摘要：本期体育人工智能学术研究共检索到英文相关文献300篇，研究热点主要集中在人工智能在体育教学、体育统计、体育产业、体育训练中的应用、人工智能神经网络对运动员成绩影响的研究、实时人工智能方法在体育运动中的应用、人工智能驱动的体育教育转型等。检索结果：1）关键词共词

本期体育人工智能学术研究共检索到英文相关文献300篇，研究热点主要集中在人工智能在体育教学、体育统计、体育产业、体育训练中的应用、人工智能神经网络对运动员成绩影响的研究、实时人工智能方法在体育运动中的应用、人工智能驱动的体育教育转型等。检索结果：1）关键词共词分析。提取关键词1738个，经过数据清洗后关键词有1440个，词频为3及以上的关键词有41个，累计百分比为15.48%，高频关键词有人工智能、机器学习、体育活动、深度学习、运动、传感器、锻炼、训练等，生成可视化知识图谱（见下图）。2）来源期刊分析。涉及期刊205种，其中载文4篇及以上的期刊有11种，所载文献累计百分比为23.67%，刊载体育人工智能前三位的期刊分别为：SCIENTIFIC REPORTS（JCR学科分区Q1），FRONTIERS IN PSYCHOLOGY （JCR学科分区Q2），FRONTIERS IN PUBLIC HEALTH（JCR学科分区Q1、Q1）。3）学科交叉分析。引用文献总计18333篇，最多的频次为8次，排名前三位的分别为Deep Residual Learning for Image Recognition 、World Health Organization 2020 guidelines on physical activity and sedentary behaviour、Attention Is All You Need。4）学术关注度分析。文献级别用量最高的是71次，排名前三位的分别为Bioinspired bicontinuous adhesive hydrogel for wearable strain sensor with high sensitivity and a wide working range、Hybrid Additive Manufacturing of Shear-Stiffening Elastomer Composites for Enhanced Mechanical Properties and Intelligent Wearable Applications、The 2025 motile active matter roadmap。

可放大查看

01

Ji XY, Samsudin SB, Hassan MZB, et al. The Application of Suitable Sports Games for Junior High School Students Based on Deep Learning and Artificial Intelligence[J]. SCIENTIFIC REPORTS, MAY 16 2025, vol.15, issue 1.

ABSTRACT:

In the contemporary educational environment, junior high school students' physical education is facing the challenge of improving teaching quality, strengthening students' physique, and cultivating lifelong physical habits. Traditional physical education teaching methods are limited by resources, feedback efficiency and other factors, and it is difficult to meet students' personalized learning needs. With the rapid development of artificial intelligence and deep learning technology, a new opportunity is provided for physical education innovation. This study intends to develop a Spatial Temporal-Graph Convolutional Network (ST-GCN) action detection algorithm based on the MediaPipe framework. This is achieved by integrating deep learning and artificial intelligence technologies. The algorithm aims to accurately identify the performance of junior high school students in sports activities, particularly in exercises such as sit-ups. By doing so, the study seeks to enhance the adaptability and teaching quality of physical education. Finally, this approach promotes the individualized development of students. By constructing the spatio-temporal graph model of human skeletal point sequence, accurate recognition of sit-ups can be achieved. Firstly, the algorithm obtains the data of human skeleton points through attitude estimation technology. Then it constructs a spatio-temporal graph model, which represents human skeleton points as nodes in the graph and the connectivity between nodes as edges. In HMDB51 dataset, the proposed average detection accuracy of ST-GCN action recognition algorithm based on MediaPipe framework reaches 88.3%. The proposed method has advantages in long-term prediction (> 500ms), especially at 1000ms, the values of Mean Absolute Error and Mean Per Joint Position Error are 71.1 and 1.04 respectively. They are obviously lower than those of other algorithms. ST-GCN action detection algorithm based on deep learning and artificial intelligence technology can significantly improve the accuracy of action recognition in junior middle school students' sports activities, and provide an immediate and accurate feedback mechanism for physical education teaching. This approach helps students correct their movements and enhance their sports skills. Additionally, it enables teachers to gain a deeper understanding of students' physical performance. These benefits provide strong support for the implementation of differentiated teaching.

