A high school sports online and offline hybrid micro class closed-loop teaching implementation method

Abstract

The application discloses a high school sports online and offline mixed micro class closed loop teaching implementation method, and belongs to the field of intelligent sports teaching. The method relies on a five-layer architecture intelligent sports teaching system, and constructs a full-link teaching closed loop according to five stages of pre-class preview guided learning, in-class practice research, after-class consolidation expansion, multi-dimensional evaluation promotion learning and resource iteration optimization learning; precise hierarchical teaching is realized through a structured hierarchical micro class resource library, multi-dimensional student learning portrait and fusion recommendation algorithm, instant feedback and sports safety control are realized through full-process embedded AI posture recognition and wearable physiological monitoring, and process evaluation and lifelong sports literacy cultivation are implemented through a three-dimensional quantitative evaluation system. The application solves the core defects of traditional sports teaching, such as one-way infusion and online and offline fragmentation, effectively improves teaching efficiency and student participation, and can be widely applied to the teaching and reform of various college sports courses.

Description

Technical Field

[0001] This invention belongs to the field of smart sports teaching and internet education technology, and in particular relates to a closed-loop teaching method for online and offline blended micro-lessons in college sports. Background Technology

[0002] With the development of the internet and artificial intelligence technologies, higher education physical education is transforming from traditional offline, didactic teaching to a diversified and open model. Information-based teaching methods such as micro-lectures, MOOCs, and flipped classrooms are gradually entering physical education settings. The widespread adoption of online education platforms like Coursera and NetEase Open Courses has also promoted the open sharing of educational resources, providing a technological foundation for the digital transformation of physical education.

[0003] The integration of information technology into current college physical education teaching practices is still in its early stages, and existing technical solutions generally suffer from the following shortcomings:

[0004] First, the integration of online and offline learning is merely a formality, failing to form a complete closed-loop teaching chain. Most solutions are simply a combination of "uploading recorded videos online and regular offline lectures," with data fragmented between pre-class preparation, in-class practice, and post-class consolidation. The teaching content lacks progressive linkage, failing to fully realize the teaching value of the flipped classroom or support the continuous optimization and iteration of the teaching model.

[0005] Second, personalized teaching lacks technological support, making it difficult to implement differentiated instruction. Traditional teaching adopts a model of uniform content and pace, which cannot adapt to the differentiated needs of students with different physical fitness levels, starting points in sports skills, and sports preferences. Existing online resources are mostly standardized content, lacking a precise delivery mechanism based on individual student profiles, making it difficult to implement the "student-centered" educational philosophy.

[0006] Third, the feedback on motor skills is delayed, which limits the improvement of teaching efficiency. Learning sports skills relies heavily on movement correction, but the coverage of offline one-to-many teacher instruction is limited, and students' movement deviations cannot be corrected in time; most existing AI movement recognition solutions are only used for end-of-term skills assessments and are not embedded in the entire teaching process, so they cannot provide real-time guidance in the stages of preparation, practice, and consolidation, resulting in low efficiency in the transfer of skill cognition.

[0007] Fourth, the evaluation system is singular and rigid, lacking process-oriented and competency-based assessment. Current assessments generally rely on end-of-term summative skills tests and physical fitness tests, neglecting dimensions such as participation, progress, teamwork, and physical literacy during the learning process; the traceability of evaluation data is poor, failing to provide students with clear feedback on their growth, and making it difficult to support the implementation of educational goals such as "health first" and "lifelong physical education."

[0008] Fifth, there is insufficient safety management and a lack of real-time risk warning mechanisms for sports. Traditional teaching relies on teachers manually observing students' conditions, which makes it impossible to grasp the physiological load of each student in real time, and it is difficult to detect excessive or abnormal risks of exercise in a timely manner, resulting in technical blind spots in sports injury prevention and control. Summary of the Invention

[0009] This invention constructs a fully closed-loop blended teaching system that integrates flipped classrooms, intelligent profile recommendation, AI posture recognition, multi-dimensional process evaluation, and physiological load monitoring. It systematically solves the problem of fragmented instruction in traditional physical education, achieving personalized, precise, safe, and efficient skill acquisition and long-term development of physical literacy.

