Speaker diarization : "who spoke when"

Speaker diarization is the process of annotating an input audio with information that attributes temporal regions of the audio signal to their respective sources, which may include both speech and non-speech events. For speech regions, the diarization system also specifies the locations of speaker boundaries and assign relative speaker labels to each homogeneous segment of speech. In short, speaker diarization systems effectively answer the question of ‘who spoke when’. There are several important applications for speaker diarization technology, such as facilitating speaker indexing systems to allow users to directly access the relevant segments of interest within a given audio, and assisting with other downstream processes such as summarizing and parsing. When combined with automatic speech recognition (ASR) systems, the metadata extracted from a speaker diarization system can provide complementary information for ASR transcripts including the location of speaker turns and relative speaker segment labels, making the transcripts more readable. Speaker diarization output can also be used to localize the instances of specific speakers to pool data for model adaptation, which in turn boosts transcription accuracies. Speaker diarization therefore plays an important role as a preliminary step in automatic transcription of audio data. The aim of this work is to improve the usefulness and practicality of speaker diarization technology, through the reduction of diarization error rates. In particular, this research is focused on the segmentation and clustering stages within a diarization system. Although particular emphasis is placed on the broadcast news audio domain and systems developed throughout this work are also trained and tested on broadcast news data, the techniques proposed in this dissertation are also applicable to other domains including telephone conversations and meetings audio. Three main research themes were pursued: heuristic rules for speaker segmentation, modelling uncertainty in speaker model estimates, and modelling uncertainty in eigenvoice speaker modelling. The use of heuristic approaches for the speaker segmentation task was first investigated, with emphasis placed on minimizing missed boundary detections. A set of heuristic rules was proposed, to govern the detection and heuristic selection of candidate speaker segment boundaries. A second pass, using the same heuristic algorithm with a smaller window, was also proposed with the aim of improving detection of boundaries around short speaker segments. Compared to single threshold based methods, the proposed heuristic approach was shown to provide improved segmentation performance, leading to a reduction in the overall diarization error rate. Methods to model the uncertainty in speaker model estimates were developed, to address the difficulties associated with making segmentation and clustering decisions with limited data in the speaker segments. The Bayes factor, derived specifically for multivariate Gaussian speaker modelling, was introduced to account for the uncertainty of the speaker model estimates. The use of the Bayes factor also enabled the incorporation of prior information regarding the audio to aid segmentation and clustering decisions. The idea of modelling uncertainty in speaker model estimates was also extended to the eigenvoice speaker modelling framework for the speaker clustering task. Building on the application of Bayesian approaches to the speaker diarization problem, the proposed approach takes into account the uncertainty associated with the explicit estimation of the speaker factors. The proposed decision criteria, based on Bayesian theory, was shown to generally outperform their non- Bayesian counterparts.

The construction contract bidder homogeneity assumption : an empirical test

This paper describes an empirical study to test the proposition that all construction contract bidders are homogeneous ie. they can be treated as behaving collectively in an identical (statistical) manner. Examination of previous analyses of bidding data reveals a flaw in the method of standardising bids across different size contracts and a new procedure is proposed which involves the estimation of a contract datum. Three independent sets of bidding data were then subjected to this procedure and estimates of the necessary distributional parameters obtained. These were then tested against the bidder homogeneity assumption resulting in the conclusion that the assumption may be appropriate for a three parameter log-normal shape, but not for scale and location.

