Improved simultaneous computation of motion detection and optical flow for object tracking

Object tracking systems require accurate segmentation of the objects from the background for effective tracking. Motion segmentation or optical flow can be used to segment incoming images. Whilst optical flow allows multiple moving targets to be separated based on their individual velocities, optical flow techniques are prone to errors caused by changing lighting and occlusions, both common in a surveillance environment. Motion segmentation techniques are more robust to fluctuating lighting and occlusions, but don't provide information on the direction of the motion. In this paper we propose a combined motion segmentation/optical flow algorithm for use in object tracking. The proposed algorithm uses the motion segmentation results to inform the optical flow calculations and ensure that optical flow is only calculated in regions of motion, and improve the performance of the optical flow around the edge of moving objects. Optical flow is calculated at pixel resolution and tracking of flow vectors is employed to improve performance and detect discontinuities, which can indicate the location of overlaps between objects. The algorithm is evaluated by attempting to extract a moving target within the flow images, given expected horizontal and vertical movement (i.e. the algorithms intended use for object tracking). Results show that the proposed algorithm outperforms other widely used optical flow techniques for this surveillance application.

Computational fluid dynamics for improved bioreactor design and 3D culture

The complex relationship between the hydrodynamic environment and surrounding tissues directly impacts on the design and production of clinically useful grafts and implants. Tissue engineers have generally seen bioreactors as 'black boxes' within which tissue engineering constructs (TECs) are cultured. It is accepted that a more detailed description of fluid mechanics and nutrient transport within process equipment can be achieved by using computational fluid dynamics (CFD) technology. This review discusses applications of CFD for tissue engineering-related bioreactors -- fluid flow processes have direct implications on cellular responses such as attachment, migration and proliferation. We conclude that CFD should be seen as an invaluable tool for analyzing and visualizing the impact of fluidic forces and stresses on cells and TECs.

The Revised Transactional Model (RTM) of occupational stress and coping : an improved process approach

Despite more than three decades of research, there is a limited understanding of the transactional processes of appraisal, stress and coping. This has led to calls for more focused research on the entire process that underlies these variables. To date, there remains a paucity of such research. The present study examined Lazarus and Folkman’s (1984) transactional model of stress and coping. One hundred and twenty nine Australian participants with full time employment (i.e. nurses and administration employees) were recruited. There were 49 male (age mean = 34, SD = 10.51) and 80 female (age mean = 36, SD = 10.31) participants. The analysis of three path models indicated that in addition to the original paths, which were found in Lazarus and Folkman’s transactional model (primary appraisal-->secondary appraisal-->stress-->coping), there were also direct links between primary appraisal and stress level time one and between stress level time one to stress level time two. This study has provided additional insights into the transactional process which will extend our understanding of how individuals appraise, cope and experience occupational stress.

Improving automatic speaker verification using SVM techniques

Automatic recognition of people is an active field of research with important forensic and security applications. In these applications, it is not always possible for the subject to be in close proximity to the system. Voice represents a human behavioural trait which can be used to recognise people in such situations. Automatic Speaker Verification (ASV) is the process of verifying a persons identity through the analysis of their speech and enables recognition of a subject at a distance over a telephone channel { wired or wireless. A significant amount of research has focussed on the application of Gaussian mixture model (GMM) techniques to speaker verification systems providing state-of-the-art performance. GMM's are a type of generative classifier trained to model the probability distribution of the features used to represent a speaker. Recently introduced to the field of ASV research is the support vector machine (SVM). An SVM is a discriminative classifier requiring examples from both positive and negative classes to train a speaker model. The SVM is based on margin maximisation whereby a hyperplane attempts to separate classes in a high dimensional space. SVMs applied to the task of speaker verification have shown high potential, particularly when used to complement current GMM-based techniques in hybrid systems. This work aims to improve the performance of ASV systems using novel and innovative SVM-based techniques. Research was divided into three main themes: session variability compensation for SVMs; unsupervised model adaptation; and impostor dataset selection. The first theme investigated the differences between the GMM and SVM domains for the modelling of session variability | an aspect crucial for robust speaker verification. Techniques developed to improve the robustness of GMMbased classification were shown to bring about similar benefits to discriminative SVM classification through their integration in the hybrid GMM mean supervector SVM classifier. Further, the domains for the modelling of session variation were contrasted to find a number of common factors, however, the SVM-domain consistently provided marginally better session variation compensation. Minimal complementary information was found between the techniques due to the similarities in how they achieved their objectives. The second theme saw the proposal of a novel model for the purpose of session variation compensation in ASV systems. Continuous progressive model adaptation attempts to improve speaker models by retraining them after exploiting all encountered test utterances during normal use of the system. The introduction of the weight-based factor analysis model provided significant performance improvements of over 60% in an unsupervised scenario. SVM-based classification was then integrated into the progressive system providing further benefits in performance over the GMM counterpart. Analysis demonstrated that SVMs also hold several beneficial characteristics to the task of unsupervised model adaptation prompting further research in the area. In pursuing the final theme, an innovative background dataset selection technique was developed. This technique selects the most appropriate subset of examples from a large and diverse set of candidate impostor observations for use as the SVM background by exploiting the SVM training process. This selection was performed on a per-observation basis so as to overcome the shortcoming of the traditional heuristic-based approach to dataset selection. Results demonstrate the approach to provide performance improvements over both the use of the complete candidate dataset and the best heuristically-selected dataset whilst being only a fraction of the size. The refined dataset was also shown to generalise well to unseen corpora and be highly applicable to the selection of impostor cohorts required in alternate techniques for speaker verification.

