Neurodevelopmental conditions like Attention Deficit Hyperactivity Disorder (ADHD) and Autism Spectrum Disorder (ASD) impact a significant number of children and adults worldwide. Currently, the means of diagnosing of such conditions is carried out by experts, who employ standard questionnaires and look for certain behavioural markers through manual observation. Such methods are not only subjective, difficult to repeat, and costly but also extremely time consuming. However, with the recent surge of research into automatic facial behaviour analysis and it's varied applications, it could prove to be a potential way of tackling these diagnostic difficulties. Automatic facial expression recognition is one of the core components of this field but it has always been challenging to do it accurately in an unconstrained environment. This thesis presents a dynamic deep learning framework for robust automatic facial expression recognition. It also proposes an approach to apply this method for facial behaviour analysis which can help in the diagnosis of conditions like ADHD and ASD. The proposed facial expression algorithm uses a deep Convolutional Neural Networks (CNN) to learn models of facial Action Units (AU). It attempts to model three main distinguishing features of AUs: shape, appearance and short term dynamics, jointly in a CNN. The appearance is modelled through local image regions relevant to each AU, shape is encoded using binary masks computed from automatically detected facial landmarks and dynamics is encoded by using a short sequence of image as input to CNN. In addition, the method also employs Bi-directional Long Short Memory (BLSTM) recurrent neural networks for modelling long term dynamics. The proposed approach is evaluated on a number of databases showing state-of-the-art performance for both AU detection and intensity estimation tasks. The AU intensities estimated using this approach along with other 3D face tracking data, are used for encoding facial behaviour. The encoded facial behaviour is applied for learning models which can help in detection of ADHD and ASD. This approach was evaluated on the KOMAA database which was specially collected for this purpose. Experimental results show that facial behaviour encoded in this way provide a high discriminative power for classification of people with these conditions. It is shown that the proposed system is a potentially useful, objective and time saving contribution to the clinical diagnosis of ADHD and ASD.

This research is concerned with adaptive, probabilistic single target tracking algorithms. Though visual tracking methods have seen significant improvement, sustained ability to capture appearance changes and precisely locate the target during complex and unexpected motion remains an open problem. Three novel tracking mechanisms are proposed to address these challenges. The first is a Particle Filter based Markov Chain Monte Carlo method with sampled appearances (MCMC-SA). This adapts to changes in target appearance by combining two popular generative models: templates and histograms, maintaining multiple instances of each in an appearance pool. The proposed tracker automatically switches between models in response to variations in target appearance, exploiting the strengths of each model component. New models are added, automatically, as necessary. The second is a Particle Filter based Markov Chain Monte Carlo method with motion direction sampling, from which are derived two variations: motion sampling using a fixed direction of the centroid of all features detected (FMCMC-C) and motion sampling using kernel density estimation of direction (FMCMC-S). This utilises sparse estimates of motion direction derived from local features detected from the target. The tracker captures complex target motions efficiently using only simple components. The third tracking algorithm considered here combines these above methods to improve target localisation. This tracker comprises multiple motion and appearance models (FMCMC-MM) and automatically selects an appropriate motion and appearance model for tracking. The effectiveness of all three tracking algorithms is demonstrated using a variety of challenging video sequences. Results show that these methods considerably improve tracking performance when compared with state of the art appearance-based tracking frameworks.

Since the 1980s, two paradigms have dominated the representation of formatted electronic documents: flowable and fixed. Flowable formats, such as HTML, EPUB, or those used by word processors, allow documents to scale to any arbitrary page size, but typographical compromises must be made since the layout is computed in real time, and is re-computed each time the document is displayed. Conversely, fixed formats such as SVG or PDF are afforded the potential for arbitrarily complex typography, but are constrained to the fixed layout that is set at the time of creation. With the recent surge in popularity of low-powered portable reading devices -- from tablets to e-readers to mobile phones -- there is an expectation that documents should scale to any size, maintain their high-quality typography, and not provide unnecessary strain on an already overloaded battery. This thesis defines a novel paradigm for electronic document representation -- the Malleable Document -- whereby documents are partially typeset at the time of creation, leaving enough flexibility that their content can be flowed to arbitrary page sizes with minimal computation. One tradeoff encountered is that of increased file size, and this is addressed with a bespoke, computationally-light compression scheme. A sample implementation is presented that transforms documents from a source format into Malleable Document format, alongside a lightweight display engine that enables the documents to be viewed and resized on a wide range of devices, mobile and otherwise. Reviews of the technical aspects and a user study to evaluate the quality of the system's rendering and layout show that the Malleable Document paradigm is a promising alternative to both fixed and flowable formats, and builds upon the best of both approaches.

