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.

Managing memory and reducing I/O cost for correlation matrix calculation in bioinformatics

The generation of a correlation matrix from a large set of long gene sequences is a common requirement in many bioinformatics problems such as phylogenetic analysis. The generation is not only computationally intensive but also requires significant memory resources as, typically, few gene sequences can be simultaneously stored in primary memory. The standard practice in such computation is to use frequent input/output (I/O) operations. Therefore, minimizing the number of these operations will yield much faster run-times. This paper develops an approach for the faster and scalable computing of large-size correlation matrices through the full use of available memory and a reduced number of I/O operations. The approach is scalable in the sense that the same algorithms can be executed on different computing platforms with different amounts of memory and can be applied to different problems with different correlation matrix sizes. The significant performance improvement of the approach over the existing approaches is demonstrated through benchmark examples.

Conceptual model linking knowledge management with organisational performance

Better management of knowledge assets has the potential to improve business processes and increase productivity. This fact has led to considerable interest in recent years in the knowledge management (KM) phenomenon, and in the main dimensions that can impact on its application in construction. However, a lack of a systematic way of assessing KM initia-tives’ contribution towards achieving organisational business objectives is evident. This paper describes the first stage of a research project intended to develop, and empirically test, a KM input-process-output framework comprising unique and well-defined theoretical constructs representing the KM process and its internal and external determinants in the context of con-struction. The paper presents the underlying principles used in operationally defining each construct through the use of extant KM literature. The KM process itself is explicitly mod-elled via a number of clearly articulated phases that ultimately lead to knowledge utilisation and capitalisation, which in turn adds value or otherwise to meeting defined business objec-tives. The main objective of the model is to reduce the impact of subjectivity in assessing the contribution made by KM practices and initiatives toward achieving performance improvements.

Modelling the Contribution of Knowledge Management towards Achieving Orgnisational Business Objectives

Knowledge management (KM) continues to receive mounting interest within the construction industry due to its potential to offer solutions for organisations seeking competitive advantage. This paper presents a KM input-process-output conceptual model comprising unique and well-defined theoretical constructs representing KM practices and their internal and external determinants in the context of construction. The paper also presents the underlying principles used in operationally defining each construct using extant KM literature, and offers a number of testable hypotheses that capture the inter-relationships between the identified constructs.

The implications of cross-regional differences for the design of In-vehicle Information Systems : a comparison of Australian and Chinese drivers

The increasing global distribution of automobiles necessitates that the design of In-vehicle Information Systems (IVIS) is appropriate for the regions to which they are being exported. Differences between regions such as culture, environment and traffic context can influence the needs, usability and acceptance of IVIS. This paper describes two studies aimed at identifying regional differences in IVIS design needs and preferences across drivers from Australia and China to determine the impact of any differences on IVIS design. Using a questionnaire and interaction clinics, the influence of cultural values and driving patterns on drivers' preferences for, and comprehension of, surface- and interaction-level aspects of IVIS interfaces was explored. Similarities and differences were found between the two regional groups in terms of preferences for IVIS input control types and labels and in the comprehension of IVIS functions. Specifically, Chinese drivers preferred symbols and Chinese characters over English words and were less successful (compared to Australians) at comprehending English abbreviations, particularly for complex IVIS functions. Implications in terms of the current trend to introduce Western-styled interfaces into other regions with little or no adaptation are discussed.

Alcohol and other drugs in the Australian construction industry : a pathway for safety focused cultural change

