Spectral coding methods for speech compression and speaker identification

This thesis investigates aspects of encoding the speech spectrum at low bit rates, with extensions to the effect of such coding on automatic speaker identification. Vector quantization (VQ) is a technique for jointly quantizing a block of samples at once, in order to reduce the bit rate of a coding system. The major drawback in using VQ is the complexity of the encoder. Recent research has indicated the potential applicability of the VQ method to speech when product code vector quantization (PCVQ) techniques are utilized. The focus of this research is the efficient representation, calculation and utilization of the speech model as stored in the PCVQ codebook. In this thesis, several VQ approaches are evaluated, and the efficacy of two training algorithms is compared experimentally. It is then shown that these productcode vector quantization algorithms may be augmented with lossless compression algorithms, thus yielding an improved overall compression rate. An approach using a statistical model for the vector codebook indices for subsequent lossless compression is introduced. This coupling of lossy compression and lossless compression enables further compression gain. It is demonstrated that this approach is able to reduce the bit rate requirement from the current 24 bits per 20 millisecond frame to below 20, using a standard spectral distortion metric for comparison. Several fast-search VQ methods for use in speech spectrum coding have been evaluated. The usefulness of fast-search algorithms is highly dependent upon the source characteristics and, although previous research has been undertaken for coding of images using VQ codebooks trained with the source samples directly, the product-code structured codebooks for speech spectrum quantization place new constraints on the search methodology. The second major focus of the research is an investigation of the effect of lowrate spectral compression methods on the task of automatic speaker identification. The motivation for this aspect of the research arose from a need to simultaneously preserve the speech quality and intelligibility and to provide for machine-based automatic speaker recognition using the compressed speech. This is important because there are several emerging applications of speaker identification where compressed speech is involved. Examples include mobile communications where the speech has been highly compressed, or where a database of speech material has been assembled and stored in compressed form. Although these two application areas have the same objective - that of maximizing the identification rate - the starting points are quite different. On the one hand, the speech material used for training the identification algorithm may or may not be available in compressed form. On the other hand, the new test material on which identification is to be based may only be available in compressed form. Using the spectral parameters which have been stored in compressed form, two main classes of speaker identification algorithm are examined. Some studies have been conducted in the past on bandwidth-limited speaker identification, but the use of short-term spectral compression deserves separate investigation. Combining the major aspects of the research, some important design guidelines for the construction of an identification model when based on the use of compressed speech are put forward.

Visual arts, technology and education : how can teaching and learning in high school visual arts classrooms be enriched by the use of computer technology?

Whilst a variety of studies has appeared over the last decade addressing the gap between the potential promised by computers and the reality experienced in the classroom by teachers and students, few have specifically addressed the situation as it pertains to the visual arts classroom. The aim of this study was to explore the reality of the classroom use of computers for three visual arts highschool teachers and determine how computer technology might enrich visual arts teaching and learning. An action research approach was employed to enable the researcher to understand the situation from the teachers' points of view while contributing to their professional practice. The wider social context surrounding this study is characterised by an increase in visual communications brought about by rapid advances in computer technology. The powerful combination of visual imagery and computer technology is illustrated by continuing developments in the print, film and television industries. In particular, the recent growth of interactive multimedia epitomises this combination and is significant to this study as it represents a new form of publishing of great interest to educators and artists alike. In this social context, visual arts education has a significant role to play. By cultivating a critical awareness of the implications of technology use and promoting a creative approach to the application of computer technology within the visual arts, visual arts education is in a position to provide an essential service to students who will leave high school to participate in a visual information age as both consumers and producers.