The process through which new blood vessels are formed within the body is known as angiogenesis. An essential part of our survival, it has also been implicated more recently in many diseases both in terms of induced growth, and abnormal vascular structure. Angiogenesis is characterized as two processes, the development of a vascular network during embryonic growth and the production of new blood vessels. This work focuses on the latter, and seeks to develop a robust, three-dimensional model for simulating blood vessel growth and the attendant processes of blood flow and mass transfer within the simulated system. A system was developed which utilises medical imaging scan data (specifically, MicroCT) as the initial conditions from which a network of vessels is grown. This is combined with GPU accelerated simulations of fluid dynamics, with the intention of providing a technique for future use in predictive medicine and therapeutic simulation.

For several decades researchers around the globe have been avidly investigating practical solutions to the enduring problem of understanding visual content within an image. One might think of the quest as an effort to emulate human visual system. Despite all the endeavours, the simplest of visual tasks to us humans, such as optical segmentation of objects, remain a significant challenge for machines. In a few occasions where a computer's processing power is adequate to accomplish the task, the issue of public alienation towards autonomous solutions to critical applications remains unresolved. The principal purpose of this thesis is to propose innovative computer vision, machine learning, and pattern recognition techniques that exploit abstract knowledge of human beings in practical models using facile yet effective methodologies. High-level information provided by users in the decision making loop of such interactive systems enhances the efficacy of vision algorithms, whilst simultaneously machines reduce users' labour by filtering results and completing mundane tasks on their behalf. In this thesis, we initially draw a vivid picture of interactive approaches to vision tasks prior to scrutinising relevant aspects of human in the loop methodologies and highlighting their current shortcomings in object recognition applications. Our survey of literature unveils that the difficulty in harnessing users' abstract knowledge is amongst major complications of human in the loop algorithms. We therefore propose two novel methodologies to capture and model such high-level sources of information. One solution builds innovative textual descriptors that are compatible with discriminative classifiers. The other is based on the random naive Bayes algorithm and is suitable for generative classification frameworks. We further investigate the infamous problem of fusing images' low-level and users' high-level information sources. Our next contribution is therefore a novel random forest based human in the loop framework that efficiently fuses visual features of images with user provided information for fast predictions and a superior classification performance. User abstract knowledge in this method is harnessed in shape of user's answers to perceptual questions about images. In contrast to generative Bayesian frameworks, this is a direct discriminative approach that enables information source fusion in the preliminary stages of the prediction process. We subsequently reveal inventive generative frameworks that model each source of information individually and determine the most effective for the purpose of class label prediction. We propose two innovative and intelligent human in the loop fusion algorithms. Our first algorithm is a modified naive Bayes greedy technique, while our second solution is based on a feedforward neural network. Through experiments on a variety of datasets, we show that our novel intelligent fusion methods of information source selection outperform their competitors in tasks of fine-grained visual categorisation. We additionally present methodologies to reduce unnecessary human involvement in mundane tasks by only focusing on cases where their invaluable abstract knowledge is of utter importance. Our proposed algorithm is based on information theory and recent image annotation techniques. It determines the most efficient sequence of information to obtain from humans involved in the decision making loop, in order to minimise their unnecessary engagement in routine tasks. This approach allows them to be concerned with more abstract functions instead. Our experimental results reveal faster achievement of peak performance in contrast to alternative random ranking systems. Our final major contribution in this thesis is a novel remedy for the curse of dimensionality in pattern recognition problems. It is theoretically based on mutual information and Fano's inequality. Our approach separates the most discriminative descriptors and has the capability to enhance the accuracy of classification algorithms. The process of selecting a subset of relevant features is vital for designing robust human in the loop vision models. Our selection techniques eliminate redundant and irrelevant visual and textual features, and therefore its influence on improvement of various human in the loop algorithms proves to be fundamental in our experiments.