Inside the boardroom : exploring board member interactions

This study aims to open-up the black box of the boardroom by directly observing directors’ interactions during meetings to better understand board processes. Design/methodology/approach: We analyse videotaped observations of board meetings at two Australian companies to develop insights into what directors do in meetings and how they participate in decision-making processes. The direct observations are triangulated with semi-structured interviews, mini-surveys and document reviews. Findings: Our analyses lead to two key findings: (i) while board meetings appear similar at a surface-level, boardroom interactions vary significantly at a deeper level (i.e. board members participate differently during different stages of discussions) and (ii) factors at multiple levels of analysis explain differences in interaction patterns, revealing the complex and nested nature of boardroom discussions. Research implications: By documenting significant intra- and inter-board meeting differences our study (i) challenges the widespread notion of board meetings as rather homogeneous and monolithic, (ii) points towards agenda items as a new unit of analysis (iii) highlights the need for more multi-level analyses in a board setting. Practical implications: While policy makers have been largely occupied with the “right” board composition, our findings suggest that decision outcomes or roles’ execution could be potentially affected by interactions at a board level. Differences in board meeting styles might explain prior ambiguous board structure-performance results, enhancing the need for greater normative consideration of how boards do their work. Originality/value: Our study complements existing research on boardroom dynamics and provides a systematic account of director interactions during board meetings.

Online calibration of stereo rigs for long-term autonomy

Stereo-based visual odometry algorithms are heavily dependent on an accurate calibration of the rigidly fixed stereo pair. Even small shifts in the rigid transform between the cameras can impact on feature matching and 3D scene triangulation, adversely affecting pose estimates and applications dependent on long-term autonomy. In many field-based scenarios where vibration, knocks and pressure change affect a robotic vehicle, maintaining an accurate stereo calibration cannot be guaranteed over long periods. This paper presents a novel method of recalibrating overlapping stereo camera rigs from online visual data while simultaneously providing an up-to-date and up-to-scale pose estimate. The proposed technique implements a novel form of partitioned bundle adjustment that explicitly includes the homogeneous transform between a stereo camera pair to generate an optimal calibration. Pose estimates are computed in parallel to the calibration, providing online recalibration which seamlessly integrates into a stereo visual odometry framework. We present results demonstrating accurate performance of the algorithm on both simulated scenarios and real data gathered from a wide-baseline stereo pair on a ground vehicle traversing urban roads.

The use of electronic portal imaging to verify patient position during intensity-modulated radiotherapy delivered by the dynamic MLC technique

Purpose: The precise shape of the three-dimensional dose distributions created by intensity-modulated radiotherapy means that the verification of patient position and setup is crucial to the outcome of the treatment. In this paper, we investigate and compare the use of two different image calibration procedures that allow extraction of patient anatomy from measured electronic portal images of intensity-modulated treatment beams. Methods and Materials: Electronic portal images of the intensity-modulated treatment beam delivered using the dynamic multileaf collimator technique were acquired. The images were formed by measuring a series of frames or segments throughout the delivery of the beams. The frames were then summed to produce an integrated portal image of the delivered beam. Two different methods for calibrating the integrated image were investigated with the aim of removing the intensity modulations of the beam. The first involved a simple point-by-point division of the integrated image by a single calibration image of the intensity-modulated beam delivered to a homogeneous polymethyl methacrylate (PMMA) phantom. The second calibration method is known as the quadratic calibration method and required a series of calibration images of the intensity-modulated beam delivered to different thicknesses of homogeneous PMMA blocks. Measurements were made using two different detector systems: a Varian amorphous silicon flat-panel imager and a Theraview camera-based system. The methods were tested first using a contrast phantom before images were acquired of intensity-modulated radiotherapy treatment delivered to the prostate and pelvic nodes of cancer patients at the Royal Marsden Hospital. Results: The results indicate that the calibration methods can be used to remove the intensity modulations of the beam, making it possible to see the outlines of bony anatomy that could be used for patient position verification. This was shown for both posterior and lateral delivered fields. Conclusions: Very little difference between the two calibration methods was observed, so the simpler division method, requiring only the single extra calibration measurement and much simpler computation, was the favored method. This new method could provide a complementary tool to existing position verification methods, and it has the advantage that it is completely passive, requiring no further dose to the patient and using only the treatment fields.

How long does it take for aquifer recharge or aquifer discharge processes to reach steady state?