Data-driven background dataset selection for SVM-based speaker verification

The recently proposed data-driven background dataset refinement technique provides a means of selecting an informative background for support vector machine (SVM)-based speaker verification systems. This paper investigates the characteristics of the impostor examples in such highly-informative background datasets. Data-driven dataset refinement individually evaluates the suitability of candidate impostor examples for the SVM background prior to selecting the highest-ranking examples as a refined background dataset. Further, the characteristics of the refined dataset were analysed to investigate the desired traits of an informative SVM background. The most informative examples of the refined dataset were found to consist of large amounts of active speech and distinctive language characteristics. The data-driven refinement technique was shown to filter the set of candidate impostor examples to produce a more disperse representation of the impostor population in the SVM kernel space, thereby reducing the number of redundant and less-informative examples in the background dataset. Furthermore, data-driven refinement was shown to provide performance gains when applied to the difficult task of refining a small candidate dataset that was mis-matched to the evaluation conditions.

Exploiting multiple feature sets in data-driven impostor dataset selection for speaker verification

This study assesses the recently proposed data-driven background dataset refinement technique for speaker verification using alternate SVM feature sets to the GMM supervector features for which it was originally designed. The performance improvements brought about in each trialled SVM configuration demonstrate the versatility of background dataset refinement. This work also extends on the originally proposed technique to exploit support vector coefficients as an impostor suitability metric in the data-driven selection process. Using support vector coefficients improved the performance of the refined datasets in the evaluation of unseen data. Further, attempts are made to exploit the differences in impostor example suitability measures from varying features spaces to provide added robustness.

Improved image contrast of the bone-muscle interface with 3T MRI compared to 1.5T MRI [Abstract]

Virtual 3D models of long bones are increasingly being used for implant design and research applications. The current gold standard for the acquisition of such data is Computed Tomography (CT) scanning. Due to radiation exposure, CT is generally limited to the imaging of clinical cases and cadaver specimens. Magnetic Resonance Imaging (MRI) does not involve ionising radiation and therefore can be used to image selected healthy human volunteers for research purposes. The feasibility of MRI as alternative to CT for the acquisition of morphological bone data of the lower extremity has been demonstrated in recent studies [1, 2]. Some of the current limitations of MRI are long scanning times and difficulties with image segmentation in certain anatomical regions due to poor contrast between bone and surrounding muscle tissues. Higher field strength scanners promise to offer faster imaging times or better image quality. In this study image quality at 1.5T is quantitatively compared to images acquired at 3T. --------- The femora of five human volunteers were scanned using 1.5T and 3T MRI scanners from the same manufacturer (Siemens) with similar imaging protocols. A 3D flash sequence was used with TE = 4.66 ms, flip angle = 15° and voxel size = 0.5 × 0.5 × 1 mm. PA-Matrix and body matrix coils were used to cover the lower limb and pelvis respectively. Signal to noise ratio (SNR) [3] and contrast to noise ratio (CNR) [3] of the axial images from the proximal, shaft and distal regions were used to assess the quality of images from the 1.5T and 3T scanners. The SNR was calculated for the muscle and bone-marrow in the axial images. The CNR was calculated for the muscle to cortex and cortex to bone marrow interfaces, respectively. --------- Preliminary results (one volunteer) show that the SNR of muscle for the shaft and distal regions was higher in 3T images (11.65 and 17.60) than 1.5T images (8.12 and 8.11). For the proximal region the SNR of muscles was higher in 1.5T images (7.52) than 3T images (6.78). The SNR of bone marrow was slightly higher in 1.5T images for both proximal and shaft regions, while it was lower in the distal region compared to 3T images. The CNR between muscle and bone of all three regions was higher in 3T images (4.14, 6.55 and 12.99) than in 1.5T images (2.49, 3.25 and 9.89). The CNR between bone-marrow and bone was slightly higher in 1.5T images (4.87, 12.89 and 10.07) compared to 3T images (3.74, 10.83 and 10.15). These results show that the 3T images generated higher contrast between bone and the muscle tissue than the 1.5T images. It is expected that this improvement of image contrast will significantly reduce the time required for the mainly manual segmentation of the MR images. Future work will focus on optimizing the 3T imaging protocol for reducing chemical shift and susceptibility artifacts.

