A visual interactive training method, device, and medium for autism spectrum disorder.

Abstract

This specification discloses a visual interaction training method, device, and medium for autism spectrum disorder, relating to the field of assistive training technology and applied to a training system. The system includes an attentional interest target generation module, an attentional interest guidance interaction module, and a shared attentional interest guidance interaction module. The method includes: generating attentional interest targets through the attentional interest target generation module, possessing multidimensional features that arouse the visual attention of the young child; conducting attentional guidance training for the young child based on social interaction videos and attentional interest targets through the attentional interest guidance interaction module; conducting shared attentional guidance training for the young child based on pre-acquired shared attentional interaction videos and attentional interest targets of the young child's caregiver through the shared attentional guidance interaction module; and generating reward feedback prompts for the young child after the shared attentional guidance training is completed, so that the caregiver can provide action rewards to the young child based on the reward feedback prompts.

Description

Technical Field

[0001] This specification relates to the field of assistive training technology, and in particular to a visual interactive training method, device and medium for autism spectrum disorder. Background Technology

[0002] Autism spectrum disorder (ASD) is a neurodevelopmental disorder. Individuals with ASD typically exhibit two types of symptoms: problems with social communication or social interaction, and restricted or repetitive patterns of behavior, interests, or activities. Autistic individuals often experience long-term difficulties including creating and maintaining social relationships, as well as difficulties maintaining work and performing daily tasks. Clinical manifestations primarily include language impairments, social interaction impairments, restricted interests, stereotyped behaviors, and intellectual disability.

[0003] Current training techniques for autism spectrum disorder primarily target deficits in social information processing. However, very young children, such as those aged three to six months, already exhibit deficits in basic social information processing functions. These deficits manifest later in development as difficulties in facial expression recognition and comprehension, and imitation. Lacking these early foundational functions, it's difficult to directly train facial expression recognition and comprehension later on. Furthermore, existing training and intervention methods often involve complex interactive activities that young children cannot effectively participate in. Therefore, current auxiliary training methods for autism spectrum disorder are not suitable for young children. Summary of the Invention

[0004] This specification provides one or more embodiments of a visual interactive autism spectrum disorder training method, device, and medium to solve the following technical problem: existing auxiliary training methods for autism spectrum disorder are not applicable to young children.

[0005] One or more embodiments of this specification employ the following technical solutions:

[0006] This specification provides one or more embodiments of a visual interaction training method for autism spectrum disorder, applied to a training system. The training system includes an attentional interest target generation module, an attentional interest guidance interaction module, and a shared attentional interest guidance interaction module. The method includes: generating attentional interest targets for a young child using the attentional interest target generation module, wherein the attentional interest targets possess multidimensional features that arouse the young child's visual attention; conducting attentional guidance training for the young child using the attentional interest guidance interaction module based on pre-acquired social interaction videos of the young child's caregivers and the attentional interest targets; when the attentional guidance training meets preset conditions, conducting shared attentional guidance training for the young child using the shared attentional interaction module based on pre-acquired shared attentional interaction videos of the young child's caregivers and the attentional interest targets; and generating reward feedback prompts for the young child after the shared attentional guidance training is completed, so that the young child's caregivers can provide action rewards based on the reward feedback prompts.

[0007] Furthermore, through the attentional interest target generation module, attentional interest targets for young children are generated, specifically including: pre-acquiring multiple image stimuli, each image stimulus including multiple feature dimensions, wherein the feature dimensions include image color, image brightness, and image type, and the image type includes any one of animal, plant, person, and object; displaying multiple image stimuli to the young child and acquiring the young child's physiological response data under each image stimulus, wherein the physiological response data includes pupil dilation degree, arousal degree, heart rate, and heart rate variability; based on the young child's physiological response data under each image stimulus, labeling multiple image stimuli that meet the requirements to obtain multiple specified image stimuli, and generating an adaptive training image library; generating stimulus images according to the pre-constructed generative adversarial neural network model and the adaptive training image library, and using the stimulus images as attentional interest targets for the young child.

