13,882 research outputs found

    Yue yu de da ci hui, lian xu yu yin shi bie xi tong

    No full text
    Wong Yiu Wing = 粤語的大詞彙、連續語音識別系統 / 黃耀榮.Thesis (M.Phil.)--Chinese University of Hong Kong, 2000.Includes bibliographical references.Text in English; abstracts in English and Chinese.Wong Yiu Wing = Yue yu de da ci hui, lian xu yu yin shi bie xi tong / Huang Yaorong.Chapter 1 --- Introduction --- p.1Chapter 1.1 --- Progress of Large Vocabulary Continuous Speech Recognition for Chinese --- p.2Chapter 1.2 --- Objectives of the Thesis --- p.5Chapter 1.3 --- Thesis Outline --- p.6Reference --- p.7Chapter 2 --- Fundamentals of Large Vocabulary Continuous Speech Recognition for Cantonese --- p.9Chapter 2.1 --- Characteristics of Cantonese --- p.9Chapter 2.1.1 --- Cantonese Phonology --- p.9Chapter 2.1.2 --- Written Cantonese versus Spoken Cantonese --- p.12Chapter 2.2 --- Techniques for Large Vocabulary Continuous Speech Recognition --- p.13Chapter 2.2.1 --- Feature Representation of the Speech Signal --- p.14Chapter 2.2.2 --- Hidden Markov Model for Acoustic Modeling --- p.15Chapter 2.2.3 --- Search Algorithm --- p.17Chapter 2.2.4 --- Statistical Language Modeling --- p.18Chapter 2.3 --- Discussions --- p.19Reference --- p.20Chapter 3 --- Acoustic Modeling for Cantonese --- p.21Chapter 3.1 --- The Speech Database --- p.21Chapter 3.2 --- Context-Dependent Acoustic Modeling --- p.22Chapter 3.2.1 --- Context-Independent Initial / Final Models --- p.23Chapter 3.2.2 --- Construction of Context-Dependent TrilF Models from Context- Independent IF Models --- p.26Chapter 3.2.3 --- Data Sharing in Acoustic Modeling --- p.27Chapter 1. --- Sparse Data Problem --- p.27Chapter 2. --- Decision-Tree Based State Clustering --- p.28Chapter 3.3 --- Experimental Results --- p.31Chapter 3.4 --- Error Analysis and Discussions --- p.33Chapter 3.4.1 --- Recognition Accuracy vs. Model Complexity --- p.33Chapter 3.4.2 --- Initial / Final Confusion Matrices --- p.34Chapter 3.4.3 --- Analysis of Phonetic Trees --- p.39Chapter 3.4.4 --- The NULL Initial HMM --- p.42Chapter 3.4.5 --- Comments on the CUSENT Speech Corpus --- p.42References --- p.44Chapter 4 --- Language Modeling for Cantonese --- p.46Chapter 4.1 --- N-gram Language Model --- p.46Chapter 4.1.1 --- Problems in Building an N-gram Language Model --- p.47Chapter 1. --- The Zero-Probability Problem and Backoff N-gram --- p.48Chapter 4.1.2 --- Perplexity of a Language Model --- p.49Chapter 4.2 --- N-gram Modeling in Cantonese --- p.50Chapter 4.2.1 --- The Vocabulary and Word Segmentation --- p.50Chapter 4.2.2 --- Evaluation of Chinese Language Models --- p.53Chapter 4.3 --- Character-Level versus Word-Level Language Models --- p.54Chapter 4.4 --- Language Modeling in a Specific Domain --- p.57Chapter 4.4.1 --- Language Model Adaptation to the Financial Domain --- p.57Chapter 1. --- Vocabulary Refinement --- p.57Chapter 2. --- The Seed Financial Bigram --- p.58Chapter 3. --- Linear Interpolation of Two Bigram models --- p.59Chapter 4. --- Performance of the Interpolated Language Model --- p.60Chapter 4.5 --- Error Analysis and Discussions --- p.61References --- p.63Chapter 5 --- Integration of Acoustic Model and Language Model --- p.65Chapter 5.1 --- One-Pass Search versus Multi-Pass Search --- p.66Chapter 5.2 --- A Two-Pass Decoder for Chinese LVCSR --- p.68Chapter 5.2.1 --- The First Pass Search --- p.69Chapter 5.2.2 --- The Second Pass Search --- p.72Chapter 5.3 --- Experimental Results --- p.73Chapter 5.4 --- Error Analysis and Discussions --- p.75Chapter 5.4.1 --- Vocabulary and Search --- p.75Chapter 5.4.2 --- Expansion of the Syllable Lattice --- p.76Chapter 5.4.3 --- Perplexity and Recognition Accuracy --- p.78Reference --- p.80Chapter 6 --- Conclusions and Suggestions for Future Work --- p.82Chapter 6.1 --- Conclusions --- p.82Chapter 6.2 --- Suggestions for future work --- p.84Chapter 1. --- Speaker Adaptation --- p.84Chapter 2. --- Tone Recognition --- p.84Reference --- p.85Appendix I Base Syllable Table --- p.86Appendix II Phonetic Question Set --- p.8