Improved joint control using a genetic algorithm for a humanoid robot

This paper describes experiments conducted in order to simultaneously tune 15 joints of a humanoid robot. Two Genetic Algorithm (GA) based tuning methods were developed and compared against a hand-tuned solution. The system was tuned in order to minimise tracking error while at the same time achieve smooth joint motion. Joint smoothness is crucial for the accurate calculation of online ZMP estimation, a prerequisite for a closedloop dynamically stable humanoid walking gait. Results in both simulation and on a real robot are presented, demonstrating the superior smoothness performance of the GA based methods.

Improved correlation-based modal strain energy method for global damage detection of truss bridge structures

The previous investigations have shown that the modal strain energy correlation method, MSEC, could successfully identify the damage of truss bridge structures. However, it has to incorporate the sensitivity matrix to estimate damage and is not reliable in certain damage detection cases. This paper presents an improved MSEC method where the prediction of modal strain energy change vector is differently obtained by running the eigensolutions on-line in optimisation iterations. The particular trail damage treatment group maximising the fitness function close to unity is identified as the detected damage location. This improvement is then compared with the original MSEC method along with other typical correlation-based methods on the finite element model of a simple truss bridge. The contributions to damage detection accuracy of each considered mode is also weighed and discussed. The iterative searching process is operated by using genetic algorithm. The results demonstrate that the improved MSEC method suffices the demand in detecting the damage of truss bridge structures, even when noised measurement is considered.

Accurate silhouettes for surveillance : improved motion segmentation using graph cuts

Silhouettes are common features used by many applications in computer vision. For many of these algorithms to perform optimally, accurately segmenting the objects of interest from the background to extract the silhouettes is essential. Motion segmentation is a popular technique to segment moving objects from the background, however such algorithms can be prone to poor segmentation, particularly in noisy or low contrast conditions. In this paper, the work of [3] combining motion detection with graph cuts, is extended into two novel implementations that aim to allow greater uncertainty in the output of the motion segmentation, providing a less restricted input to the graph cut algorithm. The proposed algorithms are evaluated on a portion of the ETISEO dataset using hand segmented ground truth data, and an improvement in performance over the motion segmentation alone and the baseline system of [3] is shown.

Improved algebraic cryptanalysis of QUAD, Bivium and Trivium via graph partitioning on equation systems

We present a novel approach for preprocessing systems of polynomial equations via graph partitioning. The variable-sharing graph of a system of polynomial equations is defined. If such graph is disconnected, then the corresponding system of equations can be split into smaller ones that can be solved individually. This can provide a tremendous speed-up in computing the solution to the system, but is unlikely to occur either randomly or in applications. However, by deleting certain vertices on the graph, the variable-sharing graph could be disconnected in a balanced fashion, and in turn the system of polynomial equations would be separated into smaller systems of near-equal sizes. In graph theory terms, this process is equivalent to finding balanced vertex partitions with minimum-weight vertex separators. The techniques of finding these vertex partitions are discussed, and experiments are performed to evaluate its practicality for general graphs and systems of polynomial equations. Applications of this approach in algebraic cryptanalysis on symmetric ciphers are presented: For the QUAD family of stream ciphers, we show how a malicious party can manufacture conforming systems that can be easily broken. For the stream ciphers Bivium and Trivium, we nachieve significant speedups in algebraic attacks against them, mainly in a partial key guess scenario. In each of these cases, the systems of polynomial equations involved are well-suited to our graph partitioning method. These results may open a new avenue for evaluating the security of symmetric ciphers against algebraic attacks.