[0008] Furthermore, based on the physiological response data of the young child under each image stimulus, multiple image stimuli that meet the requirements are labeled to obtain multiple designated image stimuli. Specifically, when a designated image stimulus is presented, if the degree of pupil dilation, arousal and heart rate in the physiological response data of the young child all reach local maxima and the degree of heart rate variability reaches local minima, then the designated image stimulus is used as the image stimulus that meets the requirements.

[0009] Furthermore, through the shared attention interest guidance interaction module, based on the pre-acquired shared attention interaction video of the caregiver of the young child and the attention interest target point, the young child undergoes shared attention guidance training. Specifically, this includes: acquiring the shared attention interaction video of the caregiver of the young child, wherein the shared attention interaction video includes a target object; capturing the gaze point of the young child when watching the shared attention interaction video through the shared attention interest guidance interaction module; and activating the attention interest target point when the young child's gaze point does not match the target object in the shared interaction video, so as to guide the young child to pay attention to the target object through the attention interest target point.

[0010] Furthermore, through the attention-interest guided interaction module, based on the pre-acquired social interaction video of the caregiver of the young child and the attention-interest target points, attention-guided training is conducted on the young child. Specifically, this includes: acquiring the social interaction video of the caregiver of the young child; capturing the interactive attention-interest points of the young child while watching the social interaction video through the attention-interest guided interaction module; and activating the attention-interest target points when the interactive attention-interest points do not match a preset area, so as to guide the young child to pay attention to the preset area through the attention-interest target points, wherein the preset area is the core facial area of ​​the caregiver of the young child in the social interaction video.

[0011] Furthermore, the attention guidance training meets preset conditions, specifically including: conducting multiple attention guidance training sessions on the young child, obtaining the target state of the attention interest target during each attention guidance training session, wherein the target state includes an active state and an inactive state; in multiple attention guidance training sessions, when the number of times the attention interest target is in an inactive state meets a preset threshold, it is determined that the attention guidance training meets the preset conditions.

[0012] Furthermore, the shared attentional interaction videos of caregivers of young children include entering the young child's field of vision and looking at an object to the left, entering the young child's field of vision and looking at an object to the right, and entering the young child's field of vision and looking at an object behind.

[0013] Furthermore, the social interaction videos of the caregivers of the young patients include: entering the young patient's field of vision, greeting the young patient, and engaging in designated social interactions with the young patient.

[0014] This specification provides one or more embodiments of a visually interactive autism spectrum disorder training device, comprising:

[0015] At least one processor; and,

[0016] A memory communicatively connected to the at least one processor; wherein,

[0017] The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enable the at least one processor to:

[0018] The attentional interest target generation module generates attentional interest targets for young children, wherein the attentional interest targets possess multidimensional features that can arouse the visual attention of the young children. The attentional interest guidance and interaction module conducts attentional guidance training for the young children based on pre-acquired social interaction videos of the young children's caregivers and the attentional interest targets. When the attentional guidance training meets preset conditions, the shared attentional interest guidance and interaction module conducts shared attentional guidance training for the young children based on pre-acquired shared attentional interaction videos of the young children's caregivers and the attentional interest targets. After the shared attentional guidance training is completed, a reward feedback prompt is generated for the young children, so that the caregivers can provide action rewards to the young children based on the reward feedback prompt.

[0019] This specification provides one or more embodiments of a non-volatile computer storage medium storing computer-executable instructions, wherein the computer-executable instructions are configured as follows:

[0020] The attentional interest target generation module generates attentional interest targets for young children, wherein the attentional interest targets possess multidimensional features that can arouse the visual attention of the young children. The attentional interest guidance and interaction module conducts attentional guidance training for the young children based on pre-acquired social interaction videos of the young children's caregivers and the attentional interest targets. When the attentional guidance training meets preset conditions, the shared attentional interest guidance and interaction module conducts shared attentional guidance training for the young children based on pre-acquired shared attentional interaction videos of the young children's caregivers and the attentional interest targets. After the shared attentional guidance training is completed, a reward feedback prompt is generated for the young children, so that the caregivers can provide action rewards to the young children based on the reward feedback prompt.