Groundwater flow models are usually characterized as being either transient flow models or steady state flow models. Given that steady state groundwater flow conditions arise as a long time asymptotic limit of a particular transient response, it is natural for us to seek a finite estimate of the amount of time required for a particular transient flow problem to effectively reach steady state. Here, we introduce the concept of mean action time (MAT) to address a fundamental question: How long does it take for a groundwater recharge process or discharge processes to effectively reach steady state? This concept relies on identifying a cumulative distribution function, $F(t;x)$, which varies from $F(0;x)=0$ to $F(t;x) \to \infty$ as $t\to \infty$, thereby providing us with a measurement of the progress of the system towards steady state. The MAT corresponds to the mean of the associated probability density function $f(t;x) = \dfrac{dF}{dt}$, and we demonstrate that this framework provides useful analytical insight by explicitly showing how the MAT depends on the parameters in the model and the geometry of the problem. Additional theoretical results relating to the variance of $f(t;x)$, known as the variance of action time (VAT), are also presented. To test our theoretical predictions we include measurements from a laboratory–scale experiment describing flow through a homogeneous porous medium. The laboratory data confirms that the theoretical MAT predictions are in good agreement with measurements from the physical model.

Real time information provision benefit measured by macroscopic fundamental diagram

The existence of Macroscopic Fundamental Diagram (MFD), which relates space-mean density and flow, has been shown in urban networks under homogeneous traffic conditions. Since MFD represents the area-wide network traffic performances, studies on perimeter control strategies and an area traffic state estimation utilizing the MFD concept has been reported. One of the key requirements for well-defined MFD is the homogeneity of the area-wide traffic condition with links of similar properties, which is not universally expected in real world. For the practical application of the MFD concept, several researchers have identified the influencing factors for network homogeneity. However, they did not explicitly take the impact of drivers’ behaviour and information provision into account, which has a significant impact on simulation outputs. This research aims to demonstrate the effect of dynamic information provision on network performance by employing the MFD as a measurement. A microscopic simulation, AIMSUN, is chosen as an experiment platform. By changing the ratio of en-route informed drivers and pre-trip informed drivers different scenarios are simulated in order to investigate how drivers’ adaptation to the traffic congestion influences the network performance with respect to the MFD shape as well as other indicators, such as total travel time. This study confirmed the impact of information provision on the MFD shape, and addressed the usefulness of the MFD for measuring the dynamic information provision benefit.

Fabrication and mechanical and thermal behaviour of graphene oxide/epoxy nanocomposites

Bulk amount of graphite oxide was prepared by oxidation of graphite using the modified Hummers method and its ultrasonication in organic solvents yielded graphene oxide (GO). X-ray diffraction (XRD) pattern, X-ray photoelectron (XPS), Raman and Fourier transform infrared (FTIR) spectroscopy indicated the successful preparation of GO. XPS survey spectrum of GO revealed the presence of 66.6 at% C and 30.4 at% O. Scanning electron microscopy (SEM) and Transmission electron microscopy (TEM) images of the graphene oxide showed that they consist of a large amount of graphene oxide platelets with a curled morphology containing of a thin wrinkled sheet like structure. AFM image of the exfoliated GO signified that the average thickness of GO sheets is ~1.0 nm which is very similar to GO monolayer. GO/epoxy nanocomposites were prepared by typical solution mixing technique and influence of GO on mechanical and thermal properties of nanocomposites were investigated. As for the mechanical behaviour of GO/epoxy nanocomposites, 0.5 wt% GO in the nanocomposite achieved the maximum increase in the elastic modulus (~35%) and tensile strength (~7%). The TEM analysis provided clear image of microstructure with homogeneous dispersion of GO in the polymer matrix. The improved strength properties of GO/epoxy nanocomposites can be attributed to inherent strength of GO, the good dispersion and the strong interfacial interactions between the GO sheets and the polymer matrix. However, incorporation of GO showed significant negative effect on composite glass transition temperature (Tg). This may arise due to the interference of GO on curing reaction of epoxy.