A grounded theory study of technology appropriation in anaesthesia

The human-technology nexus is a strong focus of Information Systems (IS) research; however, very few studies have explored this phenomenon in anaesthesia. Anaesthesia has a long history of adoption of technological artifacts, ranging from early apparatus to present-day information systems such as electronic monitoring and pulse oximetry. This prevalence of technology in modern anaesthesia and the rich human-technology relationship provides a fertile empirical setting for IS research. This study employed a grounded theory approach that began with a broad initial guiding question and, through simultaneous data collection and analysis, uncovered a core category of technology appropriation. This emergent basic social process captures a central activity of anaesthestists and is supported by three major concepts: knowledge-directed medicine, complementary artifacts and culture of anaesthesia. The outcomes of this study are: (1) a substantive theory that integrates the aforementioned concepts and pertains to the research setting of anaesthesia and (2) a formal theory, which further develops the core category of appropriation from anaesthesia-specific to a broader, more general perspective. These outcomes fulfill the objective of a grounded theory study, being the formation of theory that describes and explains observed patterns in the empirical field. In generalizing the notion of appropriation, the formal theory is developed using the theories of Karl Marx. This Marxian model of technology appropriation is a three-tiered theoretical lens that examines appropriation behaviours at a highly abstract level, connecting the stages of natural, species and social being to the transition of a technology-as-artifact to a technology-in-use via the processes of perception, orientation and realization. The contributions of this research are two-fold: (1) the substantive model contributes to practice by providing a model that describes and explains the human-technology nexus in anaesthesia, and thereby offers potential predictive capabilities for designers and administrators to optimize future appropriations of new anaesthetic technological artifacts; and (2) the formal model contributes to research by drawing attention to the philosophical foundations of appropriation in the work of Marx, and subsequently expanding the current understanding of contemporary IS theories of adoption and appropriation.

Economic growth, inequality and technology adoption in transitional economies

The stylized facts that motivate this thesis include the diversity in growth patterns that are observed across countries during the process of economic development, and the divergence over time in income distributions both within and across countries. This thesis constructs a dynamic general equilibrium model in which technology adoption is costly and agents are heterogeneous in their initial holdings of resources. Given the households‟ resource level, this study examines how adoption costs influence the evolution of household income over time and the timing of transition to more productive technologies. The analytical results of the model constructed here characterize three growth outcomes associated with the technology adoption process depending on productivity differences between the technologies. These are appropriately labeled as „poverty trap‟, „dual economy‟ and „balanced growth‟. The model is then capable of explaining the observed diversity in growth patterns across countries, as well as divergence of incomes over time. Numerical simulations of the model furthermore illustrate features of this transition. They suggest that that differences in adoption costs account for the timing of households‟ decision to switch technology which leads to a disparity in incomes across households in the technology adoption process. Since this determines the timing of complete adoption of the technology within a country, the implications for cross-country income differences are obvious. Moreover, the timing of technology adoption appears to be impacts on patterns of growth of households, which are different across various income groups. The findings also show that, in the presence of costs associated with the adoption of more productive technologies, inequalities of income and wealth may increase over time tending to delay the convergence in income levels. Initial levels of inequalities in the resources also have an impact on the date of complete adoption of more productive technologies. The issue of increasing income inequality in the process of technology adoption opens up another direction for research. Specifically increasing inequality implies that distributive conflicts may emerge during the transitional process with political- economy consequences. The model is therefore extended to include such issues. Without any political considerations, taxes would leads to a reduction in inequality and convergence of incomes across agents. However this process is delayed if politico-economic influences are taken into account. Moreover, the political outcome is sub optimal. This is essentially due to the fact that there is a resistance associated with the complete adoption of the advanced technology.

Expert elicitation and its interface with technology : a review with a view to designing Elicitator

Expert knowledge is valuable in many modelling endeavours, particularly where data is not extensive or sufficiently robust. In Bayesian statistics, expert opinion may be formulated as informative priors, to provide an honest reflection of the current state of knowledge, before updating this with new information. Technology is increasingly being exploited to help support the process of eliciting such information. This paper reviews the benefits that have been gained from utilizing technology in this way. These benefits can be structured within a six-step elicitation design framework proposed recently (Low Choy et al., 2009). We assume that the purpose of elicitation is to formulate a Bayesian statistical prior, either to provide a standalone expert-defined model, or for updating new data within a Bayesian analysis. We also assume that the model has been pre-specified before selecting the software. In this case, technology has the most to offer to: targeting what experts know (E2), eliciting and encoding expert opinions (E4), whilst enhancing accuracy (E5), and providing an effective and efficient protocol (E6). Benefits include: -providing an environment with familiar nuances (to make the expert comfortable) where experts can explore their knowledge from various perspectives (E2); -automating tedious or repetitive tasks, thereby minimizing calculation errors, as well as encouraging interaction between elicitors and experts (E5); -cognitive gains by educating users, enabling instant feedback (E2, E4-E5), and providing alternative methods of communicating assessments and feedback information, since experts think and learn differently; and -ensuring a repeatable and transparent protocol is used (E6).