[0021] The above-mentioned technical solutions adopted in the embodiments of this specification can achieve the following beneficial effects: Through the above technical solutions, targeted attentional interest targets are generated for young children, avoiding the problem that the attentional interest targets generated cannot attract visual attention due to the differences in the cognitive level of young children. The generated attentional interest targets are more in line with the age characteristics of young children and have individual targeting. Attentional interest targets are used to conduct attention guidance training for young children, training the ability of young children to focus on the core area of ​​the caregiver's face. After the attention guidance training is completed, joint attention guidance training is conducted for young children. The training method is progressive and highly targeted, which can meet the training needs of young children. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments or prior art of this specification, the drawings used in the description of the embodiments or prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0023] Figure 1 A flowchart illustrating a visual interaction training method for autism spectrum disorder provided in an embodiment of this specification;

[0024] Figure 2 A flowchart illustrating another visual interaction training method for autism spectrum disorder provided in the embodiments of this specification;

[0025] Figure 3 This is a schematic diagram of the structure of a visually interactive autism spectrum disorder training device provided in the embodiments of this specification. Detailed Implementation

[0026] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments of this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0027] This specification provides a visual interaction training method for autism spectrum disorder. It should be noted that the execution entity in this specification's embodiments can be a server or any device with data processing capabilities. This specification's embodiments are applied to a training system, wherein the training system includes an attentional interest target generation module, an attentional interest-guided interaction module, and a shared attentional interest-guided interaction module. Figure 1 This is a flowchart illustrating a visual interaction-based training method for autism spectrum disorder, as provided in an embodiment of this specification. Figure 1 As shown, the main steps include the following:

[0028] Step S101: Generate attentional interest targets for young children through the attentional interest target generation module.

[0029] Among them, attentional interest targets possess multidimensional characteristics that can arouse visual attention in young children.

[0030] In one embodiment of this specification, "young child" refers to a child aged three to six months with autism spectrum disorder. Because children aged three to six months are in a developmental window, they have difficulty understanding language and exhibit significant differences in cognitive level compared to patients of other age groups. Therefore, it is necessary to generate attentional interest targets that can arouse the visual attention of young children. Attentional interest targets refer to points that generate and attract the visual attention of the young child and possess multi-dimensional image features.

[0031] The attentional interest target generation module generates attentional interest targets for young children, specifically including: pre-acquiring multiple image stimuli, each image stimulus including multiple feature dimensions, such as image color, image brightness, and image type, which can be any one of animal, plant, person, or object; displaying multiple image stimuli to the young child and acquiring the child's physiological response data under each image stimulus, including pupil dilation, arousal level, heart rate, and heart rate variability; based on the child's physiological response data under each image stimulus, labeling multiple image stimuli that meet the requirements to obtain multiple specified image stimuli and generating an adaptive training image library; generating stimulus images based on a pre-constructed generative adversarial neural network model and the adaptive training image library, and using these stimulus images as attentional interest targets for the young child.

[0032] Based on the physiological response data of the young child under each image stimulus, multiple image stimuli that meet the requirements are labeled to obtain multiple designated image stimuli. Specifically, when a designated image stimulus is presented, if the degree of pupil dilation, arousal and heart rate in the young child's physiological response data all reach local maxima and the degree of heart rate variability reaches local minima, then the designated image stimulus is taken as the image stimulus that meets the requirements.

[0033] In one embodiment of this specification, the attention target generation module includes multiple sets of image stimuli, each of which can be described as a set of multidimensional vectors. It should be noted that the dimension here refers to the feature dimension, which includes image color, image brightness, and image type. The image type includes animals, plants, people, and objects; animals further encompass dimensions such as birds, fish, and mammals; people further encompass dimensions such as male, female, child, teenager, middle-aged, and elderly; and objects further encompass dimensions such as vehicles and food. In one embodiment of this specification, a single image can form a total of twenty-three-dimensional descriptive vectors.

