Abstract
Martial arts involve highly complex motions with intricate poses and maneuvers. Current video analysis methods face challenges in robustly scaling across diverse data while accurately modeling the nuances of techniques and forms. This research develops a deep learning framework integrating convolutional neural networks (CNN) and Gated Recurrent Units (GRU) on Apache Spark to model martial arts poses and motions. The CNN extracts hierarchical visual features, highlighting athletes and movements in multimedia data. The GRU layers leverage temporal dynamics by processing pose sequences and actions via gating mechanisms. The GRU maintains an internal state, updating it based on current inputs and previous states. Its gates selectively retain context enabling the GRU to summarize variable-length time series into fixed pose encodings. The framework concatenates CNN visual features and GRU sequential states to provide interpretable pose representations. This granular encoding captures subtle posture details for fine-grained technique analysis. Comparative experiments validate the approach through accurate classification on UCF101 and precise pose estimation on the MADS dataset. The Spark integration enables scalable distributed training on these datasets. Specifically, the approach achieves 94% validation accuracy on martial arts action recognition. It also demonstrates robust performance with low MSE of 2.48 and MAE of 1.89 for pose estimation. The validated framework powers applications such as augmented coaching, optimized training regimes, progress tracking, and skill assessment to enhance martial arts mastery. Overall, the research pioneers an interpretable deep learning solution tailored for fine-grained martial arts training through scalable spatio-temporal modeling of complex motions.
Similar content being viewed by others
Data availability
Inquiries about data availability should be addressed to the authors.
References
Abbas R, Gu N (2023) Improving deep learning-based image super-resolution with residual learning and perceptual loss using SRGAN model. Soft Comput 27:16041–16057
Ali M, Yin B, Bilal H et al (2023) Advanced efficient strategy for detection of dark objects based on spiking network with multi-box detection. Multimed Tools Appl. https://doi.org/10.1007/s11042-023-16852-2
Alizadeh Kolagar SA, Taheri A, Meghdari AF (2023) NAO robot learns to interact with humans through imitation learning from video observation. J Intell Robot Syst 109(1):4
Aslam MS (2021) L 2–L∞ control for delayed singular Markov switch system with nonlinear actuator faults. Int J Fuzzy Syst 23(7):2297–2308
Ćenanović N, Kevrić J (2022) Mixed martial arts bout prediction using artificial intelligence. In: International symposium on innovative and interdisciplinary applications of advanced technologies. Springer International Publishing, Cham, pp 452–468
Chen Z (2019) Observer-based dissipative output feedback control for network T-S fuzzy systems under time delays with mismatch premise. Nonlinear Dyn 95:2923–2941
Chen C, Nakagawa S (2023) Physical activity for cognitive health promotion: an overview of the underlying neurobiological mechanisms. Ageing Res Rev 86:101868
Corbi A, Santos OC, Burgos D (2019) Intelligent framework for learning physics with aikido (Martial Art) and registered sensors. Sensors 19(17):3681
Dai X, Hou J, Li Q, Ullah R, Ni Z, Liu Y (2020) Reliable control design for composite-driven scheme based on delay networked T-S fuzzy system. Int J Robust Nonlinear Control 30(4):1622–1642
Dou H, Liu Y, Chen S et al (2023) A hybrid CEEMD-GMM scheme for enhancing the detection of traffic flow on highways. Soft Comput 27:16373–16388. https://doi.org/10.1007/s00500-023-09164-y
Faisal MAA, Chowdhury ME, Khandakar A, Hossain MS, Alhatou M, Mahmud S, Ara I, Sheikh SI, Ahmed MU (2022) An investigation to study the effects of Tai Chi on human gait dynamics using classical machine learning. Comput Biol Med 142:105184
Hou Y, Seydou FM, Kenderdine S (2023) Unlocking a multimodal archive of Southern Chinese martial arts through embodied cues. J Doc. https://doi.org/10.1108/JD-01-2022-0027
Illig AM, Siedhoff NE, Schwaneberg U, Davari MD (2022) A hybrid model combining evolutionary probability and machine learning leverages data-driven protein engineering. bioRxiv, pp 2022–06
Ishac K, Bourahmoune K, Carmichael M (2023) An IoT sensing platform and serious game for remote martial arts training. Sensors 23(17):7565
Kapoor MP, Sugita M, Kawaguchi M, Timm D, Kawamura A, Abe A, Okubo T (2023) Influence of iron supplementation on fatigue, mood states and sweating profiles of healthy non-anemic athletes during a training exercise: a double-blind, randomized, placebo-controlled, parallel-group study. Contemp Clin Trials Commun 32:101084
Kasiri S, Fookes C, Sridharan S, Morgan S (2017) Fine-grained action recognition of boxing punches from depth imagery. Comput vis Image Underst 159:143–153
Le VH (2020) 3-D human pose estimation in traditional martial art videos. Int J Mach Learn Comput 10(2)