02

Gao Y. The Role of Artificial Intelligence in Enhancing Sports Education and Public Health in Higher Education: Innovations in Teaching Models, Evaluation Systems, and Personalized Training[J]. FRONTIERS IN PUBLIC HEALTH, APR 30 2025, vol.13.

ABSTRACT:

With the rapid development of artificial intelligence (AI) technology, particularly in the field of physical education in higher education institutions, the application of AI has shown significant potential. AI not only offers innovative teaching models and evaluation systems for physical education, but also enhances teaching efficiency, enables personalized instruction, and improves students' athletic performance. In the context of public health, AI's role becomes even more crucial, as it assists in developing scientific exercise plans through precise motion data analysis, thereby promoting both physical and mental health. Furthermore, AI technology can drive innovation in the content and methods of public physical education teaching, providing robust support for high-quality sports education. Studies indicate that AI has optimized the physical education process, spurred the innovation of curriculum content, and facilitated the transformation of teaching models, injecting new momentum into the sustainable development of physical education in universities and the achievement of public health goals.

03

Li XY. Deep Learning to Promote Health Through Sports and Physical Training[J]. FRONTIERS IN PUBLIC HEALTH, MAY 27 2025, vol.13.

ABSTRACT:

Background Physical activity plays a crucial role in maintaining health and preventing chronic diseases. However, accurately assessing the impact of sports and physical training on health improvement remains a challenge. Recent advancements in deep learning and time-series analysis offer an opportunity to develop more personalized and accurate predictive models for assessing health improvement trends.Methods This study proposes a Health Improvement Score (HIS) prediction model based on a sequence-to-sequence deep learning architecture with Long Short-Term Memory (LSTM) networks and an attention mechanism. The model integrates heterogeneous time-series data, including physiological parameters (heart rate, blood oxygen levels, respiration rate), activity metrics (steps, distance, calories burned), sleep patterns, and body measurements. A dataset comprising 384 participants over a 32-day period was used to train and evaluate the model.Results The experimental results demonstrate that the proposed HIS prediction model outperforms traditional and machine learning-based models. It achieves 22.8% lower Mean Absolute Error (MAE), 19.3% lower Root Mean Squared Error (RMSE), 6.5% higher R2, and 7.9% higher Explained Variance Score (EVS) compared to competitive models.Conclusion The proposed HIS prediction model effectively captures complex temporal dependencies and improves the accuracy of health improvement predictions.

04

He ZL, Yang ZY, Xu JR, et al. Real-Time Accurate Determination of Table Tennis Ball and Evaluation of Player Stroke Effectiveness with Computer Vision-Based Deep Learning[J]. APPLIED SCIENCES-BASEL, MAY 12 2025, vol.15, issue 10.

ABSTRACT:

The adoption of artificial intelligence (AI) in sports training has the potential to revolutionize skill development, yet cost-effective solutions remain scarce, particularly in table tennis. To bridge this gap, we present an intelligent training system leveraging computer vision and machine learning for real-time performance analysis. The system integrates YOLOv5 for high-precision ball detection (98% accuracy) and MediaPipe for athlete posture evaluation. A dynamic time-wrapping algorithm further assesses stroke effectiveness, demonstrating statistically significant discrimination between beginner and intermediate players (p = 0.004 and Cohen's d = 0.86) in a cohort of 50 participants. By automating feedback and reducing reliance on expert observation, this system offers a scalable tool for coaching, self-training, and sports analysis. Its modular design also allows adaptation to other racket sports, highlighting broader utility in athletic training and entertainment applications.

05

Wu JH. Design and Development of Artificial Intelligence Dynamic Physical Education Teaching Resources in Human-Computer Interaction Mode[J]. JOURNAL OF EDUCATIONAL COMPUTING RESEARCH, MAY 2025.

ABSTRACT:

The objective of this research is to investigate how AI-improved dynamic physical education materials impact middle school education in physical settings. Utilizing a randomized controlled crossover approach, a 16-week study involved 120 students aged 12 to 18 to evaluate the impact of AI-enhanced physical education courses against traditional instructional techniques. Findings indicated a notable improvement in the AI group compared to traditional instruction in terms of physical fitness (23% enhancement), motor skill proficiency (31% improvement), and knowledge retention (27% enhancement) (p

06

Li ZH, Samsudin SB, Farizan NH, et al. The Intelligent Development and Preservation of Folk Sports Culture Under Artificial Intelligence[J]. APR 23 2025, vol.15, issue 1.