Contracting for infrastructure projects as credence goods

Large infrastructure projects are a major responsibility of urban and regional governments, who usually lack expertise to fully specify the demanded projects. Contractors, typically experts on such projects due to experience with similar projects elsewhere, advise of the needed design in their bids. Producing the right design is nevertheless costly. We model such infrastructure projects taking into account their credence goods feature and the costly design effort they require and examine the performance of commonly used contracting methods. We show that when building costs are homogeneous and public information, multi-stage competitive bidding involving shortlisting of two contractors and contingent compensation of both contractors on design efforts outperforms sequential search and the traditional Design-and-Build approach. If building costs are private information of the contractors and are revealed to them after design cost is sunk, sequential search may be superior to the other two methods.

Acetylcholine receptor antibody in the diagnosis of myasthenia gravis

The acetylcholine receptor (AchR) antibody assay has a key role in the diagnosis of myasthenia gravis. In this article, the role of AchR antibody assay in the diagnosis of ocular and generalized myasthenia gravis is reviewed, and compared to standard means of diagnosing the disease by clinical and electrophysiological methods.

Making the train : an examination of five perspectives of rail access and how they resonate with existing and potential passengers

Rail operators recognize a need to increase ridership in order to improve the economic viability of rail service, and to magnify the role that rail travel plays in making cities feel liveable. This study extends previous research that used cluster analysis with a small sample of rail passengers to identify five salient perspectives of rail access (Zuniga et al, 2013). In this project stage, we used correlation techniques to determine how those perspectives would resonate with two larger study populations, including a relatively homogeneous sample of university students in Brisbane, Australia and a diverse sample of rail passengers in Melbourne, Australia. Findings from Zuniga et al. (2013) described a complex typology of current passengers that was based on respondents’ subjective attitudes and perceptions rather than socio-demographic or travel behaviour characteristics commonly used for segmentation analysis. The typology included five qualitative perspectives of rail travel. Based on the transport accessibility literature, we expected to find that perspectives from that study emphasizing physical access to rail stations would be shared by current and potential rail passengers who live further from rail stations. Other perspectives might be shared among respondents who live nearby, since the relevance of distance would be diminished. The population living nearby would thus represent an important target group for increasing ridership, since making rail travel accessible to them does not require expansion of costly infrastructure such as new lines or stations. By measuring the prevalence of each perspective in a larger respondent pool, results from this study provide insight into the typical socio-demographic and travel behaviour characteristics that correspond to each perspective of intra-urban rail travel. In several instances, our quantitative findings reinforced Zuniga et al.’s (2013) qualitative descriptions of passenger types, further validating the original research. This work may directly inform rail operators’ approach to increasing ridership through marketing and improvements to service quality and station experience. Operators in other parts of Australia and internationally may also choose to replicate the study locally, to fine-tune understanding of diverse customer bases. Developing regional and international collaboration would provide additional opportunities to evaluate and benchmark service and station amenities as they address the various access dimensions.

A nanometre-scale non-periodic structural variation in high temperature superconducting ceramics and the implications for properties

Non-periodic structural variation has been found in the high Tc cuprates, YBa2Cu3O7-x and Hg0.67Pb0.33Ba2Ca2Cu 3O8+δ, by image analysis of high resolution transmission electron microscope (HRTEM) images. We use two methods for analysis of the HRTEM images. The first method is a means for measuring the bending of lattice fringes at twin planes. The second method is a low-pass filter technique which enhances information contained by diffuse-scattered electrons and reveals what appears to be an interference effect between domains of differing lattice parameter in the top and bottom of the thin foil. We believe that these methods of image analysis could be usefully applied to the many thousands of HRTEM images that have been collected by other workers in the high temperature superconductor field. This work provides direct structural evidence for phase separation in high Tc cuprates, and gives support to recent stripes models that have been proposed to explain various angle resolved photoelectron spectroscopy and nuclear magnetic resonance data. We believe that the structural variation is a response to an opening of an electronic solubility gap where holes are not uniformly distributed in the material but are confined to metallic stripes. Optimum doping may occur as a consequence of the diffuse boundaries between stripes which arise from spinodal decomposition. Theoretical ideas about the high Tc cuprates which treat the cuprates as homogeneous may need to be modified in order to take account of this type of structural variation.

