Social action completion degree evaluation method and system based on action segmentation

Abstract

The invention relates to a social action completion degree evaluation method and system based on action segmentation, and the method comprises the steps: guiding a user to make a plurality of evaluation actions, and collecting the user action data flow of each evaluation action; for each evaluation action, extracting action judgment data of each frame in the corresponding user action data stream; and for each evaluation action, calculating an evaluation index according to the action label of each corresponding frame, and outputting the evaluation index. According to the invention, through formulating the assessment process of the system, the demand for rehabilitation professionals is reduced, and the problems that the assessment professionals are insufficient and human resources are consumed due to manual assessment, and the result of manual assessment is often based on the subjective feeling of the professionals and lacks specific and objective assessment indexes are avoided.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method, system, terminal and readable storage medium for evaluating the degree of completion of a social action based on action segmentation. Background Art

[0002] Autism Spectrum Disorder (ASD), also known as autism spectrum disorder, is one of the most common neurodevelopmental disorders in children. Its clinical symptoms mainly include social communication disorders, stereotyped behaviors and narrow interests.

[0003] "Early detection, early diagnosis, and early intervention" is the consensus of autism rehabilitation strategies. Early education and behavioral intervention have been found to effectively improve many behavioral, functional, social, and cognitive defects of ASD children. Authoritative research and years of clinical experience of professional doctors have shown that through early behavioral intervention, the learning and living abilities of children with autism can be greatly improved, and the earlier the intervention, the better the effect. Among them, intervention in social skills is crucial for children with autism, because social cognitive defects exacerbate the challenges faced by autistic patients in society. Therefore, it is very important to evaluate and intervene in the social skills of children with autism.

[0004] In terms of social behavior, the existing technology has proposed the ability assessment and training methods of social behaviors such as responding to others' behaviors, continuing others' behaviors, and imagining continued interactions, which laid the foundation for the current social action assessment and intervention. At present, relevant industry practitioners and researchers are constantly committed to developing innovative intervention and assessment measures, but the existing assessment methods mainly rely on manual methods, that is, relying on professionals to make medical diagnoses through behavioral observations and formulate corresponding intervention plans. Specially trained clinical professionals are required to conduct assessments and interventions for autistic children through questionnaires and other forms. This manually performed autism assessment method has two disadvantages: first, because this process relies on experienced clinicians, the long training period and the limited number of rehabilitation professionals make it difficult to meet the growing demand for intervention and assessment services, which is very unfavorable for the management of the disease; at the same time, the results of manual assessments are often based on the subjective feelings of professionals and lack specific and objective assessment indicators, which ultimately hinders the understanding of children's rehabilitation progress and the formulation of future rehabilitation plans. Summary of the invention

[0005] The purpose of the present invention is to provide a method, system, terminal and readable storage medium for evaluating the completion of a social action based on action segmentation, aiming to solve the problem that the prior art relies on manual work when evaluating social actions, resulting in inaccurate evaluation results.

[0006] The technical solution adopted by the present invention to solve the technical problem is as follows:

[0007] The present invention provides a social action completion evaluation method based on action segmentation, and the social action completion evaluation method based on action segmentation specifically includes:

[0008] Guiding the user to perform multiple evaluation actions, and collecting a user action data stream for each evaluation action;

[0009] For each of the evaluation actions, extracting action determination data of each frame in the corresponding user action data stream;

[0010] For each of the evaluation actions, an evaluation index is calculated based on the action determination data of each corresponding frame, and the evaluation index is output.

[0011] Further, guiding the user to perform multiple evaluation actions, guiding the user to perform multiple evaluation actions, and collecting the user action data stream of each evaluation action specifically includes:

[0012] Play the guidance video of each evaluation action in sequence, and collect the video stream during each guidance video;

[0013] According to the video stream corresponding to each guiding action, the skeleton point frame sequence of each frame is extracted through the human body estimation algorithm, and the skeleton point frame sequence of each frame corresponding to each evaluation action is used as the user action data stream of the corresponding evaluation action.

[0014] Furthermore, the sequentially playing of the guidance video of each evaluation action includes:

[0015] Determining whether the user's motion amplitude is greater than a set threshold according to the skeleton point frame sequence;

[0016] If the user's movement amplitude is greater than the set threshold, the encouraging audio will be played.