[0010] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0011] A closed-loop teaching method for online and offline blended micro-lessons in college physical education is implemented based on a smart physical education teaching system consisting of a perception layer, a network layer, a data layer, a platform service layer, and an application layer. The perception layer includes wearable motion monitoring devices, multi-view AI cameras, student mobile terminals, and teacher management terminals; the network layer relies on the campus 5G / WiFi network for low-latency data transmission; the data layer includes a student profile database, a structured micro-lesson resource library, a teaching process database, and a comprehensive evaluation database; the platform service layer incorporates seven core functional modules: resource management, user profiling, intelligent recommendation, AI posture recognition, exercise load monitoring, multi-dimensional evaluation, and data analysis iteration; and the application layer provides corresponding terminal applications for students, teachers, and teaching administrators.

[0012] The method specifically includes the following steps:

[0013] S1 Pre-class Preparation and Guidance Stage

[0014] First, a structured, hierarchical, and categorized micro-lesson resource library is built. Each micro-lesson is 3-8 minutes long, adapting to the fragmented learning scenarios on mobile devices. All micro-lessons are tagged with multi-dimensional metadata. At the beginning of the semester, a multi-dimensional learning baseline profile is built for each student using multi-source data, and this profile is dynamically updated as the teaching process progresses. Two to three days before each offline class, teachers release teaching objectives and core tasks. The platform uses a fusion recommendation algorithm combining content matching, collaborative filtering, and individual profiling to push a combination of pre-class micro-lessons ("basic essential learning + personalized elective learning") and self-test exercises to students. After students complete the pre-class preparation, the platform collects learning behavior data and, combined with the initial evaluation results of the pre-class action videos using AI posture recognition, automatically generates a class learning pre-diagnosis report, which teachers can use to adjust their offline classroom teaching plans.

[0015] S2 In-class Practical Learning Phase

[0016] Teachers conduct flipped classroom teaching based on pre-diagnosis results, focusing on common problems in the class and allocating most of class time to student practice. The system generates group solutions through a heterogeneous grouping algorithm, ensuring balance between groups and complementarity within groups. Differentiated practice tasks are assigned to groups at different levels, and collaborative practice and mutual error correction are carried out within groups, achieving student-to-student interaction. During the teaching process, wearable devices collect students' physiological load data in real time, triggering tiered safety warnings; multi-view AI cameras capture motion images in real time, supporting teachers to provide precise on-site feedback; teachers input classroom performance data in real time, which is synchronized to students' individual learning files. At the end of the course, a results presentation and multi-faceted peer assessment session is set up, where students complete self-assessment and inter-group peer assessment, enhancing classroom participation and emotional experience.

[0017] S3 After-Class Consolidation and Extension Stage

[0018] After the course concludes, the platform automatically generates personalized reinforcement learning packages based on the sports movement knowledge graph and student performance data. These packages include three categories: movement correction, physical fitness improvement, and extended learning, and are then pushed to the student's device. Students complete the after-class exercises and upload movement videos within the specified period. The AI ​​posture recognition module automatically scores the exercises from multiple dimensions and generates a visual error correction report, providing 24 / 7 real-time feedback. The platform also features an intelligent Q&A module to answer common student questions and pushes weekly sports literacy extension resources. A multi-verified extracurricular exercise check-in mechanism, combined with social incentive functions, guides students to form daily exercise habits, extending physical education from the classroom to daily life.

[0019] S4 Multidimensional Assessment Promotes Learning Stage

[0020] A three-dimensional quantitative evaluation system is constructed, consisting of formative assessment, summative assessment, and competency assessment, with the weights of each dimension flexibly adjustable according to the course project. The platform collects teaching data throughout the entire process in real time, automatically calculates students' comprehensive scores according to preset weights, and the entire data is traceable and appealable. The system has a learning progress early warning mechanism, automatically sending reminders and support suggestions to students with abnormal learning status; personalized growth reports are automatically generated at unit and semester milestones, including physical fitness trends, skill improvement trajectories, analysis of weaknesses, and subsequent exercise suggestions, realizing learning and teaching improvement through assessment.