[0034] In one embodiment of this specification, multiple sets of image stimuli are randomly presented to young children. Eye-tracking devices are used to capture the degree of pupil dilation, facial expression analysis devices are used to capture the degree of arousal, and a contact wristband is used to acquire the child's heart rate and heart rate variability. By analyzing the physiological responses of the young children when presented with multiple sets of image stimuli, the image stimuli of interest to the young children are identified within the multiple sets of image stimuli.

[0035] In one embodiment of this specification, the image stimulus that interests the young child can be understood as the image stimulus that attracts the young child, which can be obtained through the young child's physiological response data. When an image stimulus is presented to the young child, if the child's pupil dilation, arousal level, and heart rate all reach local maxima, and the heart rate variability reaches a local minimum, then the image stimulus is considered to be a qualified image stimulus that can attract the young child's attention. It should be noted that the local maximum here refers to the maximum value corresponding to a sharp increase in physiological response data within the presentation period of the image. Conversely, the local minimum is the minimum value corresponding to a sharp decrease in physiological response data within the presentation period.

[0036] In one embodiment of this specification, after all image stimuli have been presented, all compliant image stimuli are automatically stored in an external storage device as an automatically generated adaptive training image library.

[0037] In one embodiment of this specification, the attentional interest target generation module includes a generator based on a Generative Adversarial Network (GAN) model. This generator randomly generates images that conform to the features of a training image library. Using the GAN model and the adaptive training image library from the external storage device in the previous step, a set of images is randomly generated as adaptive selection images for attentional interest target points. The generated stimulus images fully conform to the multidimensional features that can arouse visual attention in young children with autism spectrum disorder. It should be noted that when generating the adaptive selection images for attentional interest target points, multiple feature dimensions are obtained by comparing and analyzing multiple compliant image stimuli. Based on these multiple feature dimensions, the adaptive selection images for attentional interest target points are generated as target points. For example, if the compliant image stimuli are a child wearing white clothes and a white dog, then the obtained feature dimension is the white feature; that is, the white feature is a feature dimension that can attract the attention of the young child with autism spectrum disorder.

[0038] The above technical solution generates targeted attentional interest targets for each young child, avoiding the problem that the attentional interest targets generated cannot attract visual attention due to differences in the cognitive level of young children. In other words, the generated attentional interest targets are more in line with the age characteristics of young children and are individualized.

[0039] Step S102: Through the attention-interest guided interaction module, attention guidance training is conducted for young children based on pre-acquired social interaction videos of caregivers and attention-interest targets.

[0040] Young children with autism spectrum disorder often have a skewed focus of gaze when socializing with others. For example, in normal social situations, a normal person's focus is on the central area of ​​the other person's face, but a person with autism spectrum disorder usually focuses on the lower part of the face or areas away from the central area of ​​the face.

[0041] In one embodiment of this specification, an attention-guided interaction module is used to train the young child's attention based on pre-acquired social interaction videos and attention-target points of the caregiver, so as to guide the young child's attention to the core facial area of ​​the caregiver.

[0042] The attention-guided interaction module, based on pre-acquired social interaction videos of the caregivers of young children and the attention-interest target points, conducts attention guidance training for the young children. Specifically, it includes: acquiring the social interaction videos of the caregivers of young children; capturing the interactive attention-interest points of the young children while watching the social interaction videos through the attention-interest guided interaction module; and activating the attention-interest target points when the interactive attention-interest points do not match a preset area, so as to guide the young children to pay attention to the preset area through the attention-interest target points. The preset area is the core facial area of ​​the caregivers of young children in the social interaction videos.

[0043] The video of the caregiver for the young patient includes: entering the young patient's field of vision, greeting the young patient, and engaging in designated social interactions with the young patient.

