In this thesis, we propose a system to evaluate a user’s performance in a real-time
manner. By utilizing the ability of Kinect V2 sensor to capture human motion, the system
records a professional user’s motion. This recording can be used as a reference motion for
novice users to practice. This way, the presence of the professional is not needed. Later on,
the novice user can re-play the recording inside a 3-dimensional environment and follow
through in order to practice the motion. In this 3-D environment, the system presents the
motion by a 3-D character, retargeted according to the novice’s body. This retargeting system
allows a novice user to view the motion using a 3-D model with the size relative to his/her
own body size. This system also provides feedback by changing the color of the model to
indicate the correctness of user’s motion. This way, the feedback can easily be recognized by
the novice, thus making motion learning more effective.

In this thesis, we propose a system to evaluate a user’s performance in a real-time
manner. By utilizing the ability of Kinect V2 sensor to capture human motion, the system
records a professional user’s motion. This recording can be used as a reference motion for
novice users to practice. This way, the presence of the professional is not needed. Later on,
the novice user can re-play the recording inside a 3-dimensional environment and follow
through in order to practice the motion. In this 3-D environment, the system presents the
motion by a 3-D character, retargeted according to the novice’s body. This retargeting system
allows a novice user to view the motion using a 3-D model with the size relative to his/her
own body size. This system also provides feedback by changing the color of the model to
indicate the correctness of user’s motion. This way, the feedback can easily be recognized by
the novice, thus making motion learning more effective.

[1] C.-J. Su, “Personal Rehabilitation Exercise Assistant with Kinect and Dynamic Time Warping,” Int. J. Inf. Educ. Technol., vol. 3, no. 4, p. 448, 2013.
[2] P. Tormene, T. Giorgino, S. Quaglini, and M. Stefanelli, “Matching incomplete time series with dynamic time warping: an algorithm and an application to post-stroke rehabilitation,” Artif. Intell. Med., vol. 45, no. 1, pp. 11–34, 2009.
[3] J. Venugopalan, C. Cheng, T. H. Stokes, and M. D. Wang, “Kinect-based rehabilitation system for patients with traumatic brain injury.,” Conf. Proc. IEEE Eng. Med. Biol. Soc., vol. 2013, pp. 4625–8, 2013.
[4] T. Y. Lin, C. H. Hsieh, and J. Der Lee, “A kinect-based system for physical rehabilitation: Utilizing Tai Chi exercises to improve movement disorders in patients with balance ability,” Proc. - Asia Model. Symp. 2013 7th Asia Int. Conf. Math. Model. Comput. Simulation, AMS 2013, pp. 149–153, 2013.
[5] E. Velloso, A. Bulling, and H. Gellersen, “MotionMA: motion modelling and analysis by demonstration,” CHI ’13 Proc. SIGCHI Conf. Hum. Factors Comput. Syst., p. 1309, 2013.
[6] M. Gleicher, “Retargetting motion to new characters,” Proc. 25th Annu. Conf. Comput. Graph. Interact. Tech. - SIGGRAPH ’98, pp. 33–42, 1998.
[7] K.-J. C. K.-J. Choi and H.-S. K. H.-S. Ko, “On-line motion retargetting,” Proceedings. Seventh Pacific Conf. Comput. Graph. Appl. (Cat. No.PR00293), 1999.
[8] H. Sakoe and S. Chiba, “Dynamic Programming Algorithm Optimization for Spoken Word Recognition,” IEEE Trans. Acoust., vol. 26, no. 1, pp. 43–49, 1978.
[9] Y. Kim and C. H. Park, “Query by humming by using scaled dynamic time warping,” Proc. - 2013 Int. Conf. Signal-Image Technol. Internet-Based Syst. SITIS 2013, pp. 1–5, 2013.
[10] H. Li, X. Wan, Y. Liang, and S. Gao, “Dynamic Time Warping Based on Cubic Spline Interpolation for Time Series Data Mining,” 2014 IEEE Int. Conf. Data Min. Work., pp. 19–26, 2014.
[11] P. K. Pisharady and M. Saerbeck, “Robust gesture detection and recognition using dynamic time warping and multi-class probability estimates,” Proc. 2013 IEEE Symp. Comput. Intell. Multimedia, Signal Vis. Process. CIMSIVP 2013 - 2013 IEEE Symp. Ser. Comput. Intell. SSCI 2013, pp. 30–36, 2013.
[12] R. Srivastava and P. Sinha, “Hand Movements and Gestures Characterization Using Quaternion Dynamic Time Warping Technique,” vol. 16, no. 5, pp. 1333–1341, 2016.
[13] A. Mori, S. Uchida, R. Kurazume, R. Taniguchi, T. Hasegawa, and H. Sakoe, “Early Recognition and Prediction of Gestures Conventional Gesture Recognition Algo- rithm Based on Dynamic Programming Early Recognition Based on DP,” pp. 18–21, 2006.
[14] Microsoft, “Kinect for Windows Developer Toolkit v1.8.” [Online]. Available: http://www.microsoft.com/en-us/download/details.aspx?id=40276. [Accessed: 01-Jan-2016].
[15] “Xbox One Kinect Teardown,” 2014. [Online]. Available: https://www.ifixit.com/Teardown/Xbox+One+Kinect+Teardown/19725. [Accessed: 16-May-2016].
[16] M. Szymczyk, “How Does The Kinect 2 Compare To The Kinect 1.” [Online]. Available: http://zugara.com/how-does-the-kinect-2-compare-to-the-kinect-1. [Accessed: 24-May-2016].
[17] S. Sempena, N. U. Maulidevi, and P. R. Aryan, “Human action recognition using Dynamic Time Warping,” Proc. 2011 Int. Conf. Electr. Eng. Informatics, no. July, pp. 1–5, 2011.
[18] T. Takala and R. Pugliese, “Reality-based User Interface System.” [Online]. Available: http://blog.ruisystem.net/about/. [Accessed: 16-May-2016].