In 1976, the Indonesian government began to implement a new curriculum known as Curriculum 75. This was intended for all government primary, junior secondary and senior secondary schools. This field study, which is written for non-Indonesian readers, examines the Social Science component of Curriculum 75 using elements of the Stake model of curriculum evaluation. The purpose of the evaluation is to show, within the Stake model, the intended antecedents, transactions and outcomes of the Social Science Curriculum; that is, the specifications of the curriculum documents and associated texts before implementation in the classroom. The principal argument of this evaluation is that the worth of the curriculum is discovered in the elements of control which are manifest in intentions. This evaluation selects three issues for closer examination and evaluation; economic development, political culture and the world view which are portrayed in the curriculum. This closer examination reveals that knowledge of economic development is emphasised more than knowledge of political culture. In turn, knowledge of the the world outside and beyond Indonesia which constitute the world view is the least important of the three issues, and particularly at the primary level receives very little attention whatsoever. The specific outcomes of the Social Science Curriculum show that the type of knowledge which is emphasised is cognitive understanding rather than the formation of attitudes and values. Pupils are expected to learn factual knowledge rather than develop personal and social values. The central theme or argument which this evaluation pursues is the notion of curriculum control. Curriculum 75 is the creation of an educational bureaucracy within a bureaucratic state. When understood within the context of the Jackson critique of contemporary Indonesia, curriculum is a bureaucratic function of the state. A principal element of that function is the exercising of control which, for the purpose of this field study, is revealed in a bureaucratic and a curriculum form. The curriculum form of control is the central notion in this evaluation. For the Social Science Curriculum to be successful in the Indonesian classroom, the Curriculum should specify contents and methods in such prescriptive detail that teachers are given a clear programme of what is expected of them. Teachers can only be successful when the Curriculum clearly communicates what teachers are required to do. Control in the sense of constraint is a vital element in that communication. The characteristics of curriculum control are described in terms of the Bernstein concepts of coding and framing. By tracing coding and framing techniques in the Social Science Curriculum in general and each of the three issues in particular, this study shows that the antecedents (as defined by Stake) are strongly controlled (through a Bernstein-type collection code). On the other hand, some curriculum control is lost in the transactions (as defined by Stake) because teaching and learning methods are not sufficiently framed. Future refinements in this Curriculum should concentrate on strengthening curriculum control through increasing the level of specification and prescription about learning methods.

Preventive care for the elderly originated with a study in Great Britain in 1964 that reported a large number of unmet health needs in the elderly and advocated early intervention. Subsequent randomised controlled trials (RCT) used a broad assessment of health including bio-medical, functional, psychological and social /environmental components but inconsistently demonstrated improved outcome for the elderly. 'Health checks' were introduced for all patients in British general practice in 1990. European and American models of care evolved similarly and justify a multidisciplinary team assessment, thorough training of assessment staff and medical supervision of recommendations. Two literature reviews published in 2000 have not reported sound evidence in favour of health assessments. Medicare funding of health assessments for the Australians aged 75 years and over was introduced in November 1999. A protocol for conducting 75+ Health Assessment (75+ HA) was developed and a pilot study was conducted in Yarrawonga in 1995 to initiate Australian research of this model of care. A RCT in the Adelaide Western Division of General Practice tested this model of care. The intervention group (n=50) had two 75+ HA one year apart. The control group (n=50) was left to usual care and had a 75+ HA one year later. Demographic data and the Short Form-36 were used to ensure both groups were comparable. Primary outcome measures did not demonstrate statistically significant reduction in problems nor mortality in the intervention (75+ HA) group compared to the control group. Significant improvements in secondary outcome measures in the intervention group were in self-rated health, depression score and decreased numbers reporting falls. 75+ HAs have been widely taken up by Australian general practitioners. It is no longer possible to conduct a RCT due to the inability to find a legitimate control group. Recommendations arising form this literature review and RCT include; evaluation studies of 75+ HA, concentration on a functional model of health and that nurses or allied health professionals should conduct the assessment in the elderly person's home. A consistent framework for analysis of 75+ HA is proposed. The elderly can be conceived to occupy one of 3 cohorts defined by their function state: No impairment of Activities of Daily Living (ADL), Impairment of Instrumental ADL only or Impairment of Basic ADL. The elderly without ADL impairment have not been demonstrated to benefit from 75+ HA and should be left to access the acute care stream of health services. The most disabled elderly with Basic ADL impairment have not consistently been shown to benefit from 75+ HA probably because they need a more intense level of community care. They should have Care Plans renewed regularly, as tested in the Australian Coordinated Care Trials. The cohort with Instrumental ADL impairment only seems most likely to benefit from annual 75+ HA. An evaluation of screening the elderly for Instrumental and Basic ADL impairment and providing appropriate services for each cohort is recommended. / Thesis (M.D.)--Department of General Practice, 2001.