[0044] The attention guidance training meets the preset conditions, specifically including: conducting multiple attention guidance training sessions on the young child, obtaining the target state of the attention interest target during each attention guidance training session, wherein the target state includes an active state and an inactive state; in multiple attention guidance training sessions, when the number of times the attention interest target is in an inactive state meets the preset threshold, the attention guidance training is determined to meet the preset conditions.

[0045] In one embodiment of this specification, the attention guidance module can be set in a portable terminal, such as a tablet computer or mobile phone. The social interaction video of the caregiver of the young child is recorded in advance through the portable terminal. The social interaction video includes: the caregiver entering the young child's field of vision, greeting the young child, and engaging in designated social interactions with the young child. The designated social interactions are relatively simple social types.

[0046] In one embodiment of this specification, the attentional interest points of young children watching the social interaction video are captured using an eye-tracking device installed on a portable terminal. When a mismatch is detected between the attentional interest point and the core facial region of the caregiver in the social interaction video, a generated attentional interest target point is automatically activated, arousing the child's attention and guiding the child's attention to the core facial region in the video. When the child's attentional point coincides with the core facial region of the caregiver in the social interaction video, the generated attentional interest target point automatically disappears.

[0047] In one embodiment of this specification, multiple attention-guided training sessions are conducted on young children. The target state of the attentional interest target is obtained during each training session. This target state includes an active state and an inactive state. That is, if the attentional interest target does not appear during each training session, it is inactive; if it appears, it is active. During multiple attention-guided training sessions, when the number of times the attentional interest target is inactive meets a preset threshold, the attention-guided training is deemed to have met the preset conditions. In other words, if the training result consistently shows that the attentional interest target does not appear more than 5 times during each training session, the attention-guided training is considered to have met the conditions, and the next training step can proceed.

[0048] Step S103: When the attention guidance training meets the preset conditions, the joint attention guidance training is carried out on the young child through the joint attention interest guidance interaction module, based on the pre-acquired joint attention interaction video and attention interest target of the caregiver of the young child.

[0049] In one embodiment of this specification, after the attention guidance training for young children is completed, it is necessary to train their joint attention guidance. In reality, normal children will follow the action after the caregiver makes a visual shift, for example, if the caregiver looks at an object on the left, a normal child will follow and look at the object on the left. Young children with disabilities have a deficiency in joint attention and therefore need joint attention guidance training.

[0050] The shared attention interest guidance interaction module trains the young child's attention based on pre-acquired shared attention interaction videos of the caregiver and the target attention point. Specifically, this includes: acquiring the shared attention interaction video of the caregiver, which includes a target object; capturing the young child's gaze point while watching the video; and activating the target attention point when the child's gaze point does not match the target object in the video, thereby guiding the child's attention to the target object. The shared attention interaction video includes images of the caregiver looking at an object to the left, looking at an object to the right, and looking at an object behind the child.

[0051] In one embodiment of this specification, the shared attention interest-guided interactive module can be installed in a portable terminal, such as a mobile phone or tablet computer. The portable terminal pre-records shared attention interaction videos of the caregiver for the young child, specifically including: entering the child's field of vision and looking at an object to the left; entering the child's field of vision and looking at an object to the right; and entering the child's field of vision and looking at an object behind.

[0052] In one embodiment of this specification, the gaze point of a young child watching a collaborative attentional interactive video is captured using an eye-tracking device mounted on a portable terminal. When the child's gaze point does not match a target object in the collaborative interactive video, an attentional interest target is activated to guide the child's attention to the target object. In other words, when a mismatch between the gaze point and a target object in the video is detected, a generated attentional interest target is automatically activated, stimulating the child's attention and guiding them to focus on the target object in the video. When the child's attention point coincides with the core area of ​​a face in the video, the generated attentional interest target automatically disappears.

[0053] Step S104: After the joint attention guidance training is completed, a reward feedback prompt is generated for the young child so that the caregiver can reward the young child with actions based on the reward feedback prompt.

