Cross likelihood ratio based speaker clustering using eigenvoice models

This paper proposes the use of eigenvoice modeling techniques with the Cross Likelihood Ratio (CLR) as a criterion for speaker clustering within a speaker diarization system. The CLR has previously been shown to be a robust decision criterion for speaker clustering using Gaussian Mixture Models. Recently, eigenvoice modeling techniques have become increasingly popular, due to its ability to adequately represent a speaker based on sparse training data, as well as an improved capture of differences in speaker characteristics. This paper hence proposes that it would be beneficial to capitalize on the advantages of eigenvoice modeling in a CLR framework. Results obtained on the 2002 Rich Transcription (RT-02) Evaluation dataset show an improved clustering performance, resulting in a 35.1% relative improvement in the overall Diarization Error Rate (DER) compared to the baseline system.

Diversity interventions for a socially sustainable construction industry

Major construction sites in Australia have an above average presence of ethnic minorities. These groups and the interfaces between them require effective management in order to meet the social imperatives of sustainable design and construction. A survey of 1155 workers and 204 managers on Sydney construction sites respectively, found a significant level of normalisation of negative forms of cross cultural interaction. Yet it was also found that anti-racism programs are not currently a management priority and that they generally lack sophisticated community relations aspects. This paper presents the results of a desk-top study of leading global companies within and outside the construction sector which have won international awards and recognition for their cultural diversity strategies. A key insight is that the companies profiled see diversity as a key resource and as an opportunity rather than a risk which is best harnessed through long-term and on-going commitment of senior management. These leading companies also recognise that cultural diversity strategies operate at three levels - in terms of its relationship with its own workforce; its relationship with its clients and; its relationships with the communities in which it operates - and if properly managed it can be a source of competitive advantage.

i-vector based speaker recognition on short utterances

Robust speaker verification on short utterances remains a key consideration when deploying automatic speaker recognition, as many real world applications often have access to only limited duration speech data. This paper explores how the recent technologies focused around total variability modeling behave when training and testing utterance lengths are reduced. Results are presented which provide a comparison of Joint Factor Analysis (JFA) and i-vector based systems including various compensation techniques; Within-Class Covariance Normalization (WCCN), LDA, Scatter Difference Nuisance Attribute Projection (SDNAP) and Gaussian Probabilistic Linear Discriminant Analysis (GPLDA). Speaker verification performance for utterances with as little as 2 sec of data taken from the NIST Speaker Recognition Evaluations are presented to provide a clearer picture of the current performance characteristics of these techniques in short utterance conditions.

‘Show Me Your Wiki and I’ll Show you Mine’: Using Online Interactive Media to Improve Academic Writing and Research in a Public Health Under-Graduate Cohort

The number of Internet users in Australia has been steadily increasing, with over 10.9 million people currently subscribed to an internet provider (ABS, 2011). Over the past year, the most avid users of the Internet were 15 – 24 year olds, with approximately 95% accessing the internet on a regular basis (ABS, Social Trends, 2011). While the internet has been described as fundamental to higher education students, social and leisure internet tools are also increasingly being used by these students to generate and maintain their social and professional networks and interactions (Duffy & Bruns 2006). Rapid technological advancements have enabled greater and faster access to information for learning and education (Hemmi et al, 2009; Glassman and Kang, 2011). As such, we sought to integrate interactive, online social media into the assessment profile of a Public Health undergraduate cohort at the Queensland University of Technology (QUT). The aim of this exercise was to engage students to both develop and showcase their research on a range of complex, contemporary health issues within the online forum of Wikispaces (http://www.wikispaces.com/) for review and critique by their peers. We applied Bandura’s Social Learning Theory (SLT) to analyse the interactive processes from which students developed deeper and more sustained learning, and via which their overall academic writing standards were raised. This paper outlines the assessment task, and the students’ feedback on their learning outcomes in relation to the Attentional, Retentional, Motor Reproduction, and Motivational Processes outlined by Bandura in SLT. We conceptualise the findings in a theoretical model, and discuss the implications for this approach within the broader tertiary environment.

