Automatic speaker recognition under adverse conditions

Speaker verification is the process of verifying the identity of a person by analysing their speech. There are several important applications for automatic speaker verification (ASV) technology including suspect identification, tracking terrorists and detecting a person’s presence at a remote location in the surveillance domain, as well as person authentication for phone banking and credit card transactions in the private sector. Telephones and telephony networks provide a natural medium for these applications. The aim of this work is to improve the usefulness of ASV technology for practical applications in the presence of adverse conditions. In a telephony environment, background noise, handset mismatch, channel distortions, room acoustics and restrictions on the available testing and training data are common sources of errors for ASV systems. Two research themes were pursued to overcome these adverse conditions: Modelling mismatch and modelling uncertainty. To directly address the performance degradation incurred through mismatched conditions it was proposed to directly model this mismatch. Feature mapping was evaluated for combating handset mismatch and was extended through the use of a blind clustering algorithm to remove the need for accurate handset labels for the training data. Mismatch modelling was then generalised by explicitly modelling the session conditions as a constrained offset of the speaker model means. This session variability modelling approach enabled the modelling of arbitrary sources of mismatch, including handset type, and halved the error rates in many cases. Methods to model the uncertainty in speaker model estimates and verification scores were developed to address the difficulties of limited training and testing data. The Bayes factor was introduced to account for the uncertainty of the speaker model estimates in testing by applying Bayesian theory to the verification criterion, with improved performance in matched conditions. Modelling the uncertainty in the verification score itself met with significant success. Estimating a confidence interval for the "true" verification score enabled an order of magnitude reduction in the average quantity of speech required to make a confident verification decision based on a threshold. The confidence measures developed in this work may also have significant applications for forensic speaker verification tasks.

Probabilistic visual recognition of artificial landmarks for simultaneous localization and mapping

Probabilistic robotics, most often applied to the problem of simultaneous localisation and mapping (SLAM), requires measures of uncertainly to accompany observations of the environment. This paper describes how uncertainly can be characterised for a vision system that locates coloured landmark in a typical laboratory environment. The paper describes a model of the uncertainly in segmentation, the internal camera model and the mounting of the camera on the robot. It =plains the implementation of the system on a laboratory robot, and provides experimental results that show the coherence of the uncertainly model,

The use of phase in complex spectrum subtraction for robust speech recognition

In this paper we propose a new method for utilising phase information by complementing it with traditional magnitude-only spectral subtraction speech enhancement through Complex Spectrum Subtraction (CSS). The proposed approach has the following advantages over traditional magnitude-only spectral subtraction: (a) it introduces complementary information to the enhancement algorithm; (b) it reduces the total number of algorithmic parameters, and; (c) is designed for improving clean speech magnitude spectra and is therefore suitable for both automatic speech recognition (ASR) and speech perception applications. Oracle-based ASR experiments verify this approach, showing an average of 20% relative word accuracy improvements when accurate estimates of the phase spectrum are available. Based on sinusoidal analysis and assuming stationarity between observations (which is shown to be better approximated as the frame rate is increased), this paper also proposes a novel method for acquiring the phase information called Phase Estimation via Delay Projection (PEDEP). Further oracle ASR experiments validate the potential for the proposed PEDEP technique in ideal conditions. Realistic implementation of CSS with PEDEP shows performance comparable to state of the art spectral subtraction techniques in a range of 15-20 dB signal-to-noise ratio environments. These results clearly demonstrate the potential for using phase spectra in spectral subtractive enhancement applications, and at the same time highlight the need for deriving more accurate phase estimates in a wider range of noise conditions.

Focus-score weighted super-resolution for uncooperative iris recognition at a distance and on the move

Uncooperative iris identification systems at a distance and on the move often suffer from poor resolution and poor focus of the captured iris images. The lack of pixel resolution and well-focused images significantly degrades the iris recognition performance. This paper proposes a new approach to incorporate the focus score into a reconstruction-based super-resolution process to generate a high resolution iris image from a low resolution and focus inconsistent video sequence of an eye. A reconstruction-based technique, which can incorporate middle and high frequency components from multiple low resolution frames into one desired super-resolved frame without introducing false high frequency components, is used. A new focus assessment approach is proposed for uncooperative iris at a distance and on the move to improve performance for variations in lighting, size and occlusion. A novel fusion scheme is then proposed to incorporate the proposed focus score into the super-resolution process. The experiments conducted on the The Multiple Biometric Grand Challenge portal database shows that our proposed approach achieves an EER of 2.1%, outperforming the existing state-of-the-art averaging signal-level fusion approach by 19.2% and the robust mean super-resolution approach by 8.7%.