Background: Anecdotal evidence from the infrastructure and building sectors highlights issues of alcohol and other drugs (AODs) and its association with safety risk on construction sites. Currently, there is no clear evidence on the prevalence and risk of AOD use among Australian construction workers and there is limited evidential guidance regarding how to effectively address such an issue. Aims: The current research aims to scientifically evaluate the use of AODs within the Australian construction industry in order to reduce the potential resulting safety and performance impacts and engender a cultural change in the workforce. A nationally consistent and collaborative approach across the workforce will be adopted. Methods: A national assessment of the use of AODs was conducted in participating organisations across three states. The World Health Organisation’s Alcohol Use Disorders Identification Test (AUDIT) was used to measure alcohol use. Illicit drug use, ‘readiness to change’, impediments to reducing impairment, feasibility of proposed interventions, and employee attitudes and knowledge regarding AOD was also measured through a combination of survey items and interviews. Through an educative approach and consultation with employers, employees, union groups and leaders in applied AOD research, this assessment was used to inform and support cultural change management of AOD use in the industry. Results: Results (n=494) indicate that as in the general population, a proportion of those sampled in the construction sector may be at risk of hazardous alcohol consumption. A total of 286 respondents (58%) scored above the cut-off cumulative score for risky or hazardous alcohol use. Other drug use was also identified as a major issue. Interview responses and input from all project partners is presented within a guiding principle framework for cultural change. Conclusions: Results support the need for evidence-based, comprehensive and tailored responses in the workplace. This paper will discuss the final results in the context of facilitating cultural change in the construction industry.

A Systematic review of a statistical vacuum circuit breaker model for condition monitoring

Vacuum circuit breaker (VCB) overvoltage failure and its catastrophic failures during shunt reactor switching have been analyzed through computer simulations for multiple reignitions with a statistical VCB model found in the literature. However, a systematic review (SR) that is related to the multiple reignitions with a statistical VCB model does not yet exist. Therefore, this paper aims to analyze and explore the multiple reignitions with a statistical VCB model. It examines the salient points, research gaps and limitations of the multiple reignition phenomenon to assist with future investigations following the SR search. Based on the SR results, seven issues and two approaches to enhance the current statistical VCB model are identified. These results will be useful as an input to improve the computer modeling accuracy as well as the development of a reignition switch model with point-on-wave controlled switching for condition monitoring

Slotted cooperative communication in half duplex relay networks

In cooperative communication systems, several wireless communication terminals collaborate to form a virtual-multiple antenna array system and exploit the spatial diversity to achieve a better performance. This thesis proposes a practical slotted protocol for cooperative communication systems with half-duplex single antennas. The performance of the proposed slotted cooperative communication protocol is evaluated in terms of the pairwise error probability and the bit error rate. The proposed protocol achieves the multiple-input single-output performance bound with a novel relay ordering and scheduling strategy.

Computer experiment – a case study for modelling and simulation of manufacturing systems

Deterministic computer simulation of physical experiments is now a common technique in science and engineering. Often, physical experiments are too time consuming, expensive or impossible to conduct. Complex computer models or codes, rather than physical experiments lead to the study of computer experiments, which are used to investigate many scientific phenomena. A computer experiment consists of a number of runs of the computer code with different input choices. The Design and Analysis of Computer Experiments is a rapidly growing technique in statistical experimental design. This paper aims to discuss some practical issues when designing a computer simulation and/or experiments for manufacturing systems. A case study approach is reviewed and presented.

Some practical issues in the design and analysis of computer experiments

Deterministic computer simulations of physical experiments are now common techniques in science and engineering. Often, physical experiments are too time consuming, expensive or impossible to conduct. Complex computer models or codes, rather than physical experiments lead to the study of computer experiments, which are used to investigate many scientific phenomena of this nature. A computer experiment consists of a number of runs of the computer code with different input choices. The Design and Analysis of Computer Experiments is a rapidly growing technique in statistical experimental design. This thesis investigates some practical issues in the design and analysis of computer experiments and attempts to answer some of the questions faced by experimenters using computer experiments. In particular, the question of the number of computer experiments and how they should be augmented is studied and attention is given to when the response is a function over time.

Implementation guide for surveillance of Clostridium difficile infection

The Implementation Guide for hospital surveillance of Clostridium difficile infection (CDI) has been produced by the Healthcare Associated Infection (HAI) Technical Working Group of the Australian Commission on Safety and Quality in Health Care (ACSQHC), and endorsed by the HAI Advisory Group. State jurisdictions and the ACSQHC have representatives on the Technical Working Group, and have had input into this document. The Guide is intended to be used by Australian hospitals and organisations to support the implementation of hospital-identified Clostridium difficile infection (CDI) surveillance using the endorsed case definition in this guide. It has been produced to support consistency of surveillance activities and is not intended to replace clinical assessment of infection for patient management.

