Please use this identifier to cite or link to this item: http://hdl.handle.net/11455/9130
DC FieldValueLanguage
dc.contributor莊家峰zh_TW
dc.contributorChia-Feng Juangen_US
dc.contributor.author王柏軒zh_TW
dc.contributor.authorWang, Po-Hsuanen_US
dc.contributor.other電機工程學系所zh_TW
dc.date2013en_US
dc.date.accessioned2014-06-06T06:42:43Z-
dc.date.available2014-06-06T06:42:43Z-
dc.identifierU0005-1608201315021000en_US
dc.identifier.citation[1] J. S. Jang, “ANFIS: Adaptive-network-based fuzzy inference system,” IEEE Trans. Syst., Man, Cybern., vol. 23, no. 3, pp. 665–685, May 1993. [2] C. F. Juang and C. T. Lin, “An on-line self-constructing neural fuzzy inference network and its applications,” IEEE Trans. Fuzzy Systems, vol. 6, no. 1, pp. 12-32, Feb. 1998. [3] N. K. Kasabov and Q. Song, “DENFIS: Dynamic evolving neural-fuzzy inference system and its application for time-series prediction,” IEEE Trans. on Fuzzy Systems, vol. 10, no. 2, pp. 144 -154, April 2002. [4] P.P. Angelov and D. P. Filev, “An approach to online identification of Takagi-Sugeno fuzzy models,” IEEE Trans. on Systems, Man and Cybernetics, Part B: Cybernetics, vol. 34, no. 1, pp. 484-498, 2004. [5] E. D. Lughofer, “FLEXFIS: A robust incremental learning approach for evolving Takagi-Sugeno fuzzy models,” IEEE Trans. Fuzzy Systems, vol. 16, no. 6, pp. 1393-1410, Dec. 2008. [6] J. D. Rubio, “SOFMLS: Online self-organizing fuzzy modified least-squares network,” IEEE Trans. Fuzzy Systems, vol. 17, no. 6, pp. 1296-1309, Dec. 2009. [7] C. F. Juang, T. C. Chen, and W. Y. Cheng, “Speedup of implementing fuzzy neural networks with high-dimensional inputs through parallel processing on graphic processing units,” IEEE Trans. Fuzzy Systems, vol. 19, no. 4, pp. 717-728, Aug. 2011. [8] P. Angelov, “Fuzzily connected multi-model systems evolving autonomously from data streams,” IEEE Trans. Syst., Man, and Cyber. -Part B, Cyber., vol.41, no. 4, pp. 898-910, Aug. 2011. [9] J. M. Mendel, Uncertain Rule-Based Fuzzy Logic System: Introduction and New Directions, Prentice Hall, Upper Saddle River, NJ2001 [10] H. Hagras, “A hierarchical type-2 fuzzy logic control architecture for autonomous mobile robots,” IEEE Trans. Fuzzy Systems, vol. 12, no. 524-539, 2004. [11] C. F. Juang and C. H. Hsu, “Reinforcement ant optimized fuzzy controller for mobile-robot wall-following control,” IEEE Trans. Industrial Electronics, vol. 56, no. 10, pp. 3931-3940, Oct. 2009. [12] Q. Liang and J. M. Mendel, “Equalization of nonlinear time-varying channels using type-2 fuzzy adaptive filters,” IEEE Trans. Fuzzy systems, vol. 8, no. 551-563, 2000. [13] J. Zeng and Z. Q. Liu, “Type-2 fuzzy hidden Markov models and their application to speech recognition,” IEEE Trans. Fuzzy Systems, vol. 14, no. 3, pp. 454-467, June 2006. [14] C. H. Lee, Y. C. Lin, and W. Y. Lai, “Systems identification using type-2 fuzzy neural network (Type-2 FNN) systems,” Proc. IEEE Int. Symp. Computational Intelligence in Robotics and Automation, vol. 3, pp. 1264-1269, 2003. [15] J. M. Mendel, “Computing derivatives in interval type-2 fuzzy logic system,” IEEE Trans. On Fuzzy Systems, vol. 12. no. 1, pp. 84-98, Feb. 2004. [16] J. R. Castro, O. Castillo, P. Melin, A. Rodriguez-Diaz, “A hybrid learning algorithm for a class of interval type-2 fuzzy neural networks,” Information Sciences, vol. 179, pp. 2175-2193, 2009. [17] O. U ncu and . B. , “Discrete