[0017] Furthermore, for each of the evaluated actions, extracting the action determination data of each frame in the corresponding user action data stream specifically includes:

[0018] Inputting the skeleton point frame sequence into a skeleton point action segmentation algorithm;

[0019] The skeleton point action segmentation algorithm outputs the action label and confidence of each frame in the user action data stream;

[0020] The action label and the confidence level are used as the action determination data.

[0021] Furthermore, for each of the evaluation actions, the evaluation index is calculated according to the action determination data of each corresponding frame, specifically including:

[0022] For each of the evaluated actions, calculating the action completion degree according to the action labels and the confidences of the corresponding frames;

[0023] For each of the evaluated actions, calculating the action accuracy according to the action labels and the confidences of the corresponding frames;

[0024] The action completion degree and the action accuracy are used as the evaluation indicators.

[0025] Furthermore, the calculation of the action completion degree is:

[0026]

[0027] Among them, Ac represents the degree of action completion, t0 represents the starting time of the evaluation action, and T represents the time length of the evaluation action. represents the predicted probability that the action in the i-th frame is the correct action c0, It means that the predicted probability of the action in the i-th frame being the correct action c0 is the largest. I is the indicator function. When the input content is true, it takes the value 1, and when the input content is false, it takes the value 0.

[0028] Furthermore, the calculation of the action accuracy is:

[0029]

[0030] Among them, Aa represents the action accuracy, t0 represents the starting time of the evaluation action, and T represents the time length of the evaluation action. represents the predicted probability that the action in the i-th frame is the correct action c0, It means that the predicted probability of the action in the i-th frame being the correct action c0 is the largest. I is the indicator function. When the input content is true, it takes the value 1, and when the input content is false, it takes the value 0.

[0031] In addition, to achieve the above-mentioned purpose, the present invention also provides a social action completion evaluation system based on action segmentation, and the social action completion evaluation system based on action segmentation includes:

[0032] A guidance collection module, used for guiding the user to perform multiple evaluation actions and collecting a user action data stream for each evaluation action;

[0033] A label extraction module, used for extracting the action determination data of each frame in the corresponding user action data stream for each evaluation action;

[0034] The evaluation calculation module is used to calculate the evaluation index for each evaluation action according to the action determination data of each corresponding frame, and output the evaluation index.

[0035] In addition, to achieve the above-mentioned purpose, the present invention also provides a terminal, which includes: a memory, a processor, and a social action completion assessment program based on action segmentation stored on the memory and run on the processor, and when the social action completion assessment program based on action segmentation is executed by the processor, the terminal is controlled to implement the steps of the social action completion assessment method based on action segmentation as described above.

[0036] In addition, to achieve the above-mentioned purpose, the present invention also provides a readable storage medium, which stores a social action completion assessment program based on action segmentation, and when the social action completion assessment program based on action segmentation is executed by a processor, it implements the steps of the social action completion assessment method based on action segmentation as described above.

[0037] The present invention adopts the above technical solution to achieve the following effects:

[0038] The present invention reduces the demand for rehabilitation professionals by formulating a systematic evaluation process, avoiding the problems of insufficient evaluation professionals and waste of human resources caused by manual evaluation, as well as the fact that the results of manual evaluation are often based on the subjective feelings of professionals and lack specific and objective evaluation indicators. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a flowchart of the steps of a method for evaluating the completion of a social action based on action segmentation in a preferred embodiment of the present invention;

[0040] Figure 2 It is a task flow chart of a method for evaluating the completion of a social action based on action segmentation in a preferred embodiment of the present invention;

[0041] Figure 3 It is a structural diagram of a social action completion evaluation system based on action segmentation in a preferred embodiment of the present invention;

[0042] Figure 4 A schematic diagram of an operating environment of a preferred embodiment of a terminal of the present invention. DETAILED DESCRIPTION

[0043] In order to make the purpose, technical solution and advantages of the present invention clearer and more specific, the present invention is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0044] Embodiment 1

[0045] See also Figure 1 and Figure 2Embodiment 1 of the present application is a social action completion evaluation method based on action segmentation, which mainly evaluates the social behavior actions of autistic children, combines the functional game paradigm of social actions with human posture estimation and skeleton point action segmentation algorithm, and realizes a fully automated fine-grained evaluation of the social action behaviors of autistic children, so as to make certain contributions to the formulation of rehabilitation plans for autistic children in social actions and autism research.