Li Q, Hou J (2021) Fault detection for asynchronous T-S fuzzy networked Markov jump systems with new event-triggered scheme. IET Control Theory Appl 15(11):1461–1473
Li K, Dong G, Gao Q (2023) Martial arts enhances working memory and attention in school-aged children: a functional near-infrared spectroscopy study. J Exp Child Psychol 235:105725
Litimein H, Huang ZY, Aslam MS (2023) Circular formation control with collision avoidance based on probabilistic position. Intell Autom Soft Comput 37(1)
Liu R, Jimenez ML, Haraszti CJ (2023) Martial arts and combat sports. The youth athlete. Academic Press, London, pp 849–861
Lv H (2021) Martial arts competitive decision-making algorithm based on improved BP neural network. J Healthc Eng 2021:1–8
Mahajan A, Jain R (2022) Martial art move detection and validation using the K-nearest neighbor algorithm. J High School Sci 6(4)
Manshadi M, Ranjbar E, Ghasab Sedehi R, Hassani N, Jafarnia Dabanloo N (2022) Postural balance for selection of martial artists using machine learning techniques. J Exerc Health Sci 2(1):1–12
Menon A, Siddig A, Muntean CH, Pathak P, Jilani M, Stynes P (2023) A machine learning framework for shuttlecock tracking and player service fault detection. In: International conference on deep learning theory and applications. Springer Nature Switzerland, Cham, pp 71–83
Morasso P (2023) Mental simulation of actions for learning optimal poses. Cogn Robot 3:185–200
Muhammad SA, Qaisar I, Majid A, Shamrooz S (2023) Adaptive event-triggered robust H∞ control for Takagi-Sugeno fuzzy networked Markov jump systems with time-varying delay. Asian J. Control 25(1):213–228
Niles BL, Grossman S, McQuade M, Grossman D, Kaiser AP, Muccio B, Warner B, Wang C, Mori DL (2023) Study protocol for a revised randomized trial: remotely delivered Tai Chi and wellness for Gulf War illness. Contemp Clin Trials 125:107045
Qaisar I, Majid A, Ramaraj P (2021) Design of sliding mode controller for sensor/actuator fault with unknown input observer for satellite control system. Soft Comput 25(24):14993–15003
Santos OC (2019) Artificial intelligence in psychomotor learning: modeling human motion from inertial sensor data. Int J Artif Intell Tools 28(04):1940006
Shamrooz MA, Zhenhua MA (2023) Output regulation for time-delayed Takagi-Sugeno fuzzy model with networked control system. Hacet J Math Stat 2023:1–21
Straiotto BG, Marwan N, James DC, Seeley PJ (2023) Recurrence analysis discriminates martial art movement patterns. Eur Phys J Spec Top 232(1):151–159
Tan N, Ren T (2023) A study on pulmonary functions in sports-players using machine learning and computational analysis. Soft Comput. https://doi.org/10.1007/s00500-023-08477-2
Tao L (2021) Application of data mining in the analysis of martial arts athlete competition skills and tactics. J Healthc Eng 2021:5574152
Tapaswi M, Kumar V, Laptev I (2021) Long term spatio-temporal modeling for action detection. Comput vis Image Underst 210:103242
Tropin Y, Podrigalo L, Boychenko N, Podrihalo O, Volodchenko O, Volskyi D, Roztorhui M (2023) Analyzing predictive approaches in martial arts research. Pedag Phys Cult Sports 27(4):321–330
Ullah R, Dai X, Sheng A (2020a) Event-triggered scheme for fault detection and isolation of non-linear system with time-varying delay. IET Control Theory Appl 14(16):2429–2438
Ullah R, Li Y, Aslam MS, Sheng A (2020b) Event-triggered dissipative observer-based control for delay dependent T-S fuzzy singular systems. IEEE Access 8:134276–134289
Wang J, Huang A (2021) Image recognition of martial arts movements based on FPGA and image processing. Microprocess Microsyst 80:103536
Wu Q, Li X, Wang K et al (2023) Regional feature fusion for on-road detection of objects using camera and 3D-LiDAR in high-speed autonomous vehicles. Soft Comput 27:18195–18213. https://doi.org/10.1007/s00500-023-09278-3
Xue M, Liu Y, Cai X (2023) Evaluation of prostate multi parameter bone structures for martial arts practitioners based on magnetic resonance imaging. J Radiat Res Appl Sci 16(2):100549
Yang W, Wang J, Shi J (2021) Video quality evaluation toward complicated sport activities for clustering analysis. Futur Gener Comput Syst 119:43–49
Zhang L, Qi T, Tao L (2022) Posture positioning estimation for players based on attention mechanism and hierarchical context. Neural Comput Appl. https://doi.org/10.1007/s00521-022-07800-6
Zhenhua M, Ullah R, Li Y, Sheng A, Majid A (2022) Stability and admissibility analysis of T-S descriptive systems and its applications. Soft Comput 26(15):7159–7166
Zhuang J, Sun J, Yuan G (2021) Arrhythmia diagnosis of young martial arts athletes based on deep learning for smart medical care. Neural Comput Appl 20:14641–14652
Funding
No funding was provided for the completion of this study.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
No conflict of interest has been declared by the authors.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Chen, G. An interpretable composite CNN and GRU for fine-grained martial arts motion modeling using big data analytics and machine learning. Soft Comput 28, 2223–2243 (2024). https://doi.org/10.1007/s00500-023-09565-z
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00500-023-09565-z