本論文提出使用單一加速度計及特徵降維之類神經辨識器於人類動作辨識。我們使用一個無線三軸加速度計來收集加速度資料，並發展以類神經網路辨識器為基礎的辨識策略來辨識八種人類日常生活動作。我們發展兩種不同的方法來建構辨識器：離線建構及線上建構。
在離線建構方法中，我們使用各個擊破的方法，先把動態與靜態的動作分開後，再各自辨識這兩類所屬的動作。由於類神經網路在辨識問題中能產生複雜的區分曲面，因此我們採用類神經網路為辨識器的主要架構。此外；我們採用一個以共同主成分分析(CPCA)為基礎的監督式特徵子集合選取(FSS)方法，來定義出有意義的特徵子集合並簡化辨識器架構來達到令人滿意的辨識率。而在線上建構方法中，我們假定辨識任務是在完整訓練資料不可預先取得之即時環境下執行。我們提出一個動態線性識別分析(LDA)，在不儲存所有訓練資料下，能夠以增加及刪減模式來動態更新散佈矩陣。並用一個以動態LDA之結果所建構的模糊基底函數(FBF)辨識器來作動作辨識。
最後，我們在實際實驗中評估以離線方法及線上方法所建構的辨識器之效能。我們的實驗結果已經成功地驗證下列二點：(1)兩種方法之有效性，(2) 模糊基底函數辨識器和所提出的動態LDA之整合可減低計算負擔並可在線上增加資料/群組及刪減群組模式下達到滿意的效能。

This thesis presents neural classifiers with feature dimension reduction methods for human activity recognition using only one accelerometer. We used one wireless triaxial accelerometer to collect acceleration data and develop recognition schemes with neural-network-based classifiers to classify eight types of daily human activities. We developed two different approaches for the construction of classifiers: an offline construction and an online construction.
In the offline construction, we utilized a divide-and-conquer approach that separates dynamic activities from static activities preliminarily and recognizes these two different types of activities separately. Neural networks were adopted as the classifiers due to their capability of generating complex discriminating surfaces for recognition tasks. A supervised feature subset selection (FSS) method based on common principal component analysis (CPCA) was proposed to determine significant feature subsets and compact classifier structures with satisfactory recognition accuracy. In the online construction, we assume the recognition task is performed in a real-time environment where the complete training data may not be available beforehand. We proposed a dynamic linear discriminant analysis (LDA) which can dynamically update the scatter matrices in both incremental and descremental modes without storing all the training data in memory. A fuzzy basis function (FBF) classifier constructed by the results of the dynamic LDA was used for activity recognition.
Finally, we evaluated the performance of the classifiers constructed by the offline and online approaches in practical experiments. Our experimental results have successfully validated: 1) the effectiveness of both approaches, and 2) the integration of a FBF classifier and the proposed dynamic LDA can reduce computational burden and achieve satisfactory performance for online adding data/class and deleting existing classes.

[1] P. N. Belhumeur, J. P. Hespanha, D. J. Kriegman, “Eigenfaces vs. Fisherfaces: recognition using class specific linear projection,” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 19, no. 7, pp. 711-720, Jul. 1997.
[2] C.V.C. Bouten, K.T.M. Koekkoek, M. Verduin, R. Kodde and J.D. Janssen, “A triaxial accelerometer and portable data processing unit for the assessment of daily physical activity.” IEEE Trans. on Biomedical Engineering, vol. 44, no. 3, pp. 136-147, 1997.
[3] A. Ben-Hur, D. Horn, H.T. Siegelmann and V. Vapnik, “Support vector clustering, ” Journal of Machine Learning Research, vol. 2, pp. 125-137, 2001.
[4] J. Ben-Arie, Z. Wang, P. Pandit and S. Rajaram, “Human activity recognition using multidimensional indexing,” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 24 , no. 8, pp. 1091–1104, 2002.
[5] L. Bao and S. S. Intille, “Activity recognition from user-annotated acceleration data,” Proceedings of the 2nd International Conference on Pervasive Computing and Communications, 2004, Lecture Notes in Computer Science, vol. 3001, pp. 1-17, 2004.