Low-cost hardware speech enhancement for improved speech recognition in automotive environments

Voice recognition is one of the key enablers to reduce driver distraction as in-vehicle systems become more and more complex. With the integration of voice recognition in vehicles, safety and usability are improved as the driver’s eyes and hands are not required to operate system controls. Whilst speaker independent voice recognition is well developed, performance in high noise environments (e.g. vehicles) is still limited. La Trobe University and Queensland University of Technology have developed a low-cost hardware-based speech enhancement system for automotive environments based on spectral subtraction and delay–sum beamforming techniques. The enhancement algorithms have been optimised using authentic Australian English collected under typical driving conditions. Performance tests conducted using speech data collected under variety of vehicle noise conditions demonstrate a word recognition rate improvement in the order of 10% or more under the noisiest conditions. Currently developed to a proof of concept stage there is potential for even greater performance improvement.

A research opportunity : “3D-ROSE” : virtual reality preparation for children having radiotherapy or imaging procedures

Professor Christian Langton is a medical physicist at Queensland University of Technology in Brisbane. He has developed a way of preparing children who are about to have either radiotherapy or MRI imaging procedures and is seeking research partners to develop and test these further. This is a great opportunity for nurses interested in research, and who have access to a children’s hospital, to work with Professor Langton on some truly innovative, multidisciplinary research.

Attractiveness of venture idea amongst expert entrepreneurs : a conjoint analysis

Principal Topic Venture ideas are at the heart of entrepreneurship (Davidsson, 2004). However, we are yet to learn what factors drive entrepreneurs’ perceptions of the attractiveness of venture ideas, and what the relative importance of these factors are for their decision to pursue an idea. The expected financial gain is one factor that will obviously influence the perceived attractiveness of a venture idea (Shepherd & DeTienne, 2005). In addition, the degree of novelty of venture ideas along one or more dimensions such as new products/services, new method of production, enter into new markets/customer and new method of promotion may affect their attractiveness (Schumpeter, 1934). Further, according to the notion of an individual-opportunity nexus venture ideas are closely associated with certain individual characteristics (relatedness). Shane (2000) empirically identified that individual’s prior knowledge is closely associated with the recognition of venture ideas. Sarasvathy’s (2001; 2008) Effectuation theory proposes a high degree of relatedness between venture ideas and the resource position of the individual. This study examines how entrepreneurs weigh considerations of different forms of novelty and relatedness as well as potential financial gain in assessing the attractiveness of venture ideas. Method I use conjoint analysis to determine how expert entrepreneurs develop preferences for venture ideas which involved with different degrees of novelty, relatedness and potential gain. The conjoint analysis estimates respondents’ preferences in terms of utilities (or part-worth) for each level of novelty, relatedness and potential gain of venture ideas. A sample of 32 expert entrepreneurs who were awarded young entrepreneurship awards were selected for the study. Each respondent was interviewed providing with 32 scenarios which explicate different combinations of possible profiles open them into consideration. Results and Implications Results indicate that while the respondents do not prefer mere imitation they receive higher utility for low to medium degree of newness suggesting that high degrees of newness are fraught with greater risk and/or greater resource needs. Respondents pay considerable weight on alignment with the knowledge and skills they already posses in choosing particular venture idea. The initial resource position of entrepreneurs is not equally important. Even though expected potential financial gain gives substantial utility, result indicate that it is not a dominant factor for the attractiveness of venture idea.

Ontology-based image annotation

With regard to the long-standing problem of the semantic gap between low-level image features and high-level human knowledge, the image retrieval community has recently shifted its emphasis from low-level features analysis to high-level image semantics extrac- tion. User studies reveal that users tend to seek information using high-level semantics. Therefore, image semantics extraction is of great importance to content-based image retrieval because it allows the users to freely express what images they want. Semantic content annotation is the basis for semantic content retrieval. The aim of image anno- tation is to automatically obtain keywords that can be used to represent the content of images. The major research challenges in image semantic annotation are: what is the basic unit of semantic representation? how can the semantic unit be linked to high-level image knowledge? how can the contextual information be stored and utilized for image annotation? In this thesis, the Semantic Web technology (i.e. ontology) is introduced to the image semantic annotation problem. Semantic Web, the next generation web, aims at mak- ing the content of whatever type of media not only understandable to humans but also to machines. Due to the large amounts of multimedia data prevalent on the Web, re- searchers and industries are beginning to pay more attention to the Multimedia Semantic Web. The Semantic Web technology provides a new opportunity for multimedia-based applications, but the research in this area is still in its infancy. Whether ontology can be used to improve image annotation and how to best use ontology in semantic repre- sentation and extraction is still a worth-while investigation. This thesis deals with the problem of image semantic annotation using ontology and machine learning techniques in four phases as below. 1) Salient object extraction. A salient object servers as the basic unit in image semantic extraction as it captures the common visual property of the objects. Image segmen- tation is often used as the �rst step for detecting salient objects, but most segmenta- tion algorithms often fail to generate meaningful regions due to over-segmentation and under-segmentation. We develop a new salient object detection algorithm by combining multiple homogeneity criteria in a region merging framework. 2) Ontology construction. Since real-world objects tend to exist in a context within their environment, contextual information has been increasingly used for improving object recognition. In the ontology construction phase, visual-contextual ontologies are built from a large set of fully segmented and annotated images. The ontologies are composed of several types of concepts (i.e. mid-level and high-level concepts), and domain contextual knowledge. The visual-contextual ontologies stand as a user-friendly interface between low-level features and high-level concepts. 3) Image objects annotation. In this phase, each object is labelled with a mid-level concept in ontologies. First, a set of candidate labels are obtained by training Support Vectors Machines with features extracted from salient objects. After that, contextual knowledge contained in ontologies is used to obtain the �nal labels by removing the ambiguity concepts. 4) Scene semantic annotation. The scene semantic extraction phase is to get the scene type by using both mid-level concepts and domain contextual knowledge in ontologies. Domain contextual knowledge is used to create scene con�guration that describes which objects co-exist with which scene type more frequently. The scene con�guration is represented in a probabilistic graph model, and probabilistic inference is employed to calculate the scene type given an annotated image. To evaluate the proposed methods, a series of experiments have been conducted in a large set of fully annotated outdoor scene images. These include a subset of the Corel database, a subset of the LabelMe dataset, the evaluation dataset of localized semantics in images, the spatial context evaluation dataset, and the segmented and annotated IAPR TC-12 benchmark.