[0021] S5 Resource Iteration and Optimization Phase

[0022] The platform's data analysis module regularly conducts in-depth analysis of the entire teaching process, examining micro-lesson resource quality, teaching model effectiveness, and overall teaching outcomes. It automatically identifies low-quality micro-lessons and inefficient teaching segments, generating optimization suggestions. Teachers then iteratively update micro-lesson resources based on the analysis results, supplementing any gaps in their teaching. Simultaneously, they adjust online pre-learning strategies and offline teaching organization methods, continuously refining the blended learning process to ultimately form a self-optimizing, complete teaching loop encompassing "teaching implementation - data collection - evaluation feedback - optimization iteration."

[0023] Furthermore,

[0024] The beneficial effects of the present invention after adopting the above structure are as follows:

[0025] (1) By connecting the whole process of pre-class, in-class and post-class through fragmented micro-lessons, data exchange and content linkage of each link are realized, forming a teaching loop that can be self-iterated; the flipped classroom model is implemented, which greatly increases the proportion of effective practice time in class and significantly improves the efficiency of classroom teaching.

[0026] (2) Through multi-dimensional student profiles and integrated intelligent recommendation algorithms, the micro-lesson resources and practice tasks are accurately pushed in a tiered manner to meet the learning needs of students with different foundations; AI action recognition provides real-time one-on-one guidance at all times, making up for the lack of energy of offline teachers, truly realizing individualized teaching, and fully stimulating students' initiative in self-learning.

[0027] (3) Open up multi-dimensional interaction channels between teachers and students, students and students, and humans and machines, and build a multi-directional information interaction system. While improving sports skills, strengthen students' collaborative ability, expression ability and self-learning ability, and shape a comprehensive quality to adapt to social competition.

[0028] (4) Break away from the assessment model that focuses solely on grades, and construct a three-dimensional evaluation system that covers process, results, and qualities. Replace the single horizontal comparison with vertical growth evaluation. Extend physical education from the classroom to daily life by expanding resources, sports check-in, and personalized sports prescriptions. Strengthen students' health awareness and sports habits, and implement the educational concepts of "health first" and "lifelong sports".

[0029] (5) Real-time monitoring of physiological load through wearable devices enables risk classification and early warning of sports activities, effectively reducing the incidence of sports injuries; multi-view AI action recognition enables accurate error correction, improving the scientific and professional level of physical education teaching. Attached Figure Description

[0030] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.

[0031] Figure 1This is a flowchart illustrating the overall method of an embodiment of the present invention;

[0032] Figure 2 This is a diagram of the intelligent physical education teaching system architecture upon which the present invention is based.

[0033] Figure 3 This is a schematic diagram of the workflow of the pre-class preparation and guidance stage of this invention. Detailed Implementation

[0034] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0035] This embodiment takes the "Stationary One-Handed Shoulder Shot" teaching unit (4 class hours) of a college's public physical education basketball elective course as an example to explain in detail the implementation process of this method.

[0036] Preliminary System Deployment and Preparation

[0037] The smart sports teaching system used in this embodiment is built on a five-layer architecture: the perception layer is equipped with 400 smart sports wristbands, 16 multi-angle AI cameras for the venue, a WeChat mini-program for students' mobile devices, and a tablet management terminal for teachers; the network layer relies on the campus 5G network to realize data transmission; the data layer deploys a student profile database, a micro-lesson resource database, a teaching process database, and an evaluation database; the platform service layer integrates seven core functional modules; and the application layer grants corresponding permissions to students, teachers, and teaching administrators respectively.

[0038] At the beginning of the semester, 320 students completed their basketball elective course selection. The platform connected with students' physical fitness test data, collected students' self-assessment results on basic basketball skills, sports experience, injury and contraindication information, and generated an initial learning baseline profile for each student. The teaching team uploaded 48 micro-lessons to the cloud platform to accompany the course. Each lesson is 3-6 minutes long and covers modules such as basketball rules, basic dribbling, passing and receiving, shooting techniques, common mistake corrections, and physical training. Each micro-lesson is annotated with metadata such as knowledge points, difficulty level, and key points of movements.

