自從色彩-深度感測器問世後，此種感測器便被應用於運動訓練系統中，透過感測器記錄使用者的運動過程，並從其中擷取出人體骨架資訊，藉由重播回顧其運動狀況或骨架資訊，使用者或電腦程式能夠判斷使用者之瞬間姿態標準與否。此種評估方式並不能適當地表現出使用者之姿態與時間關係，因此本研究提出一軟體架構（１）希望能透過此架構讓專業人士能夠建立針對連續動作之標準動作，並以此建立自動評估系統（２）使用者做相同動作後，可和標準動作比對。
此架構將專業人士示範動作之人體骨架資訊經由適當之前處理以及特徵擷取後，藉由分群演算法轉換出連續動作之序列，再藉由此動作序列訓練對應動作單元之隱馬可夫模型。
使用者則將其練習過程以色彩-深度感測器記錄，並透過同樣方式產生完整運動過程之序列，並使用適當方法分割為動作單元後逐一比對每個完成訓練之隱馬可夫模型，自動評估出受訓練者之運動流程與專業人士示範接近與否，得到回饋後反覆練習以達成訓練之目的，令使用者能夠如預期改善其運動行為。

Since the launch of RGB-D sensors, these sensors are applied to exercise training systems. They are used to record users’ exercise processes and extract human skeletal data. By monitoring/reviewing users’ exercise video or skeletal data, users or a computer program could check if the poses are correct, especially for key poses. This assessment of a key pose does not appropriately present the relationship between a user’s posture and time. This research proposes a software framework (1) for professionals to build standard reference key poses of some exercise, (2) users would perform the same exercise, and the system automatically performs assessment.
This framework transforms the professionals’ demonstration into sequences of continuous movements through preprocessing, feature extracting and a clustering algorithm. These sequences of continuous movements become training data sources of Hidden Markov Models that correspond to each movement primitive.
A user records his/her training process by RGB-D sensors, and through the same way above to generate sequences of the entire training process. These sequences are segmented into movement primitives, and compared to each trained HMMs. Thereby automatically assess if the training process is close to the professional’s demonstration. After viewing the feedback of training process and practicing repeatedly to reach the goal of training, the user is expected to gain improvements in the exercise.

[1] C.-C. Chen, Markerless Exercise Training Toolkit Using Multiple RGB-D based Sensors, Master Thesis, Department of Engineering Science, National Cheng Kung University, Tainan, Taiwan, 2015.
[2] T.-L. Cheng, An IoT-Based Exercise Training System, Master Thesis, Department of Engineering Science, National Cheng Kung University, Tainan, Taiwan, 2016.
[3] F. Liu, Y. Wang, Q. Wang, L. Zhang and W. Zeng, "A new gait recognition method using kinect via deterministic learning," in 2016 12th World Congress on Intelligent Control and Automation (WCICA), Guilin, China, 12-15 Jun. 2016.
[4] W. Z. W. Z. Abiddin, R. Jailani, A. R. Omar and I. M. Yassin, "Development of MATLAB Kinect Skeletal Tracking System (MKSTS) for gait analysis," in 2016 IEEE Symposium on Computer Applications & Industrial Electronics (ISCAIE), Batu Feringghi, Malaysia, 30-31 May 2016.
[5] Z. Huang, A. Nagata, M. Kanai-Pak, J. Maeda, Y. Kitajima, M. Nakamura, K. Aida, N. Kuwahara, T. Ogata and J. Ota, "Self-Help Training System for Nursing Students to Learn Patient Transfer Skills," IEEE Transactions on Learning Technologies, vol. 7, no. 4, pp. 319-332, Oct.-Dec. 2014.
[6] Y.-Y. Lin, A Cloud Platform with Decentralized and Secure Mechanisms for Managing Individual Files, Master Thesis, Department of Engineering Science, National Cheng Kung University, Tainan, Taiwan, 2016.
[7] Y.-C. Chen, H.-J. Lee and K.-H. Lin, "Measurement of body joint angles for physical therapy based on mean shift tracking using two low cost Kinect images," in 2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), Milan, Italy, 25-29 Aug. 2015.
[8] S.-H. Han, H.-G. Kim and H.-J. Choi, "A Novel Concept of the Rehabilitation Training Coach Robot for Patients with Disability," in 2017 18th IEEE International Conference on Mobile Data Management (MDM), Daejeon, South Korea, 29 May-1 Jun. 2017.
[9] S.-H. Han, H.-G. Kim and H.-J. Choi, "Rehabilitation posture correction using deep neural network," in 2017 IEEE International Conference on Big Data and Smart Computing (BigComp), Jeju, South Korea, 13-16 Feb. 2017.
[10] J. Hermus, C. Hays, M. Adamski, H. Lider, J. Westlund, A. Scholp, J. Webster and B. Buehring, "Posture monitor for vibration exercise training," in 2015 IEEE Great Lakes Biomedical Conference (GLBC), Milwaukee, WI, USA, 14-17 May 2015.
[11] L. Liu, M. Yin and S. Mehrotra, "Preventing venous thromboembolism in patients admitted by hospital using Microsoft Kinect V2," in 2015 IEEE Biomedical Circuits and Systems Conference (BioCAS), Atlanta, GA, USA, 22-24 Oct. 2015.
[12] N. A. Azis, Y.-S. Jeong, H.-J. Choi and Y. Iraqi, "Weighted averaging fusion for multi-view skeletal data and its application in action recognition," IET Computer Vision, vol. 10, no. 2, pp. 134-142, Mar. 2016.
[13] J. F.-S. Lin, M. Karg and D. Kulić, "Movement Primitive Segmentation for Human Motion Modeling: A Framework for Analysis," IEEE Transactions on Human-Machine Systems, vol. 46, no. 3, pp. 325-339, Jun. 2016.
[14] D. Wu, L. Pigou, P.-J. Kindermans, N. D.-H. Le, L. Shao, J. Dambre and Jean-Marc, "Deep Dynamic Neural Networks for Multimodal Gesture Segmentation and Recognition," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 38, no. 8, pp. 1583-1597, 1 Aug. 2016.
[15] L. Zhang, Z. Cheng, Y. Gan, G. Zhu, P. Shen and J. Song, "Fast human whole body motion imitation algorithm for humanoid robots," in 2016 IEEE International Conference on Robotics and Biomimetics (ROBIO), Qingdao, China, 3-7 Dec. 2016.
[16] Heresy's Space, "OpenNI 2 Introduction," [Online]. Available: https://kheresy.wordpress.com/2012/12/21/basic-openni-2/. [Accessed Jun. 2017].
[17] L. E. Baum and T. Petrie, "Statistical Inference for Probabilistic Functions of Finite State Markov Chains," The Annals of Mathematical Statistics, vol. 37, no. 6, pp. 1554-1563, 4 Apr. 1966.
[18] L. R. Rabiner, "A tutorial on hidden Markov models and selected applications in speech recognition," Proceedings of the IEEE, vol. 77, no. 2, Feb. 1989.
[19] "Baum–Welch algorithm," Wikimedia Foundation, [Online]. Available: https://en.wikipedia.org/wiki/Baum–Welch_algorithm. [Accessed Jun. 2017].
[20] V. AJ, "Error bounds for convolutional codes and an asymptotically optimum decoding algorithm," IEEE Transactions on Information Theory, vol. 13, no. 2, pp. 260-269, Apr. 1967.