使用影片分析進行體育訓練改變了教練和運動員評估表現和訓練的方式。
本文介紹了一種基於機器學習的方法，用於羽球揮拍的關鍵畫面偵測，旨在通過視覺化改善初學者的學習體驗。
我們提出的方法使用MediaPipe框架提取骨骼關節點的3D座標，這些座標作為神經網絡模型的輸入特徵，預測羽球揮拍影片中的關鍵畫面位置。
為了協助研究，我們與國立清華大學羽球校隊合作，收集了一個羽球揮拍影片資料集作為訓練資料。我們的神經網絡模型集成在基於安卓系統開發的行動裝置APP中，允許學習者錄製他們的揮拍並與專業選手進行比較，從而增強了傳統的學習過程。
比較結果顯示，我們的多層感知器模型在準確性和可靠性方面優於其他現有方法。
我們也對於APP是否能夠幫助使用者學習進行了實驗，證明該應用程式對學習者表現、動機和自我認知的積極影響。
本研究的主要貢獻包括開發了一個穩健的關鍵畫面偵測模型、實際應用的開發，以及通過實驗和研究進行廣泛的驗證。

The use of video analysis in sports training has revolutionized the way coaches and players evaluate performance and develop strategies.
This paper presents a machine learning-based approach for keyframe detection in badminton swings aimed at improving the learning experience for beginners through visualization and real-time feedback.
Our proposed method uses the MediaPipe framework to extract 3D coordinates of skeleton keypoints, which serve as input features for a multilayer perception(MLP) model that accurately predicts keyframe positions in badminton swing videos.
To support the research, we collaborated with the National Tsing Hua University badminton team to collect a dataset of badminton swing videos for training purposes.
Our MLP model is integrated into a mobile app developed for Android systems, allowing learners to record their swings and compare them with professional players, thereby enhancing the traditional learning process.
Comparative studies show that our MLP model outperforms other existing methods in terms of accuracy and reliability.
User studies demonstrate the app's positive impact on performance, motivation, and self-perception.
The primary contributions of this research include the development of a robust keyframe detection model, practical application development, and extensive validation through experiments and research.

[1] Valentin Bazarevsky, Ivan Grishchenko, Karthik Raveendran, Tyler Zhu, Fan Zhang, and Matthias Grundmann. Blazepose: On-device real-time body pose tracking, 2020.
[2] G´erard Biau and Erwan Scornet. A random forest guided tour. Test, 25:197–227, 2016.
[3] Shyh-Ching Chi. A guide to badminton training. Taipei: Shi-Da Shu-Yuan, 2002.
[4] HJ Choi, Mike Hughes, and Peter O’Donoghue. Ergonomics issues and human-computer interaction in sport: A case study evaluation of a match analysis system for basketball. In Symposium Proceedings 6th IACSS, pages 31–35, 2007.
[5] Junyoung Chung, Caglar Gulcehre, KyungHyun Cho, and Yoshua Bengio. Empirical evaluation of gated recurrent neural networks on sequence modeling, 2014.
[6] Alan Dix. Human-computer interaction. Pearson Education, 2004.
[7] Klaus Greff, Rupesh K Srivastava, Jan Koutn´ık, Bas R Steunebrink, and J¨urgen Schmidhuber. Lstm: A search space odyssey. IEEE transactions on neural networks and learning systems, 28(10):2222–2232, 2016.
[8] J.F. Hair. Multivariate Data Analysis. Prentice Hall, 1998.
[9] Hui Han, Wen-Yuan Wang, and Bing-Huan Mao. Borderline-smote: A new over-sampling method in imbalanced data sets learning. In De-Shuang Huang, Xiao-Ping Zhang, and Guang-Bin Huang, editors, Advances in Intelligent Computing, pages 878–887, Berlin, Heidelberg, 2005. Springer Berlin Heidelberg.
[10] Haibo He and Edwardo A Garcia. Learning from imbalanced data. IEEE Transactions on knowledge and data engineering, 21(9):1263–1284, 2009.
[11] Shih-Yu Huang, Kuei-Pin Kuo, and Yi-Hsuan Lin. A golf swing analysis system using wii balance boardand kinect sensors for novice players. Multimed Tools Appl., 74:10679–10696, 2015.
[12] Magnus Ibh, Stella Grasshof, Dan Witzner, and Pascal Madeleine. Tempose: a new skeleton-based transformer model designed for fine-grained motion recognition in badminton. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 5199–5208, 2023.