interval type-2 fuzzy system models using uncertainty in learning parameters,” IEEE Trans. Fuzzy Systems, vol. 15, no. 1, pp. 90-106, Feb. 2007. [18] C. F. Juang and Y. W. Tsao, “A self-evolving interval type-2 fuzzy neural network with on-line structure and parameter learning,” IEEE Trans. Fuzzy Systems, vol. 16, no. 6, pp. 1411-1424, Dec. 2008. [19] C. F. Juang and Y. W. Tsao, “A type-2 self-organizing neural fuzzy system and its FPGA implementation,” IEEE Trans. Syst., Man, and Cyber., Part B: Cybernetics, vol. 38, no. 6, pp. 1537-1548, Dec. 2008. [20] S. M. Zhou, J. M. Garibaldi, R. I. John, and F. Chiclana, “On constructing parsimonious type-2 fuzzy logic systems via influential rule selection,” IEEE Trans. Fuzzy Systems, vol. 17, no. 3, pp. 654-667, June 2009. [21] W. Y. Cheng and C. F. Juang*, “A fuzzy model with online incremental SVM and margin-selective gradient descent learning for classification problems,” accepted to be published in IEEE Trans. Fuzzy Systems, April 2013. [22] C. F. Juang, S. H. Shiu, and S. J. Shiu, “Fuzzy system learned through fuzzy clustering and support vector machine for human skin color segmentation,” IEEE Trans. Syst., Man, and Cyber., Part A: Systems and Humans, vol. 37, no. 6, pp. 1077-1087, 2007. [23] V. Vapnik, “The Nature of Statistical Learning Theory,” New York : Springer - Verlag, 1995. [24] N. Cristianini and J. S.-Taylor, “An Introduction to Support Vector Machines And Other Kernel-based Learning Methods,” Cambridge University Press, 2000. [25] O. Chapelle, P. Haffner, and V. Vapnik, “Support vector machines for histogram-based image classification,” IEEE Trans. Neural Networks, vol. 10, pp. 1055-1064, 1999. [26] B. Sch lkopf, K. Sung, C. Burges, F. Girosi, P. Niyogi, T. Poggio, and V. Vapnik, “Comparing support vector machines with Gaussian kernels to radial basis function classifiers,” IEEE Trans. Signal Processing, vol. 45, pp. 2758-2765, 1997. [27] Y. Kim and H. Ling, “Human Activity Classification Based on Micro-Doppler Signatures Using a Support Vector Machine,” IEEE Trans. Geoscience and Remote Sensing, vol. 47, no. 5, pp. 1328 – 1337, 2009. [28] S. Essid, G. Richard, and B. David, “Musical instrument recognition by pairwise classification strategies,” IEEE Trans. Audio, Speech, and Language Processing, vol. 14, no. 4, pp. 1401 – 1412, 2006 [29] C. F. Juang, Y. Y. Lin, and R. B. Huang, “Dynamic system modeling using a recurrent interval-valued fuzzy neural network and its hardware implementation,” Fuzzy Sets and Systems, vol. 179, no. 1, pp. 83-99, Sep. 2011. [30] C. F. Juang and Hong-Jhang Jiang, “Stereo Camera-based Real-Time 3D Virtual Human Construction with Interactive Entertainment Applications,”, July 2012. [31] A. Orriols-Puig, J. Casillas, and E. Bernado-Mansilla, “Genetic-based machine learning systems are competitive for pattern recognition,” Evol. Intel., vol. 1, pp. 209-232, 2008. [32] C. C. Chang and C. J. Lin, “A library for support vector machines LIBSVM,” [Online]. Available: http://www.csie.ntu.edu.tw/~cjlin/libsvm/index.html. [33] C. F. Juang and Kai-Jie Juang, “Reduced Interval Type-2 Neural Fuzzy System Using Weighted Bound Set Boundaries for Computation Speedup and FPGA Implementation,”, July 2012. [34] C. F. Juang and K. J. Juang, “Reduced interval type-2 neural fuzzy system using weighted bound-set boundary operation for computation speedup and chip implementation,” IEEE Trans. Fuzzy Systems, vol. 21, no. 3, pp. 477-491, June 2013.en_US