Analysis of stream cipher based authenticated encryption schemes

Authenticated Encryption (AE) is the cryptographic process of providing simultaneous confidentiality and integrity protection to messages. This approach is more efficient than applying a two-step process of providing confidentiality for a message by encrypting the message, and in a separate pass providing integrity protection by generating a Message Authentication Code (MAC). AE using symmetric ciphers can be provided by either stream ciphers with built in authentication mechanisms or block ciphers using appropriate modes of operation. However, stream ciphers have the potential for higher performance and smaller footprint in hardware and/or software than block ciphers. This property makes stream ciphers suitable for resource constrained environments, where storage and computational power are limited. There have been several recent stream cipher proposals that claim to provide AE. These ciphers can be analysed using existing techniques that consider confidentiality or integrity separately; however currently there is no existing framework for the analysis of AE stream ciphers that analyses these two properties simultaneously. This thesis introduces a novel framework for the analysis of AE using stream cipher algorithms. This thesis analyzes the mechanisms for providing confidentiality and for providing integrity in AE algorithms using stream ciphers. There is a greater emphasis on the analysis of the integrity mechanisms, as there is little in the public literature on this, in the context of authenticated encryption. The thesis has four main contributions as follows. The first contribution is the design of a framework that can be used to classify AE stream ciphers based on three characteristics. The first classification applies Bellare and Namprempre's work on the the order in which encryption and authentication processes take place. The second classification is based on the method used for accumulating the input message (either directly or indirectly) into the into the internal states of the cipher to generate a MAC. The third classification is based on whether the sequence that is used to provide encryption and authentication is generated using a single key and initial vector, or two keys and two initial vectors. The second contribution is the application of an existing algebraic method to analyse the confidentiality algorithms of two AE stream ciphers; namely SSS and ZUC. The algebraic method is based on considering the nonlinear filter (NLF) of these ciphers as a combiner with memory. This method enables us to construct equations for the NLF that relate the (inputs, outputs and memory of the combiner) to the output keystream. We show that both of these ciphers are secure from this type of algebraic attack. We conclude that using a keydependent SBox in the NLF twice, and using two different SBoxes in the NLF of ZUC, prevents this type of algebraic attack. The third contribution is a new general matrix based model for MAC generation where the input message is injected directly into the internal state. This model describes the accumulation process when the input message is injected directly into the internal state of a nonlinear filter generator. We show that three recently proposed AE stream ciphers can be considered as instances of this model; namely SSS, NLSv2 and SOBER-128. Our model is more general than a previous investigations into direct injection. Possible forgery attacks against this model are investigated. It is shown that using a nonlinear filter in the accumulation process of the input message when either the input message or the initial states of the register is unknown prevents forgery attacks based on collisions. The last contribution is a new general matrix based model for MAC generation where the input message is injected indirectly into the internal state. This model uses the input message as a controller to accumulate a keystream sequence into an accumulation register. We show that three current AE stream ciphers can be considered as instances of this model; namely ZUC, Grain-128a and Sfinks. We establish the conditions under which the model is susceptible to forgery and side-channel attacks.

Robust image hash functions using higher order spectra

Robust hashing is an emerging field that can be used to hash certain data types in applications unsuitable for traditional cryptographic hashing methods. Traditional hashing functions have been used extensively for data/message integrity, data/message authentication, efficient file identification and password verification. These applications are possible because the hashing process is compressive, allowing for efficient comparisons in the hash domain but non-invertible meaning hashes can be used without revealing the original data. These techniques were developed with deterministic (non-changing) inputs such as files and passwords. For such data types a 1-bit or one character change can be significant, as a result the hashing process is sensitive to any change in the input. Unfortunately, there are certain applications where input data are not perfectly deterministic and minor changes cannot be avoided. Digital images and biometric features are two types of data where such changes exist but do not alter the meaning or appearance of the input. For such data types cryptographic hash functions cannot be usefully applied. In light of this, robust hashing has been developed as an alternative to cryptographic hashing and is designed to be robust to minor changes in the input. Although similar in name, robust hashing is fundamentally different from cryptographic hashing. Current robust hashing techniques are not based on cryptographic methods, but instead on pattern recognition techniques. Modern robust hashing algorithms consist of feature extraction followed by a randomization stage that introduces non-invertibility and compression, followed by quantization and binary encoding to produce a binary hash output. In order to preserve robustness of the extracted features, most randomization methods are linear and this is detrimental to the security aspects required of hash functions. Furthermore, the quantization and encoding stages used to binarize real-valued features requires the learning of appropriate quantization thresholds. How these thresholds are learnt has an important effect on hashing accuracy and the mere presence of such thresholds are a source of information leakage that can reduce hashing security. This dissertation outlines a systematic investigation of the quantization and encoding stages of robust hash functions. While existing literature has focused on the importance of quantization scheme, this research is the first to emphasise the importance of the quantizer training on both hashing accuracy and hashing security. The quantizer training process is presented in a statistical framework which allows a theoretical analysis of the effects of quantizer training on hashing performance. This is experimentally verified using a number of baseline robust image hashing algorithms over a large database of real world images. This dissertation also proposes a new randomization method for robust image hashing based on Higher Order Spectra (HOS) and Radon projections. The method is non-linear and this is an essential requirement for non-invertibility. The method is also designed to produce features more suited for quantization and encoding. The system can operate without the need for quantizer training, is more easily encoded and displays improved hashing performance when compared to existing robust image hashing algorithms. The dissertation also shows how the HOS method can be adapted to work with biometric features obtained from 2D and 3D face images.

