進階搜尋


   電子論文尚未授權公開,紙本請查館藏目錄
(※如查詢不到或館藏狀況顯示「閉架不公開」,表示該本論文不在書庫,無法取用。)
系統識別號 U0026-2308202021354300
論文名稱(中文) 籃球比賽防守軌跡之自迴歸生成
論文名稱(英文) Autoregressive Generation for Basketball Defensive Trajectory
校院名稱 成功大學
系所名稱(中) 資訊工程學系
系所名稱(英) Institute of Computer Science and Information Engineering
學年度 108
學期 2
出版年 109
研究生(中文) 張圜華
研究生(英文) Yuan-Hua Chang
學號 P76074402
學位類別 碩士
語文別 英文
論文頁數 56頁
口試委員 指導教授-朱威達
共同指導教授-胡敏君
口試委員-朱宏國
口試委員-楊奕軒
中文關鍵字 籃球  防守策略  自迴歸模型 
英文關鍵字 basketball  defensive strategies  autoregressive model 
學科別分類
中文摘要 在VR中進行籃球戰術訓練有研究證明是有效的。而在VR籃球訓練系統中,雖然可以讓教練在電子戰術板上草擬進攻軌跡,但是若能提供即時的生成防守軌跡,將會提高訓練系統的實用性。因此為了能夠提供更好的虛擬訓練體驗,我們提出基於自迴歸特性的方法來模擬更逼真的防守軌跡。在我們所提出的方法中,我們藉由先前防守軌跡以及當前進攻者和球的位置作為輸入,並且設計出一種基於可微分位置採樣算法和因果卷積機制的生成模型,來學習玩家位置之間的關係。我們以客觀和主觀的方式評估生成防守軌跡與真實防守軌跡之間的相似性。為了進行客觀評估,我們以球員的防守位置、運動速度和加速度,以及基於沃羅諾伊算法計算球員在球場上的可進攻空間和防守壓力來比較生成軌跡和真實軌跡之間的差異。而為了進行主觀評估,我們邀請70名受試者進行問卷調查,讓他們判斷影片中顯示的防守軌跡是否為真實。根據問卷調查結果,受試者很難分辨真實的籃球防守軌跡和生成的防守軌跡區的差異。這表示我們所提出的自回歸模型可以生成真實的防守軌跡。
英文摘要 Tactics learning in VR has been proved to be effective for basketball training. In VR training system, the coach can input offensive trajectories by drawing via an electronic tactic board, but defensive trajectories should be generated automatically to improve the efficiency and usability. To provide a better virtual training process, we aim to simulate more realistic defensive trajectories based on an autoregressive method. In the proposed method, the previous defensive trajectories, the current offender positions, and the current ball position are taken as the input. Then, a generative model based on a differential position sampling algorithm and a causal convolution mechanism is designed to learn the relation between player positions. The similarity between the generated defensive trajectory and the real defensive trajectory is evaluated in both objective and subjective manners. For objective evaluation, we compare the defensive position, movement speed, and acceleration difference between the generated trajectories and the real ones. In addition, we calculate the empty space for the offender and the defensive pressure based on the Voronoi algorithm to compare defensive trajectories. For subjective evaluation, we recruited 70 experimenters to conduct questionnaires for judging whether the defensive trajectories shown in the video is realistic. According to the questionnaire result, the experimenters are difficult to distinguish the real basketball defensive trajectories from the generated ones. This implies that the proposed autoregressive model can generate realistic defensive trajectories.