ABSTRACT:

To promote the intelligent development and preservation of folk sports culture, this work proposes a model grounded in the Cycle-Consistent Generative Adversarial Network (CycleGAN) to produce high-quality human images that recreate traditional sports movements. In order to improve the performance of the model, a discriminative mechanism for pose consistency and identity consistency is innovatively designed, and an appearance consistency loss function is introduced. Finally, the effectiveness of the model in image generation is verified. Experiments conducted on the Deep Fashion and Market-1501 datasets suggest that compared to other models, the proposed model achieves superior visual quality and realism in the generated images. In ablation experiments, the model incorporating the appearance consistency loss achieves improvements of 1.49%, 1.76%, and 2.2% in image inception score, structural similarity index, and diversity score, respectively, compared to the best-performing comparative models. This demonstrates the effectiveness of this loss function in improving image quality. Moreover, the proposed model excels across multiple evaluation metrics when compared to other models. In authenticity discrimination experiments, the generated images have a 58.25% probability of being judged as real, significantly surpassing other models. In addition, the results on the folk sports culture action dataset also show that the model proposed performs excellently in multiple indicators, and it particularly has an advantage in the balance between image diversity and quality. These results indicate that the CycleGAN model better reproduces the details and realism of folk sports movements. This finding provides strong technical support for the digital preservation and development of traditional sports culture.

07

Zhou XH, Zhu Y. Application of the Intelligent Back Propagation Neural Network in the Optimization of Sports Industry Structure[J]. SCIENTIFIC REPORTS, APR 8 2025, vol.15, issue 1.

ABSTRACT:

To explore the application potential of the intelligent Back Propagation Neural Network (BPNN) in the optimization of sports industry structure, a new intelligent BPNN model is constructed in this study. Firstly, the development status of the sports industry is introduced. Secondly, the principle and structure of intelligent BPNN are analyzed in detail. Finally, the BPNN model's architecture is optimized, and experiments verify the optimized model's effectiveness. The experimental dataset selected is the Kaggle-Sports Category dataset. The experimental results show that the proposed optimized model achieves a high score of 0.90 in user satisfaction. Meanwhile, it significantly outperforms the compared model in economic benefits, with a gain rate of 0.95 in box office revenue. In addition, although the proposed optimized model has slightly higher operating costs than other models, its excellent performance in resource utilization and economic benefits is sufficient to fill this gap. These experimental results prove the optimized model's application value in optimizing sports industry structure. This study provides valuable references for using intelligent technology, especially intelligent BPNN, to maximize the sports industry structure.

08

Lozzi D, Di Pompeo I, Marcaccio M, et al. AI-Powered Analysis of Eye Tracker Data in Basketball Game[J]. SENSORS, JUN 5 2025, vol.25, issue 11.

ABSTRACT:

This paper outlines a new system for processing of eye-tracking data in basketball live games with two pre-trained Artificial Intelligence (AI) models. blueThe system is designed to process and extract features from data of basketball coaches and referees, recorded with the Pupil Labs Neon Eye Tracker, a device that is specifically optimized for video analysis. The research aims to present a tool useful for understanding their visual attention patterns during the game, what they are attending to, for how long, and their physiological responses, blueas is evidenced through pupil size changes. AI models are used to monitor events and actions within the game and correlate these with eye-tracking data to provide understanding into referees' and coaches' cognitive processes and decision-making. This research contributes to the knowledge of sport psychology and performance analysis by introducing the potential of Artificial Intelligence (AI)-based eye-tracking analysis in sport with wearable technology and light neural networks that are capable of running in real time.

09

Mu K,Wang ZL,Tang JZ, et al. The Satisfaction of Ecological Environment in Sports Public Services by Artificial Intelligence and Big Data[J]. SCIENTIFIC REPORTS, APR 13 2025, vol.15, issue 1.