[0046] The method includes two parts: a functional game task performed at the front end and a collection and evaluation task performed at the back end. The functional game task is a functional game for social action evaluation, which is used to guide and teach autistic children so that they can make a series of social actions required in social games during the evaluation stage. Then, videos of children making a series of actions in each evaluation stage are collected, and the evaluation index score of each action of the child is obtained through human posture estimation, skeleton point action segmentation algorithm and index calculation formula, thereby generating an evaluation report. Specifically, the social action completion evaluation method based on action segmentation includes the following steps:

[0047] S1. Guide the user to perform multiple evaluation actions, and collect the user action data stream of each evaluation action.

[0048] Specifically, guiding users to make evaluation actions and collecting character action data require two tasks to be performed simultaneously. The functional game task and the collection and evaluation task are started at the same time. The functional game is used to display game content and guide characters to interact. The collection and evaluation task collects video streams in real time in the background at this time, and provides the functional game with a skeleton point stream for real-time human posture estimation. In each evaluation stage, the skeleton point action segmentation algorithm and the indicator calculation formula are used to generate an evaluation report for each action.

[0049] For functional game tasks, it includes an induction stage and an evaluation stage. The induction stage includes an interaction induction stage and a social cognition induction stage. The evaluation stage includes a continuation behavior stage, a hypothetical interaction stage and a social interaction stage.

[0050] Among them, the interactive guidance stage is used to guide children to understand that the actions they take can interact with the system characters. In the interactive guidance stage, the voice prompts of the virtual cartoon characters in the functional game interface are mainly used to make children understand that their actions will be seen by the characters in the game. Therefore, at this time, the cartoon character poses a simple static pose, and the voice prompts the child: "You can see the actions done by the children." Then, the children are asked to imitate the poses of the cartoon characters through voice, and the skeleton point stream returned by the real-time human posture estimation algorithm in the acquisition and evaluation system is used to make similarity judgments. If the judgment threshold is reached, the child is praised through voice. If the child does not make the corresponding action, the child is encouraged to make the action. The guidance stage can be passed after the action is made.

[0051] The social cognition guidance stage is used to guide children to understand the actions they should take in the corresponding social scenarios. In the social cognition guidance stage, the voice explanations of the virtual cartoon characters in the functional game interface are mainly used to make children simply understand the game rules of social interaction. Through the actions and voice explanations of the cartoon characters, children can learn some basic instructions and social actions, such as clapping, bowing, waving, washing faces, etc. After each action is explained, ask the children whether they understand these actions. If they understand, let the children imitate the postures of the cartoon characters to enter the evaluation stage. If the children do not pose for a long time, the voice and action explanations will be played again.

[0052] In all three stages of the evaluation phase, cartoon characters issue social instructions and ask children to complete the actions in the instructions. Because initiating instructions and then receiving instructions and performing actions are all social, all three stages are called the evaluation stage of social interaction.

[0053] Among them, the continued behavior stage is used to enable children to perform social behavior actions that imitate and continue the behavior of characters in the game, also known as the imitation behavior stage. At this time, the functional game interface displays some dynamic actions performed by cartoon characters, such as clapping, jumping, squatting, stomping, turning around, waving and other simple dynamic actions. After completing an action, the child is prompted by voice to imitate the action just performed by the cartoon character for 5 seconds. Each action is given 5 seconds. After 5 seconds, the child will be given corresponding stop instructions and praise based on the degree of change in the child’s real-time posture (this feedback is only generated based on whether the child has a large posture change, and does not actually judge whether the posture is correct. The real evaluation data is generated by the skeleton point action segmentation data after the end of each stage). In this stage, a total of 5-10 actions are continued to be imitated.

