Discovering the best feature extraction and selection algorithms for spontaneous facial expression recognition

Feature extraction and selection are critical processes in developing facial expression recognition (FER) systems. While many algorithms have been proposed for these processes, direct comparison between texture, geometry and their fusion, as well as between multiple selection algorithms has not been found for spontaneous FER. This paper addresses this issue by proposing a unified framework for a comparative study on the widely used texture (LBP, Gabor and SIFT) and geometric (FAP) features, using Adaboost, mRMR and SVM feature selection algorithms. Our experiments on the Feedtum and NVIE databases demonstrate the benefits of fusing geometric and texture features, where SIFT+FAP shows the best performance, while mRMR outperforms Adaboost and SVM. In terms of computational time, LBP and Gabor perform better than SIFT. The optimal combination of SIFT+FAP+mRMR also exhibits a state-of-the-art performance.

Probabilistic matching of image sets for video-based face recognition

We address the problem of face recognition on video by employing the recently proposed probabilistic linear discrimi-nant analysis (PLDA). The PLDA has been shown to be robust against pose and expression in image-based face recognition. In this research, the method is extended and applied to video where image set to image set matching is performed. We investigate two approaches of computing similarities between image sets using the PLDA: the closest pair approach and the holistic sets approach. To better model face appearances in video, we also propose the heteroscedastic version of the PLDA which learns the within-class covariance of each individual separately. Our experi-ments on the VidTIMIT and Honda datasets show that the combination of the heteroscedastic PLDA and the closest pair approach achieves the best performance.

On the application of the probabilistic linear discriminant analysis to face recognition across expression

Facial expression is one of the main issues of face recognition in uncontrolled environments. In this paper, we apply the probabilistic linear discriminant analysis (PLDA) method to recognize faces across expressions. Several PLDA approaches are tested and cross-evaluated on the Cohn-Kanade and JAFFE databases. With less samples per gallery subject, high recognition rates comparable to previous works have been achieved indicating the robustness of the approaches. Among the approaches, the mixture of PLDAs has demonstrated better performances. The experimental results also indicate that facial regions around the cheeks, eyes, and eyebrows are more discriminative than regions around the mouth, jaw, chin, and nose.

Design for disassembly : an architectural strategy

A study of historic examples of buildings that were designed for disassembly reveals a number of important lessons in the technology employed. These lessons can inform designers such that they may better design for disassembly to attempt to increase the rates of reuse and recycling in the building industry.

Adaptive bilateral filtering using saliency map for deblocking low bit rate videos

This paper proposes a novel approach to video deblocking which performs perceptually adaptive bilateral filtering by considering color, intensity, and motion features in a holistic manner. The method is based on bilateral filter which is an effective smoothing filter that preserves edges. The bilateral filter parameters are adaptive and avoid over-blurring of texture regions and at the same time eliminate blocking artefacts in the smooth region and areas of slow motion content. This is achieved by using a saliency map to control the strength of the filter for each individual point in the image based on its perceptual importance. The experimental results demonstrate that the proposed algorithm is effective in deblocking highly compressed video sequences and to avoid over-blurring of edges and textures in salient regions of image.

Copyright enforcement in the networked society : guiding principles for protecting copyright

This paper argues that any future copyright policy should be proportional and flexible and be developed from a clear and evidence-based approach. An approach is required that carefully balances the incentives and rewards provided to economic rights holders against fundamental rights of privacy, self-expression, due process and the user rights embodied in copyright law to protect access, learning, critique, and reuse. This paper also suggests that while adequate enforcement measures are certainly part of a solution to a well functioning lawful, enforcement alone can never solve the root cause of unlawful file-sharing, since it utterly fails to address supply-side market barriers. Focus on enforcement measures alone continues to leave out a legitimate but un-served market demand, susceptible to unlawful alternatives. A competitive and consumer friendly digital content market and an appropriate legal framework to enable easy lawful access to digital content are essential preconditions for the creation of a culture of lawful, rather than unlawful, consumption.