[0054] In one embodiment of this specification, the young child is trained multiple times. During each training session, if the gaze point matches the target object in the video, that is, if no attentional interest target appears, a reward feedback prompt is generated to remind the caregiver to give the young child a physical reward after the training. The physical reward can be a smile or a comforting gesture.

[0055] The above technical solution generates targeted attentional interest targets for young children, avoiding the problem that attentional interest targets may fail to attract visual attention due to differences in the cognitive levels of young children. The generated attentional interest targets are more in line with the age characteristics of young children and are individually targeted. Using these attentional interest targets, young children can undergo attention guidance training to improve their ability to focus on the core area of ​​the caregiver's face. After the attention guidance training is completed, joint attention guidance training can be conducted on young children. The training methods are progressive and highly targeted, which can meet the training needs of young children.

[0056] This specification also provides another training method for visual interaction in autism spectrum disorder through its embodiments. Figure 2 A flowchart illustrating another visual interaction training method for autism spectrum disorder provided in this specification, as shown in the embodiments. Figure 2 As shown, the main steps include the following:

[0057] First, attention is paid to the adaptive generation of interest targets. Multiple sets of images are randomly presented, and multimodal physiological data of young children are recorded during the process of viewing the images. The images are then labeled. A labeled image library is formed, and a GAN model is used to generate images that match the library features.

[0058] Then, attention-oriented interest guidance training and joint attention mechanism attention-oriented interest guidance training were conducted.

[0059] This specification also provides an embodiment of a visual interactive autism spectrum disorder training device, such as... Figure 3 As shown, the device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to:

[0060] The attentional interest target generation module generates attentional interest targets for young children, which possess multidimensional features that can arouse the visual attention of the young children. The attentional interest guidance and interaction module uses pre-acquired social interaction videos of the young children's caregivers and the attentional interest targets to conduct attentional guidance training for the young children. When the attentional guidance training meets preset conditions, the shared attentional interest guidance and interaction module uses pre-acquired shared attentional interaction videos of the young children's caregivers and the attentional interest targets to conduct shared attentional guidance training for the young children. After the shared attentional guidance training is completed, a reward feedback prompt is generated for the young children, allowing the caregivers to provide behavioral rewards based on the reward feedback prompt.

[0061] This specification also provides a non-volatile computer storage medium storing computer-executable instructions, wherein the computer-executable instructions are configured as follows:

[0062] The attentional interest target generation module generates attentional interest targets for young children, which possess multidimensional features that can arouse the visual attention of the young children. The attentional interest guidance and interaction module uses pre-acquired social interaction videos of the young children's caregivers and the attentional interest targets to conduct attentional guidance training for the young children. When the attentional guidance training meets preset conditions, the shared attentional interest guidance and interaction module uses pre-acquired shared attentional interaction videos of the young children's caregivers and the attentional interest targets to conduct shared attentional guidance training for the young children. After the shared attentional guidance training is completed, a reward feedback prompt is generated for the young children, allowing the caregivers to provide behavioral rewards based on the reward feedback prompt.

[0063] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments of apparatus, devices, and non-volatile computer storage media are basically similar to the method embodiments, so the descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0064] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0065] The above description is merely one or more embodiments of this specification and is not intended to limit this specification. Various modifications and variations can be made to the one or more embodiments of this specification by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of one or more embodiments of this specification should be included within the scope of the claims of this specification.