[0054] The imaginary interaction stage is used to make children perform social behaviors described verbally by the characters in the game. At this time, the functional game interface displays the thinking bubbles of the cartoon characters, and the bubbles will show the static actions of the cartoon characters thinking, such as: bouncing a ball, brushing teeth, running, drinking water, eating a hamburger, and washing the face. At this time, ask the children with voice actions: "Children, can you wash your face? Can you pretend to wash your face for me to see? Don't stop." Through imaginary questions, let the children make relevant imaginary actions with certain meanings within 5 seconds. After 5 seconds, the children will be given corresponding stop instructions and praises based on the degree of change in the children's actual posture (this feedback is only based on whether the child has a large posture change, and does not actually judge whether the posture is correct. The real evaluation data is generated by the skeleton point action segmentation data after the end of each stage). In this stage, a total of 5-10 imaginary interactions are carried out.

[0055] The social interaction stage is used to make children perform social actions that should be performed in the social scene in the game. The social interaction at this time is actually an advanced version of imaginary interaction, which will give the interaction a certain social significance. At this time, the functional game interface displays cartoon characters and their social scenes, and guides children to perform a series of social behaviors through scenes and voice questions, such as bowing in class, waving at parting, making birthday wishes, applause after the performance on stage, and hugs from friends, etc., which have social significance. Through the voice questions of cartoon characters, for example, "Children, the actors on the stage have finished their performance. What should children do to show their appreciation for the performers after watching the performance? Please keep making such movements." Through such questions, children are asked to make corresponding social movements within 5 seconds. After 5 seconds, the children will be given corresponding stop instructions and praises according to the degree of change in the children's real-time posture (this feedback is only generated based on whether the children have a large posture change, and does not actually have a judgment on whether the posture is correct. The real evaluation data is generated by the data of skeleton point action segmentation after the end of each stage). In this stage, a total of 5-10 social movements are performed.

[0056] During the functional game task, the collection and evaluation task collects 30FPS real-time streaming video through the camera, and transmits the streaming video to the real-time human posture estimation algorithm to extract the real-time skeleton points of each frame, and saves each frame in real time. During the functional game, the human posture skeleton points of each frame are used to judge the movement amplitude of the child within 5 seconds of the action to be performed. When the child's movement amplitude is large, it is considered that the action posture has a certain change, that is, the child is considered to have made an action, and the child is given certain praise and enters the next stage to detect whether the child has completed the current series of actions through simple rules, which is used to give the child corresponding feedback and advance the game process.

[0057] Finally, the video clips of each evaluation stage are saved for archiving. After an evaluation stage (continuation behavior stage, imaginary interaction stage, social interaction stage) is completed, the skeleton point frame sequence of this stage is used as an action data stream to enter the skeleton point action segmentation algorithm for calculating indicators.

[0058] S2. For each of the evaluated actions, extract the action determination data of each frame in the corresponding user action data stream.

[0059] In this embodiment, a skeleton point action segmentation algorithm is used to obtain the degree of completion of the action. At the end of each of the three evaluation stages, the skeleton point frame sequence of this stage transmitted by the human posture estimation algorithm is received, and the skeleton point action segmentation algorithm is used to generate an action label and a corresponding confidence level for each frame. The action label and the corresponding confidence level are the action judgment data.

[0060] In this way, we can obtain what actions the children took in each time period in each evaluation stage, and the action confidence of each frame in this action time period. The fine-grained information of each frame can be further processed by formulas. For each action required in this time period, a unified, objective and meaningful indicator of action completion and action accuracy is generated.

[0061] S3. For each evaluation action, calculate an evaluation index according to the action determination data of each corresponding frame, and output the evaluation index.

[0062] Specifically, in this embodiment, calculating the evaluation index specifically includes calculating the action completion degree of each action and calculating the action accuracy of each action.

[0063] First, the action completion degree indicates whether each action has completed the action that should be completed within the expected 5s time period. The formula is as follows:

[0064]

[0065] Among them, Ac represents the degree of completion of the action, t0 represents the expected starting time of the action, which is the starting point of the 5s window period after the cartoon character finishes asking the question, and T represents the time length for evaluating the action, which is 5s. represents the predicted probability (confidence) that the action in the i-th frame is the correct action c0, The predicted probability that the action in the i-th frame is the correct action c0 is the largest, that is, the action label of the i-th frame is predicted to be c0, and I is an indicator function. When the content is true, it takes a value of 1; when the content is false, it takes a value of 0. Therefore, the meaning of this formula is the proportion of time that the child's action is the evaluation action within the 5s period, which is used as an indicator of action completion.