論文目次 Abstract (Chinese) i
Abstract (English) ii
Table of Contents iii
List of Tables v
List of Figures vii
Chapter 1. Introduction 1
Chapter 2. Related Work 5
2.1 Basketball Analytics 5
2.2 Sequential Generative Model 7
Chapter 3. Basketball Defensive Trajectory Generation 9
3.1 Architecture 10
3.1.1 Convolutional Autoregressive Network 10
3.1.2 Differentiable Position Sampling 13
3.2 Loss Functions 15
3.2.1 Huber Loss 16
3.2.2 Standard Deviation Loss 16
3.2.3 Heuristic Loss 17
3.3 Training and Generation Process 18
3.3.1 Training Process 18
3.3.2 Generation Process 19
Chapter 4. Experimental Results 21
4.1 Dataset and Environment 21
4.2 Objective Evaluation Metrics 23
4.2.1 Average Difference of Position 24
4.2.2 Average Difference of Velocity 24
4.2.3 Average Difference of Acceleration 24
4.2.4 Average Difference of Empty Space for Offender 25
4.2.5 Average Difference of Defensive Pressure 26
4.3 Ablation Study 26
4.3.1 Impact of Loss Function 27
4.3.2 Comparison of Using Causal Convolution and Dilated Causal Convolution 32
4.3.3 Influence of the Parameter in the Basket Dependent Distance Loss 37
4.4 Comparison with Existing Methods 41
4.4.1 The Comparison of Average Difference of Position, Velocity, and Acceleration 42
4.4.2 The Comparison of Average Difference of Empty Space for Offender and Defensive Pressure 42
4.5 User Study 47
4.5.1 Questionnaire Results 47
Chapter 5. Conclusion and Future Works 52
References 53
參考文獻 [1] Stats sportvu. https://www.statsperform.com/team-performance/ basketball/.
[2] M. Abadi, A. Agarwal, P. Barham, E. Brevdo, Z. Chen, C. Citro, G. S. Corrado, A. Davis, J. Dean, M. Devin, S. Ghemawat, I. Goodfellow, A. Harp, G. Irving, M. Isard, Y. Jia, R. Jozefowicz, L. Kaiser, M. Kudlur, J. Levenberg, D. Mané, R. Monga, S. Moore, D. Murray, C. Olah, M. Schuster, J. Shlens, B. Steiner, I. Sutskever, K. Talwar, P. Tucker, V. Vanhoucke, V. Vasudevan, F. Viégas, O. Vinyals, P. Warden, M. Wattenberg, M. Wicke, Y. Yu, and X. Zheng. TensorFlow: Large-scale machine learning on heterogeneous systems, 2015. Software available from tensorflow.org.
[3] M. D. Akers, S. Wolff, and T. E. Buttross. An empirical examination of the factors affecting the success of ncaa division i college basketball teams. Journal of Business and Economic Studies, 1992.
[4] C.-Y. Chen, W. Lai, H.-Y. Hsieh, W.-H. Zheng, Y.-S. Wang, and J.-H. Chuang. Generating defensive plays in basketball games. In Proceedings of the 26th ACM international conference on Multimedia, pages 1580–1588, 2018.
[5] G. Csataljay, P. O’ Donoghue, M. Hughes, and H. Dancs. Performance indicators that distinguish winning and losing teams in basketball. International Journal of Performance Analysis in Sport, 9(1):60–66, 2009.
[6] P. Fearnhead and B. M. Taylor. On estimating the ability of nba players. Journal of quantitative analysis in sports, 7(3), 2011.
[7] A. Franks, A. Miller, L. Bornn, and K. Goldsberry. Counterpoints: Advanced defensive metrics for nba basketball. In 9th Annual MIT Sloan Sports Analytics Conference, Boston, MA, 2015.
[8] K. Goldsberry and E. Weiss. The dwight effect: A new ensemble of interior defense analytics for the nba. Sports Aptitude, LLC. Web, pages 1–11, 2013.
[9] I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y. Bengio. Generative adversarial nets. In Advances in neural information processing systems, pages 2672–2680, 2014.
[10] A. Graves. Generating sequences with recurrent neural networks. arXiv preprint arXiv:1308.0850, 2013.
[11] H.-Y. Hsieh, C.-Y. Chen, Y.-S. Wang, and J.-H. Chuang. Basketballgan: Generating basketball play simulation through sketching. In Proceedings of the 27th ACM International Conference on Multimedia, pages 720–728, 2019.