Improving ultrasound excitation systems using a flexible power supply with adjustable voltage and frequency to drive piezoelectric transducers

The ability of a piezoelectric transducer in energy conversion is rapidly expanding in several applications. Some of the industrial applications for which a high power ultrasound transducer can be used are surface cleaning, water treatment, plastic welding and food sterilization. Also, a high power ultrasound transducer plays a great role in biomedical applications such as diagnostic and therapeutic applications. An ultrasound transducer is usually applied to convert electrical energy to mechanical energy and vice versa. In some high power ultrasound system, ultrasound transducers are applied as a transmitter, as a receiver or both. As a transmitter, it converts electrical energy to mechanical energy while a receiver converts mechanical energy to electrical energy as a sensor for control system. Once a piezoelectric transducer is excited by electrical signal, piezoelectric material starts to vibrate and generates ultrasound waves. A portion of the ultrasound waves which passes through the medium will be sensed by the receiver and converted to electrical energy. To drive an ultrasound transducer, an excitation signal should be properly designed otherwise undesired signal (low quality) can deteriorate the performance of the transducer (energy conversion) and increase power consumption in the system. For instance, some portion of generated power may be delivered in unwanted frequency which is not acceptable for some applications especially for biomedical applications. To achieve better performance of the transducer, along with the quality of the excitation signal, the characteristics of the high power ultrasound transducer should be taken into consideration as well. In this regard, several simulation and experimental tests are carried out in this research to model high power ultrasound transducers and systems. During these experiments, high power ultrasound transducers are excited by several excitation signals with different amplitudes and frequencies, using a network analyser, a signal generator, a high power amplifier and a multilevel converter. Also, to analyse the behaviour of the ultrasound system, the voltage ratio of the system is measured in different tests. The voltage across transmitter is measured as an input voltage then divided by the output voltage which is measured across receiver. The results of the transducer characteristics and the ultrasound system behaviour are discussed in chapter 4 and 5 of this thesis. Each piezoelectric transducer has several resonance frequencies in which its impedance has lower magnitude as compared to non-resonance frequencies. Among these resonance frequencies, just at one of those frequencies, the magnitude of the impedance is minimum. This resonance frequency is known as the main resonance frequency of the transducer. To attain higher efficiency and deliver more power to the ultrasound system, the transducer is usually excited at the main resonance frequency. Therefore, it is important to find out this frequency and other resonance frequencies. Hereof, a frequency detection method is proposed in this research which is discussed in chapter 2. An extended electrical model of the ultrasound transducer with multiple resonance frequencies consists of several RLC legs in parallel with a capacitor. Each RLC leg represents one of the resonance frequencies of the ultrasound transducer. At resonance frequency the inductor reactance and capacitor reactance cancel out each other and the resistor of this leg represents power conversion of the system at that frequency. This concept is shown in simulation and test results presented in chapter 4. To excite a high power ultrasound transducer, a high power signal is required. Multilevel converters are usually applied to generate a high power signal but the drawback of this signal is low quality in comparison with a sinusoidal signal. In some applications like ultrasound, it is extensively important to generate a high quality signal. Several control and modulation techniques are introduced in different papers to control the output voltage of the multilevel converters. One of those techniques is harmonic elimination technique. In this technique, switching angles are chosen in such way to reduce harmonic contents in the output side. It is undeniable that increasing the number of the switching angles results in more harmonic reduction. But to have more switching angles, more output voltage levels are required which increase the number of components and cost of the converter. To improve the quality of the output voltage signal with no more components, a new harmonic elimination technique is proposed in this research. Based on this new technique, more variables (DC voltage levels and switching angles) are chosen to eliminate more low order harmonics compared to conventional harmonic elimination techniques. In conventional harmonic elimination method, DC voltage levels are same and only switching angles are calculated to eliminate harmonics. Therefore, the number of eliminated harmonic is limited by the number of switching cycles. In the proposed modulation technique, the switching angles and the DC voltage levels are calculated off-line to eliminate more harmonics. Therefore, the DC voltage levels are not equal and should be regulated. To achieve this aim, a DC/DC converter is applied to adjust the DC link voltages with several capacitors. The effect of the new harmonic elimination technique on the output quality of several single phase multilevel converters is explained in chapter 3 and 6 of this thesis. According to the electrical model of high power ultrasound transducer, this device can be modelled as parallel combinations of RLC legs with a main capacitor. The impedance diagram of the transducer in frequency domain shows it has capacitive characteristics in almost all frequencies. Therefore, using a voltage source converter to drive a high power ultrasound transducer can create significant leakage current through the transducer. It happens due to significant voltage stress (dv/dt) across the transducer. To remedy this problem, LC filters are applied in some applications. For some applications such as ultrasound, using a LC filter can deteriorate the performance of the transducer by changing its characteristics and displacing the resonance frequency of the transducer. For such a case a current source converter could be a suitable choice to overcome this problem. In this regard, a current source converter is implemented and applied to excite the high power ultrasound transducer. To control the output current and voltage, a hysteresis control and unipolar modulation are used respectively. The results of this test are explained in chapter 7.