[0066] Secondly, the action accuracy represents the average confidence (prediction probability) of each action in the expected 5s time period for the correct action. The formula is as follows:

[0067]

[0068] Among them, Aa represents the action accuracy. The meaning of this formula is the average prediction probability of all frames that are correctly predicted as the evaluation action within the time period of 5s. Therefore, it can be expressed as the accuracy of this action and can be used as an indicator of children's action control ability.

[0069] The above indicators represent the action completion and action accuracy indicators of a single action within an evaluation stage. By averaging the completion (or action accuracy) of all actions within this stage, we can obtain the total action completion (or action accuracy) indicator within this evaluation stage, which also has certain reference significance.

[0070] After the calculation of the indicators of the above stages is completed, the two indicators of each action in each stage will be printed out together with the child’s data set and the evaluation time as the report data, which will serve as the basis for the child’s score in this evaluation. At the same time, such data will also be stored in the database for later data processing, analysis and access.

[0071] Embodiment 2

[0072] See also Figure 3 Based on the above method, the present invention also provides a social action completion evaluation system based on action segmentation, and the social action completion evaluation system based on action segmentation includes:

[0073] A guiding and collecting module 51 is used to guide the user to perform multiple evaluation actions and collect a user action data stream for each evaluation action;

[0074] A label extraction module 52 is used to extract the action determination data of each frame in the corresponding user action data stream for each evaluation action;

[0075] The evaluation calculation module 53 is used to calculate the evaluation index for each evaluation action according to the action determination data of each corresponding frame, and output the evaluation index.

[0076] Embodiment 3

[0077] See also Figure 4 Based on the above method, the present invention further provides a terminal, which includes a processor 10, a memory 20 and a display 30. However, it should be understood that it is not required to implement all the components shown, and more or fewer components may be implemented instead.

[0078] The memory 20 may be an internal storage unit of the terminal in some embodiments, such as a hard disk or memory of the terminal. The memory 20 may also be an external storage device of the terminal in other embodiments, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), etc. equipped on the terminal. Further, the memory 20 may also include both an internal storage unit of the terminal and an external storage device. The memory 20 is used to store application software and various types of data installed on the terminal, such as program codes of the installation terminal, etc. The memory 20 may also be used to temporarily store data that has been output or is to be output. In one embodiment, a social action completion evaluation program 40 based on action segmentation is stored on the memory 20, and the social action completion evaluation program 40 based on action segmentation can be executed by the processor 10, thereby realizing the terminal in the present application.

[0079] In some embodiments, the processor 10 may be a central processing unit (CPU), a microprocessor or other data processing chip, used to run the program code or process data stored in the memory 20, such as executing relevant programs of the social action completion evaluation method based on action segmentation.

[0080] In some embodiments, the display 30 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light Emitting Diode) touch device, etc. The display 30 is used to display information on the terminal and to display a visual user interface.

[0081] In one embodiment, when the processor 10 executes the social action completion evaluation program 40 based on action segmentation in the memory 20 , the steps of the social action completion evaluation method based on action segmentation as described above are implemented.

[0082] Embodiment 4

[0083] This embodiment provides a storage medium, which stores a social action completion evaluation program based on action segmentation. When the social action completion evaluation program based on action segmentation is executed by a processor, the steps of the social action completion evaluation method based on action segmentation as described above are implemented.

[0084] In summary, the present invention reduces the demand for rehabilitation professionals by formulating a systematic evaluation process, avoids the problems of insufficient evaluation professionals and waste of human resources caused by manual evaluation, and the results of manual evaluation are often based on the subjective feelings of professionals and lack specific and objective evaluation indicators.

[0085] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or terminal including the element.

[0086] Of course, those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing related hardware (such as a processor, a controller, etc.) through a computer program, and the program can be stored in a computer-readable storage medium, and the program can include the processes of the above-mentioned method embodiments when executed. The storage medium can be a memory, a disk, an optical disk, etc.

[0087] It should be understood that the application of the present invention is not limited to the above examples. For ordinary technicians in this field, improvements or changes can be made based on the above description. All these improvements and changes should fall within the scope of protection of the claims attached to the present invention.