[12] P. J. Huber. Robust estimation of a location parameter. In Breakthroughs in statistics, pages 492–518. Springer, 1992.
[13] S. J. Ibáñez, J. Sampaio, S. Feu, A. Lorenzo, M. A. Gómez, and E. Ortega. Basketball game-related statistics that discriminate between teams’ seasonlong success. European journal of sport science, 8(6):369–372, 2008.
[14] S. Ibáñez, J. Sampaio, P. Buñuel, J. Giménez Fuentes-Guerra, and M. Janeira. Game statistics discriminating the final outcome of junior world basketball championship matches (portugal 1999). Journal of Human Movement Studies, 45:1–19, 01 2003.
[15] A. Karipidis, P. Fotinakis, K. Taxildaris, and J. Fatouros. Factors characterizing a successful performance in basketball. Journal of human movement studies, 41(5):385–397, 2001.
[16] D. P. Kingma and J. Ba. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980, 2014.
[17] D. P. Kingma and M. Welling. Auto-encoding variational bayes. arXiv preprint arXiv:1312.6114, 2013.
[18] A. Krizhevsky, I. Sutskever, and G. E. Hinton. Imagenet classification with deep convolutional neural networks. In Advances in neural information processing systems, pages 1097–1105, 2012.
[19] R. Maheswaran, Y.-H. Chang, A. Henehan, and S. Danesis. Deconstructing the rebound with optical tracking data. In Proceedings of the 6th annual MIT SLOAN sports analytics conference, 2012.
[20] A. v. d. Oord, S. Dieleman, H. Zen, K. Simonyan, O. Vinyals, A. Graves, N. Kalchbrenner, A. Senior, and K. Kavukcuoglu. Wavenet: A generative model for raw audio. arXiv preprint arXiv:1609.03499, 2016.
[21] F. J. Pineda. Generalization of back-propagation to recurrent neural networks. Physical review letters, 59(19):2229, 1987.
[22] L. Rabiner and B. Juang. An introduction to hidden markov models. ieee assp magazine, 3(1):4–16, 1986.
[23] J. Sampaio, E. J. Drinkwater, and N. M. Leite. Effects of season period, team quality, and playing time on basketball players’ game-related statistics. European Journal of Sport Science, 10(2):141–149, 2010.
[24] J. Sampaio, M. Janeira, S. Ibáñez, and A. Lorenzo. Discriminant analysis of game-related statistics between basketball guards, forwards and centres in three professional leagues. European journal of sport science, 6(3):173–178, 2006.
[25] W.-L. Tsai and M.-C. Hu. Training assistant: strengthen your tactical nous with proficient virtual basketball players. In ACM SIGGRAPH 2018 Posters, pages 1–2. 2018.
[26] G. Voronoi. Nouvelles applications des paramètres continus à la théorie des formes quadratiques. premier mémoire. sur quelques propriétés des formes quadratiques positives parfaites. Journal für die reine und angewandte Mathematik, 1908(133):97–102, 1908.
[27] G. U. Yule. Vii. on a method of investigating periodicities disturbed series, with special reference to wolfer’s sunspot numbers. Philosophical Transactions of the Royal Society of London. Series A, Containing Papers of a Mathematical or Physical Character, 226(636-646):267–298, 1927.
[28] Y. Zhang, Z. Gan, and L. Carin. Generating text via adversarial training. In NIPS workshop on Adversarial Training, volume 21, 2016.
論文全文使用權限
  • 同意授權校內瀏覽/列印電子全文服務,於2025-08-24起公開。
  • 同意授權校外瀏覽/列印電子全文服務,於2025-08-24起公開。


  • 如您有疑問,請聯絡圖書館
    聯絡電話:(06)2757575#65773
    聯絡E-mail:etds@email.ncku.edu.tw