Monte Carlo verification of intensity modulated radiotherapy treatments using EPIDS

Purpose: Electronic Portal Imaging Devices (EPIDs) are available with most linear accelerators (Amonuk, 2002), the current technology being amorphous silicon flat panel imagers. EPIDs are currently used routinely in patient positioning before radiotherapy treatments. There has been an increasing interest in using EPID technology tor dosimetric verification of radiotherapy treatments (van Elmpt, 2008). A straightforward technique involves the EPID panel being used to measure the fluence exiting the patient during a treatment which is then compared to a prediction of the fluence based on the treatment plan. However, there are a number of significant limitations which exist in this Method: Resulting in a limited proliferation ot this technique in a clinical environment. In this paper, we aim to present a technique of simulating IMRT fields using Monte Carlo to predict the dose in an EPID which can then be compared to the measured dose in the EPID. Materials: Measurements were made using an iView GT flat panel a-SI EPfD mounted on an Elekta Synergy linear accelerator. The images from the EPID were acquired using the XIS software (Heimann Imaging Systems). Monte Carlo simulations were performed using the BEAMnrc and DOSXVZnrc user codes. The IMRT fieids to be delivered were taken from the treatment planning system in DICOMRT format and converted into BEAMnrc and DOSXYZnrc input files using an in-house application (Crowe, 2009). Additionally. all image processing and analysis was performed using another in-house application written using the Interactive Data Language (IDL) (In Visual Information Systems). Comparison between the measured and Monte Carlo EPID images was performed using a gamma analysis (Low, 1998) incorporating dose and distance to agreement criteria. Results: The fluence maps recorded by the EPID were found to provide good agreement between measured and simulated data. Figure 1 shows an example of measured and simulated IMRT dose images and profiles in the x and y directions. "A technique for the quantitative evaluation of dose distributions", Med Phys, 25(5) May 1998 S. Crowe, 1. Kairn, A. Fielding, "The Development of a Monte Carlo system to verify Radiotherapy treatment dose calculations", Radiotherapy & Oncology, Volume 92, Supplement 1, August 2009, Pages S71-S71.