[6] T. Cover, and P. Hart, “Nearest neighbor pattern classification,” IEEE Trans. on Information Theory, vol. 13, no. 1, pp. 21-27, Jan. 1967.
[7] I. Cohen, Q. Tian, X.S. Zhou and T.S. Huang, “Feature selection using principal feature analysis,” Proceedings of the 15th International Conference on Multimedia, pp. 301-304, 2007.
[8] J. Solà i Carós, O. Chételat, P. Celka and S. Dasen, “Very low complexity algorithm for ambulatory activity classification,” Proceedings of European Medical & Biological Engineering Conference and IFMBE European Conference on Biomedical Engineering, 2005.
[9] E. S. Choi, W. C. Bang, S. J. Cho, J. Yang, D. Y. Kim, and S. R. Kim, “Beatbox music phone: gesture-based interactive mobile phone using a tri-axis accelerometer,” Proceedings of IEEE International Conf. on Industrial Technology,, pp. 97-102, 2005.
[10] A. K. Dey, “Understanding and using context,” Personal Ubiquitous Computing, vol. 5, no. 1, pp. 4-7, 2001.
[11] S. Delsey, M. Marilyn, R. John and B. Paolo, “Using hierarchical clustering methods to classify motor activities of COPD patients from wearable sensor data,” Journal of NeuroEngineering and Rehabilitation, vol. 2, pp. 1-14, 2005.
[12] M. Ermes, J. Pärkkä, J. Mäntyjärvi, I. Korhonen, “Detection of daily activities and sports with wearable sensors in controlled and uncontrolled conditions,” IEEE Trans. on Information Technology in Biomedicine, vol. 12, no.1, pp. 20–26, Jan. 2008.
[13] B. N. Flury, “Common principal components in k groups,” Journal of the American Statistical Association. vol. 79, pp. 892-898, 1984.
[14] M. Fukumi, and Y. Mitsukura, “Feature generation by simple-FLDA for pattern recognition,” Proceedings of the International Conference on Computational Intelligence for Modelling, Control and Automation and International Conference on Intelligent Agents, Web Technologies and Internet Commerce, vol. 2, pp. 730-734, Nov. 2005.
[15] J. M. Hannan and J.M. Bishop, “A comparison of fast training algorithms over two real problems,” Proceedings of IEE Int. Conf. on Neural Networks. pp. 1-6, 1997.
[16] S. Haykin, Neural networks: a comprehensive foundation, 2nd Edition. Prentice Hall, 1999.
[17] Y.A. Ivanov and A.F. Bobick, “Recognition of visual activities and interactions by stochastic parsing,” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 22, no. 8, pp. 852-872, 2000
[18] C. Igel and M. Hasken, “Empirical evaluation of the improved RPROP learning algorithms,” Neurocomputing, vol. 50, pp. 105-123, 2003.
[19] A. W. Jamie, P. Lukowicz, G. Troster, and T. E. Starner, “Activity recognition of assembly tasks using body-worn microphones and accelerometers,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 28, no. 10, pp. 1553-1567, 2006.
[20] L. Jatoba, U. Grossmann, J. Ottenbacher, W. Stork, and K. Muller-Glaser, Development of a self-constructing neuro-fuzzy inference system for online classification of physical movements, 2007 9th International Conference on e-Health Networking, pp. 332-335, Jun. 2007.
[21] W. J. Krzanowski, “Between-groups comparison of principal components,” Journal of the American Statistical Association, vol. 74, no. 367, pp. 703-707, 1979
[22] M. Kudo and J. Sklansky, “Comparison of algorithms that select features for pattern classifiers,” Pattern Recognition, vol. 33, no. 1, pp. 25-41, Jan. 2000.
[23] S. Kallio, J. Kela and J. Mantyjarvi, “Online gesture recognition system for mobile interaction,” Proceedings of IEEE International Conf. on Systems, Man and Cybernetics, vol. 3, pp. 2070- 2076, 2003.
