictures.
The training method according to claim 4, characterized in that before the random presentation of the perceptual decision-making task, the training method further includes:

Get N preset sounds;

Adjust the sound attributes of each preset sound to obtain N first sounds;

The auditory classification task is constructed based on the N first sounds.
The training method according to claim 5, characterized in that before the random presentation of the perceptual decision-making task, the training method further includes:

Determine L second pictures among the M first pictures;

Determine L second sounds among the N first sounds;

Pair the L second pictures and the L second sounds to obtain the L audio-visual stimulus pairs;

The audio-visual classification task is constructed based on the L audio-visual stimulus pairs.
The training method according to any one of claims 1 to 6, characterized in that the training method further includes:

Determine the stimulation intensity corresponding to each first picture and each first sound. The stimulation intensity is used to reflect the corresponding stimulation intensity of each first picture and each first sound when they are classified. Correct rate;

Select the picture whose stimulation intensity is the first stimulation intensity and the picture whose stimulation intensity is the second stimulation intensity among the M first pictures;

Select the sound whose stimulation intensity is the first stimulation intensity and the sound whose stimulation intensity is the second stimulation intensity among the N first sounds;

Construct a perceptual decision-making task of the first stimulation intensity based on the picture of the first stimulation intensity and the sound of the first stimulation intensity;

Based on the picture of the second stimulation intensity and the sound of the second stimulation intensity, a perceptual decision-making task of the second stimulation intensity is constructed.
A training device, characterized in that it includes:

A display unit for randomly displaying perceptual decision-making tasks. The perceptual decision-making tasks include visual classification tasks, auditory classification tasks, and visual and auditory classification tasks. The visual classification tasks include classifying M first pictures respectively, and the auditory classification tasks Including classifying N first sounds respectively, the audio-visual classification task includes classifying L audio-visual stimulus pairs respectively, each of the audio-visual stimulus pairs includes a second picture and a third image corresponding to the target in the second picture. Two sounds, where M≥2, N≥2, L≥2;

A collection unit, used to collect behavioral response data generated by the user when completing the perceptual decision-making task;

A determining unit, configured to determine a training result based on the behavioral response data, where the training result includes the accuracy of the user completing the perceptual decision-making task.
A training device, including a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that when the processor executes the computer program, it implements claims 1 to 1 The method described in any one of 7.
A computer-readable storage medium stores a computer program, characterized in that when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.