Biological hydrogen production is a renewable, carbon-neutral and clean route for hydrogen production. Purple non-sulfur (PNS) bacteria have the ability to produce biohydrogen via photofermentation process. The type of the bacterial strain used in photofermentation is known to have an important effect on hydrogen yield. In this study, the effect of different co-cultures of PNS bacteria on photofermentation process was investigated in search of improving the hydrogen yield. For this purpose, growth, hydrogen production and substrate utilization of single and co-cultures of different PNS bacteria (R. capsulatus (DSM 1710), R. capsulatus hup- v (YO3), R. palustris (DSM 127) and R. sphaeroides O.U.001 (DSM 5864)) were compared on artificial H2 production medium in 150 mL photobioreactors under continuous illumination and anaerobic conditions. In general, higher hydrogen yields were obtained via co-cultivation of two different PNS bacteria when compared with single cultures. Further increase in hydrogen yield was observed with co-cultivation of three different PNS bacteria. Co-cultures of two different PNS bacteria have resulted in up to 1.4 and 2.1 fold increase in hydrogen yield and hydrogen productivity. Whereas co-cultures of three different PNS bacteria have resulted in up to 1.6 and 2.0 fold increase in hydrogen yield and hydrogen productivity compared to single cultures. These results indicate that, defined co-cultures of PNS bacteria produce hydrogen at a higher yield and productivity, due most probably to some synergistic relationship. Further studies regarding the physiological and molecular changes need to be carried out for deeper understanding of the mechanism of hydrogen production in co-cultures.

One of the most important parameters affecting hydrogen production in photofermentation process is the type of carbon and nitrogen sources. For this reason in this research, the effect of different nitrogen sources (5mM ammonium chloride and 2mM glutamate) and acetate concentrations (40

The members of the genus Bacillus produce a wide variety of secondary metabolites with antimetabolic and pharmacological activities. These metabolites are mostly small peptides and have unusual components and chemical bonds. These metabolites are synthesized nonribosomally by multifunctional enzyme complexes called peptide synthetases. One of those small peptides, bacilysin, is a dipeptide antibiotic composed of L-alanine and L-anticapsin which is produced and excreted by certain strains of Bacillus subtilis. Proteins that are responsible to synthesize bacilysin are encoded by bac operon. It has been shown that the biosynthesis of bacilysin is under the control of quorum sensing global regulatory pathway through the action of ComQ/ComX, PhrC (CSF), ComP/ComA in a Spo0K (Opp)-dependent manner. The objective of the study is to identify the functional roles of bacilysin biosynthesis in the regulatory cascade and idiophase cell physiology operating in B. subtilis by using gel-based and gel-free proteomics techniques. For this, we employed comparative proteome-wide analysis of the bacilysin producer B. subtilis PY79 and its bacilysin nonproducer derivative bacA::lacz::erm OGU1 strain which was recently constructed by our group. Identification via GeLC analysis of 76 differentially expressed proteins from total soluble proteome of wild-type PY79 and bacilysin minus OGU1 strain indicated the direct or indirect multiple effects of bacilysin on metabolic pathways, global regulatory systems and sporulation.