Informational city

Many cities worldwide face the prospect of major transformation as the world moves towards a global information order. In this new era, urban economies are being radically altered by dynamic processes of economic and spatial restructuring. The result is the creation of ‘informational cities’ or its new and more popular name, ‘knowledge cities’. For the last two centuries, social production had been primarily understood and shaped by neo-classical economic thought that recognized only three factors of production: land, labor and capital. Knowledge, education, and intellectual capacity were secondary, if not incidental, factors. Human capital was assumed to be either embedded in labor or just one of numerous categories of capital. In the last decades, it has become apparent that knowledge is sufficiently important to deserve recognition as a fourth factor of production. Knowledge and information and the social and technological settings for their production and communication are now seen as keys to development and economic prosperity. The rise of knowledge-based opportunity has, in many cases, been accompanied by a concomitant decline in traditional industrial activity. The replacement of physical commodity production by more abstract forms of production (e.g. information, ideas, and knowledge) has, however paradoxically, reinforced the importance of central places and led to the formation of knowledge cities. Knowledge is produced, marketed and exchanged mainly in cities. Therefore, knowledge cities aim to assist decision-makers in making their cities compatible with the knowledge economy and thus able to compete with other cities. Knowledge cities enable their citizens to foster knowledge creation, knowledge exchange and innovation. They also encourage the continuous creation, sharing, evaluation, renewal and update of knowledge. To compete nationally and internationally, cities need knowledge infrastructures (e.g. universities, research and development institutes); a concentration of well-educated people; technological, mainly electronic, infrastructure; and connections to the global economy (e.g. international companies and finance institutions for trade and investment). Moreover, they must possess the people and things necessary for the production of knowledge and, as importantly, function as breeding grounds for talent and innovation. The economy of a knowledge city creates high value-added products using research, technology, and brainpower. Private and the public sectors value knowledge, spend money on its discovery and dissemination and, ultimately, harness it to create goods and services. Although many cities call themselves knowledge cities, currently, only a few cities around the world (e.g., Barcelona, Delft, Dublin, Montreal, Munich, and Stockholm) have earned that label. Many other cities aspire to the status of knowledge city through urban development programs that target knowledge-based urban development. Examples include Copenhagen, Dubai, Manchester, Melbourne, Monterrey, Singapore, and Shanghai. Knowledge-Based Urban Development To date, the development of most knowledge cities has proceeded organically as a dependent and derivative effect of global market forces. Urban and regional planning has responded slowly, and sometimes not at all, to the challenges and the opportunities of the knowledge city. That is changing, however. Knowledge-based urban development potentially brings both economic prosperity and a sustainable socio-spatial order. Its goal is to produce and circulate abstract work. The globalization of the world in the last decades of the twentieth century was a dialectical process. On one hand, as the tyranny of distance was eroded, economic networks of production and consumption were constituted at a global scale. At the same time, spatial proximity remained as important as ever, if not more so, for knowledge-based urban development. Mediated by information and communication technology, personal contact, and the medium of tacit knowledge, organizational and institutional interactions are still closely associated with spatial proximity. The clustering of knowledge production is essential for fostering innovation and wealth creation. The social benefits of knowledge-based urban development extend beyond aggregate economic growth. On the one hand is the possibility of a particularly resilient form of urban development secured in a network of connections anchored at local, national, and global coordinates. On the other hand, quality of place and life, defined by the level of public service (e.g. health and education) and by the conservation and development of the cultural, aesthetic and ecological values give cities their character and attract or repel the creative class of knowledge workers, is a prerequisite for successful knowledge-based urban development. The goal is a secure economy in a human setting: in short, smart growth or sustainable urban development.