Management strategies to harness cultural diversity on Australian constructions sites : a social identity perspective

Construction sites around the world employ large numbers of people from diverse cultural backgrounds. The effective management of this cultural diversity has important implications for the productivity, safety, health and welfare of construction workers and for the performance and reputation of firms which employ them. The findings of a three year, multi-staged study of cultural diversity management practices on construction sites are critiqued using social identity theory. This reveals that so called "best-practice" diversity management strategies may have an opposite effect to that intended. It is concluded that the management of diversity on construction projects would benefit from being informed by social identity research.

Anchored deformable face ensemble alignment

At present, many approaches have been proposed for deformable face alignment with varying degrees of success. However, the common drawback to nearly all these approaches is the inaccurate landmark registrations. The registration errors which occur are predominantly heterogeneous (i.e. low error for some frames in a sequence and higher error for others). In this paper we propose an approach for simultaneously aligning an ensemble of deformable face images stemming from the same subject given noisy heterogeneous landmark estimates. We propose that these initial noisy landmark estimates can be used as an “anchor” in conjunction with known state-of-the-art objectives for unsupervised image ensemble alignment. Impressive alignment performance is obtained using well known deformable face fitting algorithms as “anchors.

V1-inspired features induce a weighted margin in SVMs

Image representations derived from simplified models of the primary visual cortex (V1), such as HOG and SIFT, elicit good performance in a myriad of visual classification tasks including object recognition/detection, pedestrian detection and facial expression classification. A central question in the vision, learning and neuroscience communities regards why these architectures perform so well. In this paper, we offer a unique perspective to this question by subsuming the role of V1-inspired features directly within a linear support vector machine (SVM). We demonstrate that a specific class of such features in conjunction with a linear SVM can be reinterpreted as inducing a weighted margin on the Kronecker basis expansion of an image. This new viewpoint on the role of V1-inspired features allows us to answer fundamental questions on the uniqueness and redundancies of these features, and offer substantial improvements in terms of computational and storage efficiency.

On the statistical determination of optimal camera configurations in large scale surveillance networks

The selection of optimal camera configurations (camera locations, orientations etc.) for multi-camera networks remains an unsolved problem. Previous approaches largely focus on proposing various objective functions to achieve different tasks. Most of them, however, do not generalize well to large scale networks. To tackle this, we introduce a statistical formulation of the optimal selection of camera configurations as well as propose a Trans-Dimensional Simulated Annealing (TDSA) algorithm to effectively solve the problem. We compare our approach with a state-of-the-art method based on Binary Integer Programming (BIP) and show that our approach offers similar performance on small scale problems. However, we also demonstrate the capability of our approach in dealing with large scale problems and show that our approach produces better results than 2 alternative heuristics designed to deal with the scalability issue of BIP.

Championing behaviour of a project manager : determinants and outcomes

The role of individuals in the innovation process is highlighted as system integrator and/or champion in literature, however, little is known about championing role of a project manager. Our contention is that the role of the project manager (PM) is essentially of a champion to enable innovation and achieve desired project performance in construction project environment. Hypothesizing that championing behaviour is determined by a number of individual and situational factors, which in turn effects on level of innovation and project performance, we used correlation and regression analysis to test the hypotheses. A survey was carried out with project managers and project team members in 32 building and civil engineering projects in Singapore to test the hypothesized relationships. The results corroborate the importance of championing behaviour to fostering innovation and achieve better project performance.

Efficient articulated trajectory reconstruction using dynamic programming and filters

This paper considers the problem of reconstructing the motion of a 3D articulated tree from 2D point correspondences subject to some temporal prior. Hitherto, smooth motion has been encouraged using a trajectory basis, yielding a hard combinatorial problem with time complexity growing exponentially in the number of frames. Branch and bound strategies have previously attempted to curb this complexity whilst maintaining global optimality. However, they provide no guarantee of being more efficient than exhaustive search. Inspired by recent work which reconstructs general trajectories using compact high-pass filters, we develop a dynamic programming approach which scales linearly in the number of frames, leveraging the intrinsically local nature of filter interactions. Extension to affine projection enables reconstruction without estimating cameras.