Claims

1. A visual interactive autism spectrum disorder training method, characterized by, The method is applied to a training system, wherein the training system comprises an attention interest target point generation module, an attention interest guiding interaction module, and a common attention interest guiding interaction module, and the method comprises the following steps: Generating, by the attention interest target point generation module, an attention interest target point of a young patient, wherein the attention interest target point has a multi-dimensional feature capable of arousing visual attention of the young patient; Performing, by the attention interest guiding interaction module, attention guiding training on the young patient according to a pre-acquired social interaction video of a young patient caregiver and the attention interest target point; When the attention guiding training meets a preset condition, performing, by the common attention interest guiding interaction module, common attention guiding training on the young patient according to a pre-acquired common attention interaction video of the young patient caregiver and the attention interest target point; When the common attention guiding training is completed, generating a reward feedback prompt for the young patient, so that the young patient caregiver rewards the young patient with an action based on the reward feedback prompt; Performing, by the common attention interest guiding interaction module, common attention guiding training on the young patient according to a pre-acquired common attention interaction video of a young patient caregiver and the attention interest target point, specifically comprising the following steps: Acquiring the common attention interaction video of the young patient caregiver, wherein the common attention interaction video comprises a target object; Capturing, by the common attention interest guiding interaction module, a fixation point of the young patient when the young patient watches the common attention interaction video; When the fixation point of the young patient does not match the target object in the common attention interaction video, activating the attention interest target point to guide the young patient to pay attention to the target object through the attention interest target point; Performing, by the attention interest guiding interaction module, attention guiding training on the young patient according to a pre-acquired social interaction video of a young patient caregiver and the attention interest target point, specifically comprising the following steps: Acquiring the social interaction video of the young patient caregiver, and capturing, by the attention interest guiding interaction module, an interactive attention interest point of the young patient when the young patient watches the social interaction video; When the interactive attention interest point does not match a preset region, activating the attention interest target point to guide the young patient to pay attention to the preset region through the attention interest target point, wherein the preset region is a facial core region of the young patient caregiver in the social interaction video; The attention guiding training meets a preset condition, specifically comprising the following steps: Performing multiple times of attention guiding training on the young patient, and acquiring a target point state of the attention interest target point in each attention guiding training process, wherein the target point state comprises an activated state and an unactivated state; When the number of times of the unactivated state of the attention interest target point in the multiple attention guiding training meets a preset threshold, it is determined that the attention guiding training meets the preset condition.

2. The visual interactive autism spectrum disorder training method according to claim 1, wherein, Generating, by the attention interest target point generation module, an attention interest target point of a young patient, specifically comprising the following steps: Pre-acquire a plurality of image stimuli, each image stimulus comprising a plurality of feature dimensions, wherein the feature dimensions comprise image color, image brightness, and image type; Display a plurality of image stimuli to the low-age child, and acquire physiological response data of the low-age child under each image stimulus, wherein the physiological response data comprises pupil dilation degree, arousal degree, heart rate, and heart rate variability degree; Based on the physiological response data of the low-age child under each image stimulus, label a plurality of image stimuli meeting the requirements to obtain a plurality of specified image stimuli, and generate an adaptive training image library; Generate a stimulus image according to a pre-constructed generative adversarial neural network model and the adaptive training image library, and use the stimulus image as a target of attention interest of the low-age child.

3. The visual interactive autism spectrum disorder training method of claim 2, wherein, Based on the physiological response data of the low-age child under each image stimulus, label a plurality of image stimuli meeting the requirements to obtain a plurality of specified image stimuli, specifically comprising: When a specified image stimulus is presented, the pupil dilation degree, arousal degree, and heart rate in the physiological response data of the low-age child all reach local maximum values, and the heart rate variability degree reaches a local minimum value, the specified image stimulus is taken as an image stimulus meeting the requirements.

4. The visual interactive autism spectrum disorder training method of claim 1, wherein, The common attention interaction video of the low-age child caregiver includes entering the visual field of the low-age child and looking at the left object, entering the visual field of the low-age child and looking at the right object, and entering the visual field of the low-age child and looking at the rear object.

5. The visual interactive autism spectrum disorder training method of claim 1, wherein, The social interaction video of the low-age child caregiver includes entering the visual field of the low-age child and greeting the low-age child.

6. A visual interactive autism spectrum disorder training device, characterized by, The device comprises: at least one processor; and a memory connected in communication with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-5.

7. A non-volatile computer storage medium storing computer executable instructions configured to perform the method of any one of claims 1-5.

Images