Echo State Network Ensemble for Human Motion Data Temporal Phasing: A Case Study on Tennis Forehands
Temporal phasing analysis is integral to ubiquitous/“smart” coaching devices and sport science. This study presents a novel approach to autonomous temporal phasing of human motion from captured tennis activity (3D data, 66 time-series). Compared to the optimised Echo State Network (ESN) model achieving 85 % classification accuracy, the ESN ensemble system demonstrates improved classification of 95 % and 100 % accurate phasing state transitions for previously unseen motions without requiring ball impact information. The ESN ensemble model is robust to low-sampling rates (50 Hz) and unbalanced data sets containing incomplete data time-series. The demonstrated achievements are applicable to exergames, augmented coaching and rehabilitation systems advancements by enabling automated qualitative analysis of motion data and generating feedback to aid motor skill and technique improvements.
KeywordsComputational Intelligence (CI) Sport and rehabilitation Biomechanics Augmented Coaching Systems (ACS) Data analytics Human Motion Modelling and Analysis (HMMA)
The author wishes to express his appreciation to developers of Oger Toolbox (http://organic.elis.ugent.be/organic/engine) and Spyder IDE (https://pythonhosted.org/spyder/) utilised in this study. The tennis data was obtained in the Peharec polyclinic for physical therapy and rehabilitation, Pula (Croatia) in collaboration with Petar Bačić (biomechanics lab specialist and professional tennis coach). The author also wishes to express his sincere appreciation to Dr. Stefan Schliebs and Dr. Russel Pears for their valuable comments and insights.
- 2.Bačić, B.: Bridging the gap between biomechanics and artificial intelligence. In: XXIV International Symposium on Biomechanics in Sports - ISBS 2006, Salzburg, Austria, pp. 371–374 (2006)Google Scholar
- 3.Bacic, B.: Evolving connectionist systems for adaptive sports coaching. Neural Inf. Process. – Lett. Rev. 12, 53–62 (2008)Google Scholar
- 4.Bačić, B.: Connectionist methods for data analysis and modelling of human motion in sporting activities. Ph.D. thesis, School of Computer and Mathematical Sciences, AUT University, New Zealand, Auckland (2013)Google Scholar
- 5.Bačić, B.: Prototyping and user interface design for augmented coaching systems with MATLAB and Delphi: implementation of personal tennis coaching system. In: MATLAB Conference 2015, Auckland (2015)Google Scholar
- 6.Knudson, D.V.: Qualitative Diagnosis of Human Movement: Improving Performance in Sport and Exercise. Human Kinetics, Champain (2013)Google Scholar
- 9.Paugam-Moisy, H.: Spiking neuron networks a survey. In: IDIAP Research Institute (2006)Google Scholar