dc.identifier.urihttp://hdl.handle.net/11455/9130-
dc.description.abstract本論文提出精簡區間第二型類神經模糊分類器,前後件部分別使用邊界篩選梯度下降(MSGD)以及支持向量機學習。此分類器簡稱為RIT2BFC-MS.。RIT2BFC-MS前件部使用區間第二型模糊集合(IT2FSs),後件部為零階Takagi-Sugeno-Kang (TSK)形式,可採用單一數值或區間值。RIT2BFC-MS使用加權集合邊界去簡化降階運算且減少軟體的訓練時間。RIT2BFC-MS透過線上架構學習。前件部參數學習透過MSGD去減少訓練樣本與避免過度訓練。後件部參數學習透過支持向量機去賦予分類器高辨識能力。RIT2BFC-MS應用在基於視覺人體姿態分類問題。此分類系統使用兩組攝影機且從人體剪影擷取特徵值去分類四種姿態,站、彎、坐、躺。特徵值包含從人體乾淨剪影做水平與垂直投影的離散傅立葉轉換(DFT)係數與人體長寬比值。此外,將包住人體剪影的最小正方形網格均勻切割,剪影在每個均勻分割網格的面積比例做為特徵值。RIT2BFC-MS的分類性能透過姿態分類問題和幾個基準分類問題跟各個第一型模糊系統比較得到驗證。zh_TW
dc.description.abstractThis thesis proposes a reduced interval type-2 neural fuzzy classifier with margin-selective gradient descent (MSGD) and support vector learning for antecedent and consequent part learning, respectively, and the classifier is called RIT2BFC-MS. The antecedent part in the RIT2BFC-MS uses interval type-2 fuzzy sets (IT2FSs) and the consequent part uses zero-order Takagi-Sugeno-Kang (TSK)-type rules with crisp or interval values. The RIT2BFC-MS uses weighted bound-set boundaries to simplify type-reduction operation and reduce training time. The RIT2BFC-MS is built through online structure learning. The antecedent parameters are tuned through MSGD learning to reduce training samples and avoid overtraining. The consequent parameters are learned through support vector machine to endow the classifier high generalization ability. The RIT2BFC-MS is applied to a vision-based human body posture classification problem. The classification system uses two cameras and novel features extracted from segmented human silhouettes to classify the four postures of standing, bending, sitting, and lying. The novel features include discrete Fourier transform (DFT) coefficients calculated from horizontal and vertical projections of the body silhouette and length-width ratio of human body. In addition, the minimum enclosing square of the body silhouette is uniformly partitioned and the area ratio of the silhouette in each partitioned grid is calculated used as another set of features. Classification performance of the RIT2BFC-MS is verified through the posture classification problem and several benchmark classification problems and comparisons with various type-1 and type-1 fuzzy classifiers.en_US
dc.description.tableofcontentsContent Acknowledgments i Chinese Abstract ii English Abstract iii Content iv List of Figures vi List of Tables ii Chapter 1 Introduction 1 1.1 Survey And Literature Review 1 1.2 Organization of the Thesis 3 Chapter 2 RIT2BFC-MS with Crisp Consequents 4 2.1 RIT2BFC-MS(C) Functions 4 2.2 Structure Learning 7 2.3 Parameter Learning 8 2.3.1 Consequent Parameter Learning 8 2.3.2 Antecedent Parameter Learning 9 2.3.3 The Overall Parameter Learning Algorithm 13 Chapter 3 RIT2BFC-MS with Interval Consequents 15 3.1 RIT2BFC-MS(I) Functions 15 3.2 Structure and Parameter Learning 18 3.2.1 Structure and Consequent Parameter Learning 18 3.2.2 Antecedent and The Overall Parameter Learning Algorithm 20 Chapter 4 Feature Extraction and Recognition of Human Body Posture 24 4.1 Feature Extraction 24 4.2 Training and Recognition by RIT2BFC-MS 28 Chapter 5 Simulations 30 5.1 Classification Performance Verification 30 Example 1:Svmguide1 date set...……………………………………………….…30 Example 2:Splice date set….......……………………………………………….