[0039] Step S1: Pre-class Preparation and Guidance

[0040] Two days before the first lesson, the teacher will release the teaching objectives for this lesson through the teacher's terminal: master the standard action of a stationary one-handed shoulder shot, correct common force errors, and mark the core knowledge points and teaching difficulties.

[0041] The platform integrates recommendation algorithms to automatically push two essential micro-lessons to all students: "Decomposition of the One-Handed Overhead Shot in Place" and "Common Shooting Mistakes and Correction Methods," along with five self-test questions on key knowledge points. For 89 students with weak basketball fundamentals, a preliminary micro-lesson, "Hand Shape and Ball Holding Basics," is also pushed. For 67 students with excellent fundamentals, an extension micro-lesson, "Advanced Shooting Power Techniques," is added.

[0042] By the day before the lesson, the class's pre-lesson completion rate was 94.7%, the average accuracy rate of self-tests was 71.8%, and a total of 196 students uploaded short videos of their pre-lesson shooting practice. The AI ​​posture recognition module, based on the detection of 33 key points on the human body, conducted an initial assessment of students' movements. Statistics showed that 68% of students had common problems such as "excessive elbow abduction angle" and "poor lower limb force transmission." The platform automatically generated a visual pre-lesson learning report and pushed it to the teacher. Based on this, the teacher adjusted the lesson plan, prioritizing "elbow joint correction" and "whole-body force connection," and increasing the time allocated to targeted supplementary exercises.

[0043] Practical learning in step S2

[0044] This in-person class will last 90 minutes and will be conducted according to the following procedure:

[0045] 1. Class Introduction (0-10 minutes): Based on the pre-class learning report, the teacher will focus on explaining common knowledge points and frequently occurring errors in movements. The teacher will compare the standard movement skeleton model with typical student errors on a large screen, without repeating the basic movement demonstration.

[0046] 2. Group Practice (10-60 minutes): The system uses a genetic algorithm to divide the 40 students in each class into 8 groups of 5, ensuring complementary abilities within each group. Teachers assign tiered tasks: the basic group focuses on hand position and ball-holding stability practice, while the advanced group focuses on complete movements and power connection practice. Students wear smart bracelets to monitor their heart rate and other physiological data in real time. The platform issues exercise intensity warnings to two students whose heart rates consistently exceed 85% of their maximum heart rate, prompting teachers to advise them to rest. Four AI cameras on the field capture students' practice movements in real time, and teachers circulate to provide guidance, offering one-on-one correction for individual movement deviations, focusing on addressing difficulties that students cannot overcome through independent practice.

[0047] 3. Demonstration and Peer Evaluation (60-85 minutes): Each group demonstrates their shooting technique. Students complete peer evaluations via mobile devices, with evaluation dimensions including technique accuracy (40%), improvement (30%), and teamwork (30%). Evaluation data is synchronized to the platform in real time. Teachers provide feedback to each group, highlighting students with significant improvement to enhance their sense of accomplishment and participation.

[0048] 4. Class summary (85-90 minutes): The teacher summarizes the core points of this lesson and assigns homework.

[0049] Step S3: Post-lesson reinforcement and extension learning

[0050] Within 1.5 hours after the course ends, the platform automatically generates personalized after-class learning packages based on students' in-class performance, physiological load data, and knowledge mastery, combined with the basketball movement knowledge graph: two micro-lessons, "Elbow Joint Abduction Correction Training" and "Shooting Power Sequence Practice," are pushed to 21 students whose in-class movement scores are below 70; a micro-lesson, "Home Upper Limb Strength Training," is pushed to 11 students with weak physical fitness; and advanced shooting skills expansion resources are pushed to students who have mastered the skills.

[0051] Students completed after-class exercises and uploaded shooting videos within three days. The AI ​​posture recognition module automatically scored the students based on five dimensions: body posture, joint angles, movement rhythm, force sequence, and range of motion, generating a visual movement deviation report. Data showed that the class's average movement standard improved from 52 points before class to 78 points, and most students' elbow abduction problems showed significant improvement.