    Difference and dispersion: Educational research in a postmodern context

    No full text
    Difference and Dispersion is the fourth in a series of annual research papers produced by doctoral students from The Graduate School of Education, The University of Queensland, following their presentation at the School’s annual Postgraduate Research Conference in Education.\ud \ud The work featured herein celebrates the diversity of cultural and disciplinary backgrounds of education researchers who come from as far afield as Germany, Hong Kong, China, Nigeria, Russia, Singapore, Thailand and of course different parts of Australia. In keeping with a postmodern epistemology, ‘difference’ and ‘dispersion’ are key themes in apprehending the multiplicity of their research topics, methodologies, methods and speaking/writing positions.\ud \ud From widely differing contexts and situations, these writers address the consequences, implications and possibilities for education at the beginning of the third millennium. Their interest ranges from location-specific issues in schools and classrooms, change in learning contexts and processes, educational discourses and relations of power in diverse geographical settings, and the differing articulations of the local and the global in situated policy contexts. \ud \ud Conceived and developed in a spirit of ongoing dialogue with and insight to alternative views and visions of education and society, this edited collection exemplifies the quality in diversity and the high levels of scholarship and supervision at one of Australia’s finest Graduate Schools of Education

    CUHK electronic theses & dissertations collection

    No full text
    Yiu, Wing Ho.Thesis M.Phil. Chinese University of Hong Kong 2015.Includes bibliographical references (leaves 27-30).Abstracts also in Chinese.Title from PDF title page (viewed on 03, January, 2017)

    Yu zhou she xian miao zi ji qi yin zhi zhong zi de yan jiu

    No full text
    Tam, Yiu Ho = 宇宙射線渺子及其引致中子的研究 / 譚耀豪.Thesis (Ph.D.)--Chinese University of Hong Kong, 2013.Includes bibliographical references (leaves 148-153).Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web.Abstracts also in Chinese.Tam, Yiu Ho = Yu zhou she xian miao zi ji qi yin zhi zhong zi de yan jiu / Tan Yaohao

    Optimal design of a composite wing structure for a flying-wing aircraft subject to multi-constraint

    Full text link
    This thesis presents a research project and results of design and optimization of a composite wing structure for a large aircraft in flying wing configuration. The design process started from conceptual design and preliminary design, which includes initial sizing and stressing followed by numerical modelling and analysis of the wing structure. The research was then focused on the minimum weight optimization of the /composite wing structure /subject to multiple design /constraints. The modelling, analysis and optimization process has been performed by using the NASTRAN code. The methodology and technique not only make the modelling in high accuracy, but also keep the whole process within one commercial package for practical application. The example aircraft, called FW-11, is a 250-seat commercial airliner of flying wing configuration designed through our MSc students Group Design Project (GDP) in Cranfield University. Started from conceptual design in the GDP, a high-aspect-ratio and large sweepback angle flying wing configuration has been adopted. During the GDP, the author was responsible for the structural layout design and material selection. Composite material has been chosen as the preferable material for both the inner and outer wing components. Based on the derivation of structural design data in the conceptual phase, the author continued with the preliminary design of the outer wing airframe and then focused on the optimization of the composite wing structure. Cont/d

    Optimal design of a flying-wing aircraft inner wing structure configuration

    Full text link
    Flying-wing aircraft are considered to have great advantages and potentials in aerodynamic performance and weight saving. However, they also have many challenges in design. One of the biggest challenges is the structural design of the inner wing (fuselage). Unlike the conventional fuselage of a tube configuration, the flying-wing aircraft inner wing cross section is limited to a noncircular shape, which is not structurally efficient to resist the internal pressure load. In order to solve this problem, a number of configurations have been proposed by other designers such as Multi Bubble Fuselage (MBF), Vaulted Ribbed Shell (VLRS), Flat Ribbed Shell (FRS), Vaulted Shell Honeycomb Core (VLHC), Flat Sandwich Shell Honeycomb Core (FLHC), Y Braced Box Fuselage and the modified fuselage designed with Y brace replaced by vaulted shell configurations. However all these configurations still inevitably have structural weight penalty compared with optimal tube fuselage layout. This current study intends to focus on finding an optimal configuration with minimum structural weight penalty for a flying-wing concept in a preliminary design stage. A new possible inner wing configuration, in terms of aerodynamic shape and structural layout, was proposed by the author, and it might be referred as ‘Wave-Section Configuration’. The methodologies of how to obtain a structurally efficient curvature of the shape, as well as how to conduct the initial sizing were incorporated. A theoretical analysis of load transmission indicated that the Wave-Section Configuration is feasible, and this was further proved as being practical by FE analysis. Moreover, initial FE analysis and comparison of the Wave-Section Configuration with two other typical configurations, Multi Bubble Fuselage and Conventional Wing, suggested that the Wave-Section Configuration is an optimal design in terms of weight saving. However, due to limitations of the author’s research area, influences on aerodynamic performances have not yet been taken into account
    corecore