Claims

1. A social action completion evaluation method based on action segmentation, characterized in that: The social action completion evaluation method based on action segmentation specifically includes: Guiding the user to perform multiple evaluation actions, and collecting a user action data stream for each evaluation action; For each of the evaluation actions, extracting action determination data of each frame in the corresponding user action data stream; For each of the evaluation actions, an evaluation index is calculated based on the action determination data of each corresponding frame, and the evaluation index is output.

2. According to the method for evaluating the completion of a social action based on action segmentation according to claim 1, it is characterized in that: The step of guiding the user to perform a plurality of evaluation actions and collecting a user action data stream of each evaluation action specifically includes: Play the guidance video of each evaluation action in sequence, and collect the video stream during each guidance video; According to the video stream corresponding to each guiding action, the skeleton point frame sequence of each frame is extracted through the human body estimation algorithm, and the skeleton point frame sequence of each frame corresponding to each evaluation action is used as the user action data stream of the corresponding evaluation action.

3. A social action completion evaluation method based on action segmentation according to claim 2, characterized in that: The step of playing the guide video of each evaluation action in sequence includes: Determining whether the user's motion amplitude is greater than a set threshold according to the skeleton point frame sequence; If the user's movement amplitude is greater than the set threshold, the encouraging audio will be played.

4. The method for evaluating the completion of a social action based on action segmentation according to claim 2, characterized in that: For each of the evaluated actions, extracting the action determination data of each frame in the corresponding user action data stream specifically includes: Inputting the skeleton point frame sequence into a skeleton point action segmentation algorithm; The skeleton point action segmentation algorithm outputs the action label and confidence of each frame in the user action data stream; The action label and the confidence level are used as the action determination data.

5. The method for evaluating the completion of a social action based on action segmentation according to claim 4, characterized in that: For each of the evaluation actions, calculating the evaluation index according to the action determination data of each corresponding frame specifically includes: For each of the evaluated actions, calculating the action completion degree according to the action labels and the confidences of the corresponding frames; For each of the evaluated actions, calculating the action accuracy according to the action labels and the confidences of the corresponding frames; The action completion degree and the action accuracy are used as the evaluation indicators.

6. The method for evaluating the completion of a social action based on action segmentation according to claim 5, characterized in that: The calculation of the action completion degree is: Among them, Ac represents the degree of action completion, t0 represents the starting time of the evaluation action, and T represents the time length of the evaluation action. represents the predicted probability that the action in the i-th frame is the correct action c0, It means that the predicted probability of the action in the i-th frame being the correct action c0 is the largest. I is the indicator function. When the input content is true, it takes the value 1, and when the input content is false, it takes the value 0.

7. The method for evaluating the completion of a social action based on action segmentation according to claim 5, characterized in that: The calculation of the action accuracy is: Among them, Aa represents the action accuracy, t0 represents the starting time of the evaluation action, and T represents the time length of the evaluation action. represents the predicted probability that the action in the i-th frame is the correct action c0, It means that the predicted probability of the action in the i-th frame being the correct action c0 is the largest. I is the indicator function. When the input content is true, it takes the value 1, and when the input content is false, it takes the value 0.

8. A social action completion evaluation system based on action segmentation, characterized in that: The social action completion evaluation system based on action segmentation includes: A guidance collection module, used for guiding the user to perform multiple evaluation actions and collecting a user action data stream for each evaluation action; A label extraction module, used for extracting the action determination data of each frame in the corresponding user action data stream for each evaluation action; The evaluation calculation module is used to calculate the evaluation index for each evaluation action according to the action determination data of each corresponding frame, and output the evaluation index.

9. A terminal, characterized in that: The terminal includes: a memory, a processor, and a social action completion assessment program based on action segmentation stored in the memory and executable on the processor. When the social action completion assessment program based on action segmentation is executed by the processor, the terminal is controlled to implement the steps of the social action completion assessment method based on action segmentation as described in any one of claims 1-8.

10. A readable storage medium, characterized in that: The readable storage medium stores a social action completion evaluation program based on action segmentation, and when the social action completion evaluation program based on action segmentation is executed by a processor, the steps of the social action completion evaluation method based on action segmentation as described in any one of claims 1-8 are implemented.

Images