ABSTRACT:

In order to gain a more accurate understanding and enhance the relationship between the fitness ecological environment and artificial intelligence (AI)-driven sports public services, this study combines a Convolutional Neural Network (CNN) approach based on residual modules and attention mechanisms with the SERVQUAL evaluation model. The method employed involves the analysis of big data collected from questionnaire surveys, literature reviews, and interviews. This study critically examines the impact of advanced AI technologies on residents' satisfaction with the fitness ecological environment in sports public services and conducts theoretical analysis of the obtained data. The results show that the quality of sports public services empowered by AI significantly influences residents' satisfaction with the fitness ecological environment, such as running, swimming, ball games and other sports with high requirements for sports service quality and ecological environment. Only the good public sports service quality matching with them can meet the needs of the ecological environment for fitness, and stimulate the enthusiasm of the people for fitness. The study also shows that swimming, running and all kinds of ball games account for the largest proportion of all sports. To sum up, the satisfaction of residents' fitness ecological environment is greatly affected by the quality of public sports services, which is mainly reflected in the good and perfect sports environment and facilities that can provide residents with a wealth of fitness options, greatly improving the sports ecological environment. This study is helpful to realize the relationship between sports public service and sports ecological environment. It contributes to understanding the role of AI and deep learning in enhancing the correlation between sports public service and the ecological environment of sports.

10

Zhou DW, Keogh JWL,Ma YL, et al. Artificial Intelligence in Sport: A Narrative Review of Applications, Challenges and Future Trends[J]. JOURNAL OF SPORTS SCIENCES , JUN 2025.

ABSTRACT:

This narrative review explores the transformative impact of artificial intelligence (AI) in sport, covering its applications, challenges and future directions across key areas such as biomechanics, performance enhancement, sports medicine, health monitoring, coaching and talent identification. AI can potentially empower athletes to optimise movement, personalise training, improve diagnostics and accelerate rehabilitation. However, integrating AI into sport presents challenges, particularly around data privacy, ethical concerns and adoption within sport organisations. This review also addresses these issues, highlighting strategies for responsible data governance and transparency. Furthermore, the review explores the promising future trends for AI in sport, which suggest a profound impact how sport is practiced and managed globally, pointing towards an era of enhanced performance, health and inclusivity.

11

Zhao TT, Cabral J,Zhu GY. A Novel Explainable Artificial Intelligence Framework Using Knockoffs Techniques with Applications to Sports Analytics[J]. ANNALS OF OPERATIONS RESEARCH, APR 2025.

ABSTRACT:

The rapid integration of black-box Machine Learning (ML) models into critical decision-making scenarios has triggered an urgent call for transparency from stakeholders in Artificial Intelligence (AI). This call stems from growing concerns about the deployment of models whose decisions lack justification, legitimacy, and detailed explanations of their behavior. To address these concerns, Explainable Artificial Intelligence (XAI) has emerged as a crucial field, focusing on methods and processes that enable the comprehension of how AI systems make decisions, generate predictions, and execute their functions. The importance of XAI lies in its ability to provide explanations that justify a model's outputs, thereby ensuring trust and accountability in AI systems. In this work, we propose a novel XAI framework that leverages state-of-the-art statistical knockoff techniques to identify the most informative predictors while maintaining a controlled False Discovery Rate (FDR). This framework enhances informed decision-making by ensuring robust and interpretable insights. We validate our approach through synthetic data experiments, demonstrating that it can effectively identify important features with high power while providing finite-sample FDR control across various scenarios. We demonstrate the efficacy of our approach by applying it to predict the outcomes of National Football League (NFL) playoffs, a domain of significant importance in sports analytics. Our method provides invaluable insights that support strategic decision-making in the highly competitive field of professional football.

12

Han R, Yi MN, Feng W, et al. Enhancing Accuracy in Dynamic Pose Estimation for Sports Competitions Using HRPose: A Hybrid Approach Integrating SinglePose AI[J]. ALEXANDRIA ENGINEERING JOURNAL, MAY 2025, vol.127, pp.200-213.