[0052] Students can exchange practice experiences in the course discussion area, and the intelligent Q&A robot automatically answers common questions. Teachers provide daily group answers to difficult questions. Students' extracurricular exercise records are verified through wearable data and exercise trajectories and are included in their daily grades according to the rules. The platform pushes out weekly extension micro-lessons on sports culture and sports protection, and simultaneously releases information on school basketball games and club activities.

[0053] Step S4: Multidimensional Assessment Promotes Learning

[0054] This unit adopts a three-dimensional evaluation system, with the weighting set as follows: formative assessment 50%, summative assessment 40%, and competency assessment 10%. After the unit is completed, the platform automatically summarizes the data from the entire process, including pre-class preparation, in-class activities, post-class activities, and check-ins, and calculates the comprehensive unit score for each student. The entire data is traceable.

[0055] Each student receives a personalized learning report for each unit, including a trajectory of skill improvement, analysis of weaknesses, and suggestions for subsequent practice. The system triggers a learning alert for three students who fail to meet the standards in two consecutive practice sessions, sending reminders to both the students and teachers and simultaneously generating personalized support plans. Mid-term and final progress reports are generated by the platform at the end of the semester, visually presenting trends in students' physical fitness and skills and promoting the concept of lifelong physical activity.

[0056] Step S5: Resource Iteration and Optimization

[0057] Midway through the semester, the platform's data analysis module analyzed all teaching data from the entire semester: Micro-lesson quality analysis revealed that the completion rate for micro-lessons on "Basic Basketball Rules" was only 58%, with students reporting the content as dull; teaching model analysis showed that the skill improvement under the heterogeneous grouping model was significantly higher than that under the traditional uniform teaching model. Based on the analysis results, the teaching team remade the rule-related micro-lessons using NBA game examples, and after optimization, the completion rate of these micro-lessons increased to 82%. Simultaneously, the heterogeneous grouping teaching model was comprehensively promoted in subsequent teaching, continuously optimizing the teaching process and achieving a self-iterative closed-loop teaching system.

[0058] Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents. In conclusion, if those skilled in the art, inspired by this description, design similar structural methods and embodiments without departing from the inventive spirit of the invention, such designs should fall within the protection scope of the present invention.

Claims

1. A closed-loop teaching method for online and offline blended micro-lessons in college physical education, implemented based on a smart physical education teaching system consisting of a perception layer, network layer, data layer, platform service layer, and application layer, characterized in that... Includes the following steps: S1 Pre-class Preparation and Guidance: Construct a structured, hierarchical, and categorized micro-lesson resource library and a multi-dimensional student learning baseline profile. Based on teaching objectives and student profiles, intelligently push personalized pre-class micro-lessons through a fusion recommendation algorithm, collect pre-class behavior data and action data, and generate a class learning situation pre-diagnosis report. S2 In-Class Practice and Research: Based on the pre-diagnosis results of learning situation, offline flipped teaching is carried out. The heterogeneous grouping algorithm is used to realize stratified group practice, and students' physiological load data, classroom performance data and multi-angle motion video data are collected in real time to complete classroom achievement display and multi-dimensional peer evaluation. S3 Post-Class Consolidation and Extension: Based on sports movement knowledge graphs and in-class performance data, personalized consolidation learning packages are generated and pushed out. Through the AI ​​posture recognition module, multi-dimensional real-time feedback and intelligent tutoring are provided for post-class practice. Accompanying extended literacy resources and a multi-verified sports check-in mechanism cultivate lifelong sports literacy. S4 Multidimensional Assessment Promotes Learning: Constructs a three-dimensional quantitative assessment system consisting of process assessment, summative assessment, and competency assessment. Based on the entire teaching process data, it automatically calculates students' comprehensive scores, generates personalized growth feedback reports, and triggers learning progress warnings. S5 Resource Iterative Optimization: Based on full-link teaching data analysis, the quality of micro-lesson resources and the effectiveness of teaching models are dynamically updated to form a self-optimizing complete teaching loop.