Phase composition of the rapidly quenched melt of YBa2Cu3O7-y+20 mol% Y2BaCuO5

Samples of YBa2Cu3O7-y+20 mol% Y2BaCuO5, with thicknesses ranging between 50-250 μm, have been melt processed and rapidly quenched from temperatures between 985 and 1100°C by immersing them in liquid nitrogen. The phase composition and microstructures of these samples have been characterised using a combination of X-ray diffractometry, optical microscopy and scanning electron microscopy with energy dispersive X-ray spectroscopy. The quenched melt of samples quenched from temperatures greater than 985°C appears relatively homogeneous but consists of Ba2Cu3Ox (BC1.5) and BaCu2O2 (BC2) regions. At about 985°C, BaCuO2 (BC1) crystallises from the melt and most of the BC1.5 decomposes into BC1 and CuO or into BC1 and BC2. The crystallisation of BC1 induces segregation of elements in the melt and this is very significant for the melt texturing of YBCO.

Automatic segmentation of choroidal thickness in optical coherence tomography

The assessment of choroidal thickness from optical coherence tomography (OCT) images of the human choroid is an important clinical and research task, since it provides valuable information regarding the eye’s normal anatomy and physiology, and changes associated with various eye diseases and the development of refractive error. Due to the time consuming and subjective nature of manual image analysis, there is a need for the development of reliable objective automated methods of image segmentation to derive choroidal thickness measures. However, the detection of the two boundaries which delineate the choroid is a complicated and challenging task, in particular the detection of the outer choroidal boundary, due to a number of issues including: (i) the vascular ocular tissue is non-uniform and rich in non-homogeneous features, and (ii) the boundary can have a low contrast. In this paper, an automatic segmentation technique based on graph-search theory is presented to segment the inner choroidal boundary (ICB) and the outer choroidal boundary (OCB) to obtain the choroid thickness profile from OCT images. Before the segmentation, the B-scan is pre-processed to enhance the two boundaries of interest and to minimize the artifacts produced by surrounding features. The algorithm to detect the ICB is based on a simple edge filter and a directional weighted map penalty, while the algorithm to detect the OCB is based on OCT image enhancement and a dual brightness probability gradient. The method was tested on a large data set of images from a pediatric (1083 B-scans) and an adult (90 B-scans) population, which were previously manually segmented by an experienced observer. The results demonstrate the proposed method provides robust detection of the boundaries of interest and is a useful tool to extract clinical data.

Effect of the mediator in feedback mode-based SECM interrogation of indium tin-oxide and boron-doped diamond electrodes

Indium tin-oxide (ITO) and polycrystalline boron-doped diamond (BDD) have been examined in detail using the scanning electrochemical microscopy technique in feedback mode. For the interrogation of electrodes made from these materials, the choice of mediator has been varied. Using Ru(CN) 4− 6 (aq), ferrocene methanol (FcMeOH), Fe(CN) 3− 6 (aq) and Ru(NH 3) 3+ 6 (aq), approach curve experiments have been performed, and for purposes of comparison, calculations of the apparent heterogeneous electron transfer rates (k app) have been made using these data. In general, it would appear that values of k app are affected mainly by the position of the mediator reversible potential relative to the relevant semiconductor band edge (associated with majority carriers). For both the ITO (n type) and BDD (p type) electrodes, charge transfer is impeded and values are very low when using FcMeOH and Fe(CN) 3− 6 (aq) as mediators, and the use of Ru(NH 3) 3+ 6(aq) results in the largest value of k app. With ITO, the surface is chemically homogeneous and no variation is observed for any given mediator. Data is also presented where the potential of the ITO electrode is fixed using a ratio of the mediators Fe(CN) 3− 6(aq) and Fe(CN) 4− 6(aq). In stark contrast, the BDD electrode is quite the opposite and a range of k app values are observed for all mediators depending on the position on the surface. Both electrode surfaces are very flat and very smooth, and hence, for BDD, variations in feedback current imply a variation in the electrochemical activity. A comparison of the feedback current where the substrate is biased and unbiased shows a surprising degree of proportionality.