Improved power sharing among distributed generators using web based communication

This paper investigates the possibility of power sharing improvements amongst distributed generators with low cost, low bandwidth communications. Decentralized power sharing or power management can be improved significantly with low bandwidth communication. Utility intranet or a dedicated web based communication can serve the purpose. The effect of network parameter such line impedance, R/X ratio on decentralized power sharing can be compensated with correction in the decentralized control reference quantities through the low bandwidth communication. In this paper, the possible improvement is demonstrated in weak system condition, where the micro sources and the loads are not symmetrical along the rural microgrid with high R/X ratio line, creates challenge for decentralized control. In those cases the web based low bandwidth communication is economic and justified than costly advance high bandwidth communication.

Improved functional ability and independence in activities of daily living for older adults at high risk of hospital readmission : a randomised controlled trial

Objective: During hospitalisation older people often experience functional decline which impacts on their future independence. The objective of this study was to evaluate a multifaceted transitional care intervention including home-based exercise strategies for at-risk older people on functional status, independence in activities of daily living, and walking ability. Methods: A randomised controlled trial was undertaken in a metropolitan hospital in Australia with 128 patients (64 intervention, 64 control) aged over 65 years with an acute medical admission and at least one risk factor for hospital readmission. The intervention group received an individually tailored program for exercise and follow-up care which was commenced in hospital and included regular visits in hospital by a physiotherapist and a Registered Nurse, a home visit following discharge, and regular telephone follow-up for 24 weeks following discharge. The program was designed to improve health promoting behaviours, strength, stability, endurance and mobility. Data were collected at baseline, then 4, 12 and 24 weeks following discharge using the Index of Activities of Daily Living (ADL), Instrumental Index of Activities of Daily Living (IADL), and the Walking Impairment Questionnaire (Modified). Results: Significant improvements were found in the intervention group in IADL scores (p<.001), ADL scores (p<.001), and WIQ scale scores (p<.001) in comparison to the control group. The greatest improvements were found in the first four weeks following discharge. Conclusions: Early introduction of a transitional model of care incorporating a tailored exercise program and regular telephone follow-up for hospitalised at-risk older adults can improve independence and functional ability.

Towards improved speech recognition for resource poor languages

In recent times, the improved levels of accuracy obtained by Automatic Speech Recognition (ASR) technology has made it viable for use in a number of commercial products. Unfortunately, these types of applications are limited to only a few of the world’s languages, primarily because ASR development is reliant on the availability of large amounts of language specific resources. This motivates the need for techniques which reduce this language-specific, resource dependency. Ideally, these approaches should generalise across languages, thereby providing scope for rapid creation of ASR capabilities for resource poor languages. Cross Lingual ASR emerges as a means for addressing this need. Underpinning this approach is the observation that sound production is largely influenced by the physiological construction of the vocal tract, and accordingly, is human, and not language specific. As a result, a common inventory of sounds exists across languages; a property which is exploitable, as sounds from a resource poor, target language can be recognised using models trained on resource rich, source languages. One of the initial impediments to the commercial uptake of ASR technology was its fragility in more challenging environments, such as conversational telephone speech. Subsequent improvements in these environments has gained consumer confidence. Pragmatically, if cross lingual techniques are to considered a viable alternative when resources are limited, they need to perform under the same types of conditions. Accordingly, this thesis evaluates cross lingual techniques using two speech environments; clean read speech and conversational telephone speech. Languages used in evaluations are German, Mandarin, Japanese and Spanish. Results highlight that previously proposed approaches provide respectable results for simpler environments such as read speech, but degrade significantly when in the more taxing conversational environment. Two separate approaches for addressing this degradation are proposed. The first is based on deriving better target language lexical representation, in terms of the source language model set. The second, and ultimately more successful approach, focuses on improving the classification accuracy of context-dependent (CD) models, by catering for the adverse influence of languages specific phonotactic properties. Whilst the primary research goal in this thesis is directed towards improving cross lingual techniques, the catalyst for investigating its use was based on expressed interest from several organisations for an Indonesian ASR capability. In Indonesia alone, there are over 200 million speakers of some Malay variant, provides further impetus and commercial justification for speech related research on this language. Unfortunately, at the beginning of the candidature, limited research had been conducted on the Indonesian language in the field of speech science, and virtually no resources existed. This thesis details the investigative and development work dedicated towards obtaining an ASR system with a 10000 word recognition vocabulary for the Indonesian language.