[24] A. Kocsor and L. Toth, “Kernel-based feature extraction with a speech technology application,” IEEE Trans. on Signal Processing, vol. 52, no. 8, pp. 2250-2263, Aug, 2004.
[25] D. M. Karantonis, M. R. Narayanan, M. Mathie, N. H. Lovell and B. G. Celler, “Implementation of a real-time human movement classifier using a triaxial accelerometer for ambulatory monitoring,” IEEE Trans. on Information Technology in Biomedicine, vol. 10, no. 1, pp. 156-167, 2006.
[26] T. K. Kim, S. F. Wong, B. Stenger, J. Kittler and R. Cipolla, “Incremental linear discriminant analysis using sufficient spanning set approximations,” Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, pp. 1-8, 2007.
[27] C. G. Looney, Pattern Recognition Using Neural Networks: Theory and Algorithms for Engineers and Scientists. Oxford University Press, 1997.
[28] J. Lu, K.N. Plataniotis and A.N. Venetsanopoulos, “Regularization studies on LDA for face recognition,” Proceedings of 2004 International Conference on Image Processing, vol. 1, pp. 63-66, Oct, 2004.
[29] H. Liu and L. Yu, “Toward integrating feature selection algorithms for classification and clustering,” IEEE Trans. on Knowledge and Data Engineering, vol. 17, no. 4, pp. 491-502, Apr. 2005.
[30] T. S. Li, “Feature selection for classification by using a GA-based neural network approach,” Journal of the Chinese Institute of Industrial Engineers, vol. 23, no. 1, pp. 55-64, 2006.
[31] J. Mantyjarvi, J. Himberg, and T. Seppanen, “Recognizing human motion with multiple acceleration sensors,” Proceedings of IEEE Int. Conf. Systems, Man, and Cybernetics, vol. 2, pp. 747-752, 2001.
[32] M. J. Mathie, N. H. Lovell, A. C. F. Coster and B. G. Celler, “Determining activity using a triaxial accelerometer,” Proc. 2nd Joint EMBS-BMES Conf, vol. 3, pp. 2481-2482, 2002.
[33] P. Mitra, C. A. Murthy and Sankar K. Pal, “Unsupervised feature selection using feature similarity,” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 24, no. 3, pp. 301- 312, 2002.
[34] M. J. Mathie, B. G. Celler, N. H. Lovell and A. C. F. Coster, “Classification of basic daily movements using a triaxial accelerometer,” Medical and Biological Engineering and Computing, vol. 42, no. 5, pp. 679-687, Sep. 2004.
[35] U. Maurer, A. Smailagic, D. P. Siewiorek and M. Deisher, “Activity recognition and monitoring using multiple sensors on different body positions,” Proceedings of International Workshop on Wearable and Implantable Body Sensor Networks, pp. 113-116, 2006.
[36] B. Najafi, K. Aminian, A. Paraschiv-Ionescu, F. Loew, C.J. Bula and P. Robert, “Ambulatory system for human motion analysis using a kinematic sensor: monitoring of daily physical activity in the elderly,” IEEE Trans. on Biomedical Engineering, Laboratory of Motion Analysis and Measurement, vol. 50, no. 6, pp. 711-723, 2003.
[37] B. A. Pearlmutter, “Gradient calculations for dynamic recurrent neural networks: a survey,” IEEE Trans. Neural Networks, vol. 6, no. 5, pp. 1212-1228, 1995.
[38] S. Pang, S. Ozawa, and N. Kasabov, “Incremental linear discriminant analysis for classification of data streams,” IEEE Transactions on Systems, Man and Cybernetics, Part B, vol. 35, no. 5, pp. 905-914, 2005.
[39] M. Riedmiller and H. Braun, “Direct adaptive method for faster backpropagation learning: the RPROP algorithm,” Proceedings of IEEE Int. Conf. Neural Networks, vol. 1, pp. 586-591, 1993.
[40] N. Ravi, N. Dandekar, P. Mysore and M. L. Littman, “Activity recognition from accelerometer data,” Proceedings of the Seventeenth Innovative Applications of Artificial Intelligence Conference, pp. 11-18, 2005.