Tragedy and assimilation : reciprocity between surface and subject

Here we unveil a tragic triptych of three Australian women painfully painted onto the walls of interior surfaces. The woman at the centre of the triptych is Florence Broadhurst whose tragic death still remains a mystery. To the right is Australian skin illustrator Emma Hack who recreates Broadhurst’s wallpapers, mimicking their colourful patterns onto live models. Hack perfectly assimilates the models’ body into the wallpaper, camouflaging bodies except for small hints at something more in the foreground. In the process of Hack’s images, the models become statues, standing painfully still holding their breath for minutes at a time. The third woman, to the left of the triptych, is the fictional character Candy from the 2006 Australian film Candy. Candy’s traumatic struggle with addiction ends with her conveying her pain in a poem she writes on the walls of her home; culminating her tragic story into a disturbed domestic wall surface. This research tries to understand this relationship with the surface through tragedy as a reciprocal agreement between surface and subject and not a permanent transference between one state and another. What the surface provides in times of personal struggle and turmoil is a method for us to come to terms with out material existence.

Nursing research : the Australian College of Nursing position statement

Background Nursing perspectives play an important role in addressing the health priorities of today’s society. The Australian College of Nursing (ACN) acknowledges the significant contribution that nursing research has made since the first nurse researcher, Florence Nightingale, documented the factors that affected the morbidity and mortality of soldiers wounded in the Crimean war in the 1800s. The nursing profession continues to celebrate the significant contribution nursing research made to improving nursing practice and health outcomes. These significant contributions over recent years include, but are not limited to: 1. Health services research that has demonstrated the importance of nursing services and how such services are designed/organised to ensure safety and quality of care (Duffield, et al., 2011; Fernandez, et al., 2012; Middleton, et al., 2011); 2. Clinical research that has demonstrated the value of specific nursing interventions to improved health outcomes, including enhanced survival, reduced morbidity, and improved quality of life and consumer engagement (Cancer Australia and Cancer Voices Australia, 2011; Kitson, et al., 2013; Middleton, et al., 2012; Rickard, et al., 2012; Zeitz, et al., 2011); 3. Basic science research that has advanced discoveries in terms of understanding the biological mechanisms underpinning nursing interventions (Illi, et al., 2012; Kim, et al., 2012; Miaskowski, et al., 2010; Simonova, et al., 2012); 4. Epidemiological research that has advanced understanding about how individuals and populations respond to health problems (Carrington, et al., 2012); 5. Qualitative research that has advanced understanding about experiences of and responses to health and illness and the processes of care that are important to optimal outcomes (Schulman-Green, et al., 2012; Scott, et al., 2011).

The inaccuracy of national character stereotypes

Consensual stereotypes of some groups are relatively accurate, whereas others are not. Previous work suggesting that national character stereotypes are inaccurate has been criticized on several grounds. In this article we (a) provide arguments for the validity of assessed national mean trait levels as criteria for evaluating stereotype accuracy and (b) report new data on national character in 26 cultures from descriptions (N= 3323) of the typical male or female adolescent, adult, or old person in each. The average ratings were internally consistent and converged with independent stereotypes of the typical culture member, but were weakly related to objective assessments of personality. We argue that this conclusion is consistent with the broader literature on the inaccuracy of national character stereotypes

Stereotypes of age differences in personality traits : universal and accurate?

Age trajectories for personality traits are known to be similar across cultures. To address whether stereotypes of age groups reflect these age-related changes in personality, we asked participants in 26 countries (N = 3,323) to rate typical adolescents, adults, and old persons in their own country. Raters across nations tended to share similar beliefs about different age groups; adolescents were seen as impulsive, rebellious, undisciplined, preferring excitement and novelty, whereas old people were consistently considered lower on impulsivity, activity, antagonism, and Openness. These consensual age group stereotypes correlated strongly with published age differences on the five major dimensions of personality and most of 30 specific traits, using as criteria of accuracy both self-reports and observer ratings, different survey methodologies, and data from up to 50 nations. However, personal stereotypes were considerably less accurate, and consensual stereotypes tended to exaggerate differences across age groups.