2. The method for implementing closed-loop online and offline blended micro-lesson teaching in college physical education according to claim 1, characterized in that, The micro-lesson resource library in step S1 includes four categories of resources: theoretical cognition, action breakdown, error correction guidance, and extension and improvement. Each micro-lesson lasts 3-8 minutes and is labeled with knowledge point tags, difficulty level tags, suitable sports tags, physical load tags, and corresponding key action point tags through metadata.

3. The method for implementing closed-loop online and offline blended micro-lesson teaching in college physical education according to claim 1, characterized in that, The student learning baseline profile in step S1 includes data in four dimensions: physical fitness foundation, skill starting point, personalized characteristics, and learning behavior. The profile data is dynamically updated with the teaching process. The physical fitness foundation dimension is aligned with the national student physical fitness and health standard test data, and the personalized characteristics dimension includes students' sports preferences, learning goals, and injury and contraindication information.

4. The method for implementing closed-loop online and offline blended micro-lesson teaching in college physical education according to claim 1, characterized in that, The learning progress pre-diagnosis report in step S1 is generated in the following way: The AI ​​posture recognition module performs human key point detection on the pre-study action video uploaded by students, outputs the initial score of action standard and error point marking, combines the self-test scores of knowledge points and learning behavior data, counts the common knowledge point weaknesses and common action deviations of the class, and generates a visual learning progress report to be pushed to the teacher's end.

5. The method for implementing closed-loop online and offline blended micro-lesson teaching in college physical education according to claim 1, characterized in that, In step S2, the wearable motion monitoring device in the perception layer collects data on students' heart rate, exercise duration, calorie consumption, and exercise intensity in real time. The system presets safe heart rate thresholds for students with different fitness levels. When a student's heart rate exceeds the safe range or when high-intensity exercise is prolonged, the platform automatically sends a graded exercise risk warning to both the teacher and student.

6. The method for implementing closed-loop online and offline blended micro-lesson teaching in college physical education according to claim 1, characterized in that, The heterogeneous grouping algorithm in step S2 uses a genetic algorithm to generate a grouping scheme with 4-6 people in each group, ensuring that the overall ability between groups is balanced and the ability levels within groups are complementary. Teachers issue differentiated practical tasks to different groups at different levels: the basic group focuses on motion shaping practice, the intermediate group focuses on motion connection practice, and the advanced group focuses on scenario application practice.

7. The method for implementing closed-loop online and offline blended micro-lesson teaching in college physical education according to claim 1, characterized in that, The AI ​​posture recognition module in step S3 automatically scores students' after-class practice videos from five dimensions: body posture, joint angle, movement rhythm, force exertion sequence, and movement amplitude. It generates a visual movement deviation report, marking the error location, deviation value, and step-by-step correction suggestions.

8. The method for implementing closed-loop online and offline blended micro-lesson teaching in college physical education according to claim 1, characterized in that, The exercise check-in mechanism in step S3 combines wearable physiological data, GPS movement trajectory, and AI action recognition for multiple verifications to prevent cheating; verified extracurricular exercise data is converted into course grades according to preset rules, and the accompanying class exercise leaderboard and team check-in function realize social incentives.

9. The method for implementing closed-loop online and offline blended micro-lesson teaching in college physical education according to claim 1, characterized in that, In the three-dimensional quantitative evaluation system of step S4, process evaluation accounts for 40%-50% of the total score, summative evaluation accounts for 30%-40%, and competency evaluation accounts for 10%-20%. Among them, process evaluation includes an individual longitudinal progress index, which is calculated by comparing the student's own data before and after.

10. The method for implementing closed-loop online and offline blended micro-lesson teaching in college physical education according to claim 1, characterized in that, In step S5, the quality of micro-lesson resources is evaluated by statistically analyzing the completion rate, playback speed ratio, pause and replay hotspots, and mastery rate of corresponding knowledge points for different micro-lessons. Low-quality micro-lessons are automatically marked and optimization suggestions are generated. The effectiveness of the teaching model is evaluated and dynamically adjusted by comparing the skill improvement and student satisfaction under different teaching organization methods.

Images