[41] S. R. Safavian and D. Landgrebe, “A survey of decision tree classifier methodology,” IEEE Trans. on Systems, Man and Cybernetics, vol. 21, no. 3, pp. 660-674, 1991.
[42] M. Sekine, T. Tamura, M. Akay, T. Fujimoto, T. Togawa and Y. Fukui, “Discrimination of walking patterns using wavelet-based fractal analysis,” IEEE Trans. on Neural Systems and Rehabilitation Engineering, vol. 10, no. 3, pp. 188- 196, Sep 2002.
[43] K. T. Song and Y. Q. Wang, “Remote activity monitoring of the elderly using a two-axis accelerometer,” Proceedings of 2005 CACS Automatic Control Conference Tainan, Taiwan, pp. 18-23, 2005.
[44] A. Tsymbal, S. Puuronen, M. Pechenizkiy, M. Baumgarten and D. Patterson, “Eigenvector-based feature extraction for classification,” Proceedings of the Fifteenth International Florida Artificial Intelligence Research Society Conference, pp. 354-358, 2002.
[45] H. C. C. Tan, J. Kui, and C. D. Silva, “Human activities recognition by head movement using partial recurrent neural network,” Proceedings of SPIE Int. Conf. Visual Communications and Image, vol. 5150, pp. 2007-2014, 2003.
[46] M. Uray, D. Skocaj, P. M. Roth, H. Bischof and A. Leonardis, “Incremental LDA Learning by Combining Reconstructive and Discriminative Approaches,” Proceedings of the 18th British Machine Vision Conference (BMVC 2007), 2007.
[47] L.X. Wang and J.M. Mendel, “Fuzzy basis functions, universal approximation, and orthogonal least-squares learning,” IEEE Trans. on Neural Networks, vol. 3, no. 5, pp. 807–814, 1992.
[48] T. Watanabe, S. Yamagishi, H. Murakami, N. Furuse and N. Hoshimiya, “Recognition of lower limb movements by artificial neural network for eestoring gait of Hemiplegic patients by functional electrical stimulation,” Engineering in Medicine and Biology Society, 2001. Proceedings of the 23rd Annual International Conference of the IEEE, vol.2, pp. 1348-1351, 2001.
[49] S. Wang, J. Yang, N. Chen, X. Chen, and Q. Zhang, “Human activity recognition with user-free accelerometers in the sensor networks,” Proceedings of IEEE Int. Conf. Neural Networks and Brain, vol. 2, pp. 1212- 1217, 2005.
[50] Yole Développement, “Inertial MEMS markets for consumer electronics applications,” 2007, http://www.yole.fr/pagesAn/products/Report_sample/WISM.pdf
[51] J. Ye, R. Janardan, C.H. Park and H. Park, “An optimization criterion for generalized discriminant analysis on undersampled problems,” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 26, no. 8, pp. 982 – 994, Aug. 2004.
[52] J. Ye, Q. Li, H. X., H. Park, R. Janardan, and V. Kumar, “IDR/QR: an incremental dimension reduction algorithm via QR decomposition,” IEEE Transactions on Knowledge and Data Engineering, vol. 17, no. 9, pp. 1208-1222, 2005.
[53] H. Yoon, K. Yang and C. Shahabi, “Feature subset selection and feature ranking for multivariate time series,” IEEE Trans. on Knowledge and Data Engineering, vol.17, no.9, pp. 1186-1198,2005
[55] J. Yan, B. Zhang, N. Liu, S. Yan, Q. Cheng, W. Fan, Q. Yang, W. Xi and Z. Chen, “Effective and efficient dimensionality reduction for large-scale and streaming data preprocessing,” IEEE Trans. on Knowledge and Data Engineering, vol. 18, no. 3, pp. 320- 333, March, 2006.
[56] S. Zhang and T. Sim, “When Fisher meets Fukunaga-Koontz: A new look at linear discriminants,” Proceedings of 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, vol. 1, pp. 323-329, 2006.