…32 Example 3:Heart-s date set……………………………………………………......35 Example 4:Tao date set….......………………………………………………….…36 5.2 Human Body Posture Recognition 38 Chapter 6 Conclusion 42 References 43 List of Figures Figure 2. 1 An interval type-2 fuzzy set with an uncertain 4 Figure 2. 2 Distributions of the reduced fuzzy set , the two bound sets, and the defuzzified output in the RIT2BFC-MS(C). 7 Figure 2. 3 The partitioned regions MO and MS and distributions of misclassified training samples, where samples in the MO region are moved using the MSGA toward to correct directions. 10 Figure 2. 4 The RIT2BFC-MS(C) parameter training process 14 Figure 4. 1 The length and width calculation of a segmented human body 25 Figure 4. 2 Process chart of Grid-Partition 27 Figure 4. 3 Grid-Partition results 27 Figure 4. 4 The training procedure RIT2BFC-MS using one-against-all 28 Figure 4. 5 The test procedure RIT2BFC-MS using one-against-all 29 Figure 5. 1 Illustrative segmented results of the four main postures, standing, lying 39 List of Tables Table 5. 1 Test rates of different fuzzy classifiers with four rules in Example 1 31 Table 5. 2 Test rates of different fuzzy classifiers with four rules and different levels of White Gaussian noise in Example 1 31 Table 5. 3 Test rates of different fuzzy classifiers with ten rules in Example 1 32 Table 5. 4 Test rates of different fuzzy classifiers with ten rules and different levels of White Gaussian noise in Example 1 32 Table 5. 5 Test rates of different fuzzy classifiers with four rules in Example 2 33 Table 5. 6 Test rates of different fuzzy classifiers with four rules and different levels of White Gaussian noise in Example 2 34 Table 5. 7 Test rates of different fuzzy classifiers with five rules in Example 2 34 Table 5. 8 Test rates of different fuzzy classifiers with five rules and different levels of White Gaussian noise in Example 2 35 Table 5. 9 Test rates of different fuzzy classifiers with four rules in Example 3 36 Table 5. 10 Test rates of different fuzzy classifiers with four rules and different levels of White Gaussian noise in Example 3 36 Table 5. 11 Test rates of different fuzzy classifiers with four rules in Example 4 37 Table 5. 12 Test rates of different fuzzy classifiers with four rules and different levels of White Gaussian noise in Example 4 38 Table 5. 13 Test results of different fuzzy classifiers with 400 training data and five rules 40 Table 5. 14 Test results of different fuzzy classifiers with 400 training data and ten rules 40 Table 5. 15 Test results of different fuzzy classifiers with 1200 training data and five rules 41 Table 5. 16 Test results of different fuzzy classifiers with 1200 training data and ten rules 41zh_TW
dc.language.isoen_USen_US
dc.publisher電機工程學系所zh_TW
dc.relation.urihttp://www.airitilibrary.com/Publication/alDetailedMesh1?DocID=U0005-1608201315021000en_US
dc.subject類神經模糊分類器zh_TW
dc.subjectNeural Fuzzy Classifersen_US
dc.subject支持向量機學習zh_TW
dc.subject姿態辨識zh_TW
dc.subjectSupport Vector Machine Learningen_US
dc.subjectBody Posture Recongnitionen_US
dc.title使用支持向量學習之第二型類神經模糊分類器及人體姿態辨識應用zh_TW
dc.titleReduced Type-2 Neural Fuzzy Classifers With Support Vector Learning And Human Body Posture Recongnition Applicationsen_US
dc.typeThesis and Dissertationzh_TW
Appears in Collections:電機工程學系所
文件中的檔案:

取得全文請前往華藝線上圖書館



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.