ABSTRACT:

Human pose estimation plays a critical role in various applications, such as sports performance evaluation, rehabilitation, and human-computer interaction. Recent advancements in deep learning have significantly improved the accuracy and robustness of human pose estimation models. However, challenges remain in dynamic environments, especially in sports competitions, where high-speed movements, occlusions, and complex backgrounds often hinder accurate estimation. This paper proposes HRPose, a novel approach that combines HRNet for feature extraction and SinglePose AI for precise keypoint localization. It maintains high-resolution feature maps throughout the feature extraction process, enabling the model to capture fine-grained spatial details. SinglePose AI uses these features to generate and refine keypoint heatmaps, achieving accurate pose estimation even in challenging conditions. We evaluate HRPose on benchmark datasets, including the MPII Human Pose and PoseTrack datasets, and compare it with several models. Our results demonstrate that HRPose achieves superior performance in terms of mAP, precision, and robustness. Additionally, we discuss the real-time performance of HRPose and its potential applications in various domains, such as sports, healthcare, and rehabilitation. Future work will focus on improving the model's robustness to extreme conditions, such as low lighting and motion blur, and exploring its integration with multimodal data for more comprehensive analysis.

13

Wang KG, Wang L, Sun JD. The Data Analysis of Sports Training by ID3 Decision Tree Algorithm and Deep Learning[J]. SCIENTIFIC REPORTS, APR 29 2025, vol.15, issue 1.

ABSTRACT:

In order to improve the accuracy and efficiency of sports training data analysis, this paper proposes an optimized analysis model by combining Iterative Dichotomiser 3 (ID3) decision tree algorithm and deep learning model. As an important scientific tool, sports training data analysis aims to provide decision support for athletes and coaches, optimize training programs and improve sports performance through accurate data mining and model prediction. Traditional analysis methods have shortcomings in dealing with complex and multidimensional data, while analysis methods based on artificial intelligence can significantly improve the ability of feature extraction and prediction. Based on this background, this paper comprehensively evaluates the performance of each model in different dimensions by comparing six key indicators: mean square error (MSE), mean absolute error (MAE), information gain, feature importance, sports performance improvement rate and training target achievement rate. The experimental results show that the optimized model has the best MSE, and its MSE is only 1.05 under the information gain. It is significantly better than Extreme Gradient Boosting (XGBoost) of 1.48 and Capsule Networks (CapsNets) of 1.25. In terms of MAE, the minimum error of the optimized model is 0.65, while the maximum error of XGBoost is 1.11. In terms of information gain and feature importance, the optimization model is also outstanding, with the highest information gain of 1.02 and the feature importance maintained at a high level of 0.94 in many dimensions. Meanwhile, the optimized model is superior to other models in sports performance improvement rate (up to 6.71) and training target achievement rate (up to 78.32%). Therefore, this paper has certain reference significance to the field of sports training data analysis.

14

Noorbhai H, Moon S, Fukushima T. A Conceptual Framework and Review of Multi-Method Approaches for 3D Markerless Motion Capture in Sports and Exercise[J]. OURNAL OF SPORTS SCIENCES, JUN 18 2025, vol.43, issue 12, pp.1167-1174.

ABSTRACT:

The increasing diversity in motion capture technologies necessitates a structured approach to review and compare different systems. This paper presents a conceptual framework based on a review of existing motion capture methodologies, ranging from single-camera configurations to multi-camera systems enhanced with depth sensing and computer vision technology. The framework encompasses three distinct approaches: 1) single-camera with depth estimation, 2) single-camera with depth sensors, and 3) multiple cameras. Each method is detailed in terms of setup procedures, calibration techniques, advantages and disadvantages, as well as data processing workflows. The paper provides a framework and guide that can be adapted to different research and application contexts for sports and exercise, ensuring accurate and reliable 3D markerless motion capture. This framework aims to assist researchers, analysts and scientists in choosing the most suitable configuration based on their sport, specific requirements and/or constraints. By outlining the processes and considerations for each setup, this paper serves as a methodological guide, facilitating broader adoption and standardisation of advanced 3D motion capture technologies for sports and exercise. Although empirical data is not included in this paper, the focus on procedural guidelines demonstrates methodological rigour and practical implementation for 3D markerless motion capture research in sports and exercise.

责任编辑：马赛迈

来源：京津冀消息通