[13] Meng Jian, Shuai Zhang, Lifang Wu, Shijie Zhang, Xiangdong Wang, and Yonghao He. Deep key frame extraction for sport training. Neurocomputing, 328:147–156, 2019. Chinese Conference on Computer Vision 2017.
[14] Serkan Kiranyaz, Onur Avci, Osama Abdeljaber, Turker Ince, Moncef Gabbouj, and Daniel J. Inman. 1d convolutional neural networks and applications: A survey. Mechanical Systems and Signal Processing, 151:107398, 2021.
[15] Hsiao-Chin Lan, Hsuan-Chih Wang, and Wu-Chou Chen. The effect of multimedia-assisted instruction on college students’ learning of the eight-form tai chi. Physical Education Journal, 43(3):109–126, 2010. [16] Kuo-Chin Lin, Hui-Chun Hung, and SC Young. The effectiveness of adopting tablets pc in physical education of badminton skill. Physical Education Journal, 47(3):437–449, 2014.
[17] Kuo-Chin Lin, Duan-Shin Lee, Shao-An Chien, Yueh-Hui Vanessa Chiang, and Nian-Shing Chen. The effect of a badminton teaching system using kinect on badminton skill performance. Physical Education Journal, 53(2), 2020.
[18] R. Liu, J. Zhu Y. Sun, L. Tian, and U. S. Kamilov. Recovery of continuous 3D refractive index maps from discrete intensity-only measurements using neural fields. Nature Machine Intelligence, 4:781–791, 2022.
[19] I Scott MacKenzie. Human-computer interaction: An empirical research perspective. Elsevier, 2024.
[20] Marina Papastergiou and Vassilis Gerodimos. Can learning of basketball be enhanced through a web-based multimedia course? an experimental study. Education and Information Technologies, 18:459–478, 2013.
[21] N. Promrit and S. Waijanya. Model for practice badminton basic skills by using motion posture detection from video posture embedding and one-shot learning technique. In AICCC 2019, Kobe, Japan, December 2019.
[22] Sameera Ramasinghe, K. G. Manosha Chathuramali, and Ranga Rodrigo. Recognition of badminton strokes using dense trajectories. In 7th International Conference on Information and Automation for Sustainability, pages 1–6, 2014.
[23] Muhammad Atif Sarwar, Yu-Chen Lin, Yousef-Awwad Daraghmi, Ts`ı-U´ı ˙IK, and Yih-Lang Li. Skeleton based keyframe detection framework for sports action analysis: Badminton smash case. IEEE Access, 11:90891–90900, 2023.
[24] Kyvete Shatri and Kastriot Buza. The use of visualization in teaching and learning process for developing critical thinking of students. European Journal of Social Sciences Education and Research, 9:71, 01 2017.
[25] Shih-Mei Su, Chi-Kun Weng, Chien-Chih Chou, and Yang-Tao Yeh. An experimental study on different computer-assisted teaching modes for fifth grade students’ learning of volleyball. 2008 Annual Academic Journal of College Physical Education, Taipei: Physical Education General Association of Universities and Colleges, pages 207–213, 2008.
[26] Shan Suthaharan and Shan Suthaharan. Support vector machine. Machine learning models and algorithms for big data classification: thinking with examples for effective learning, pages 207–235, 2016.
[27] J. Veˇrmiˇrovsk´y. The importance of visualisation in education, 2013.
[28] Song Wei, Kuili Wang, and Xiangliang Li. Design and implementation of college sports training system based on artificial intelligence. International Journal of System Assurance Engineering and Management, 13, 06 2021.
[29] Xiang Yan, Syed Zulqarnain Gilani, Hanlin Qin, Mingtao Feng, Liang Zhang, and Ajmal Mian. Deep keyframe detection in human action videos. arXiv preprint arXiv:1804.10021, 2018.
[30] See Shin Yue, Raveendran Paramesran, and Ganesh Krishnasamy. Readyto-serve detection in badminton videos. In 2024 International Conference on Electronics, Information, and Communication (ICEIC), pages 1–5, 2024.
[31] J. Zou, Q. Liu, and Z. Yang. Development of a moodle course for schoolchildren’s table tennis learning based on competence motivation theory: Its effectiveness in comparison to traditional training method. Computers & Education, 59(2):294–303, 2012.