Optimising container processes at multimodal seaport terminals : an integrated approach and application

A Multimodal Seaport Container Terminal (MSCT) is a complex system which requires careful planning and control in order to operate efficiently. It consists of a number of subsystems that require optimisation of the operations within them, as well as synchronisation of machines and containers between the various subsystems. Inefficiency in the terminal can delay ships from their scheduled timetables, as well as cause delays in delivering containers to their inland destinations, both of which can be very costly to their operators. The purpose of this PhD thesis is to use Operations Research methodologies to optimise and synchronise these subsystems as an integrated application. An initial model is developed for the overall MSCT; however, due to a large number of assumptions that had to be made, as well as other issues, it is found to be too inaccurate and infeasible for practical use. Instead, a method of developing models for each subsystem is proposed that then be integrated with each other. Mathematical models are developed for the Storage Area System (SAS) and Intra-terminal Transportation System (ITTS). The SAS deals with the movement and assignment of containers to stacks within the storage area, both when they arrive and when they are rehandled to retrieve containers below them. The ITTS deals with scheduling the movement of containers and machines between the storage areas and other sections of the terminal, such as the berth and road/rail terminals. Various constructive heuristics are explored and compared for these models to produce good initial solutions for large-sized problems, which are otherwise impractical to compute by exact methods. These initial solutions are further improved through the use of an innovative hyper-heuristic algorithm that integrates the SAS and ITTS solutions together and optimises them through meta-heuristic techniques. The method by which the two models can interact with each other as an integrated system will be discussed, as well as how this method can be extended to the other subsystems of the MSCT.

Free our cultural archives.

Australians are the creators and custodians of a broad range of cultural materials. This material includes literary, photographic, video and audio archives. These archives should be made available to all Australians for access and reuse, as part of a pre-competitive platform which promotes the interests of the Australian public in gaining access to a diverse range of content that contributes to the development of national and cultural identity. This does not mean that all material must be made available for access and reuse for free and in an unrestricted fashion. But for publicly funded content, free and unrestricted access should be the default. The Venturous Australia report on the National Innovation System recommended that “[t]o the maximum extent possible, information, research and content funded by Australian governments – including national collections – should be made freely available over the internet as part of the global public commons.”1 The report further stated that “both for its direct and indirect benefits to Australia and for the greater global good, Australia should energetically and proudly maximise the extent to which it makes government funded content available as part of the global digital commons...

Cytokine secretion via cholesterol-rich lipid raft-associated SNAREs at the phagocytic cup

Lipopolysaccharide-activated macrophages rapidly synthesize and secrete tumor necrosis factor α (TNFα) to prime the immune system. Surface delivery of membrane carrying newly synthesized TNFα is controlled and limited by the level of soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) proteins syntaxin 4 and SNAP-23. Many functions in immune cells are coordinated from lipid rafts in the plasmamembrane, and we investigated a possible role for lipid rafts in TNFα trafficking and secretion. TNFα surface delivery and secretion were found to be cholesterol- dependent. Upon macrophage activation, syntaxin 4 was recruited to cholesterol-dependent lipid rafts, whereas its regulatory protein, Munc18c, was excluded from the rafts. Syntaxin 4 in activated macrophages localized to discrete cholesterol-dependent puncta on the plasmamembrane, particularly on filopodia. Imaging the early stages of TNFα surface distribution revealed these puncta to be the initial points of TNFα delivery. During the early stages of phagocytosis, syntaxin 4 was recruited to the phagocytic cup in a cholesterol dependent manner. Insertion of VAMP3-positive recycling endosome membrane is required for efficient ingestion of a pathogen. Without this recruitment of syntaxin 4, it is not incorporated into the plasma membrane, and phagocytosis is greatly reduced. Thus, relocation of syntaxin 4 into lipid rafts in macrophages is a critical and rate-limiting step in initiating an effective immune response.

A risk simulation framework for information infrastructure protection

Information communication and technology (ICT) systems are almost ubiquitous in the modern world. It is hard to identify any industry, or for that matter any part of society, that is not in some way dependent on these systems and their continued secure operation. Therefore the security of information infrastructures, both on an organisational and societal level, is of critical importance. Information security risk assessment is an essential part of ensuring that these systems are appropriately protected and positioned to deal with a rapidly changing threat environment. The complexity of these systems and their inter-dependencies however, introduces a similar complexity to the information security risk assessment task. This complexity suggests that information security risk assessment cannot, optimally, be undertaken manually. Information security risk assessment for individual components of the information infrastructure can be aided by the use of a software tool, a type of simulation, which concentrates on modelling failure rather than normal operational simulation. Avoiding the modelling of the operational system will once again reduce the level of complexity of the assessment task. The use of such a tool provides the opportunity to reuse information in many different ways by developing a repository of relevant information to aid in both risk assessment and management and governance and compliance activities. Widespread use of such a tool allows the opportunity for the risk models developed for individual information infrastructure components to be connected in order to develop a model of information security exposures across the entire information infrastructure. In this thesis conceptual and practical aspects of risk and its underlying epistemology are analysed to produce a model suitable for application to information security risk assessment. Based on this work prototype software has been developed to explore these concepts for information security risk assessment. Initial work has been carried out to investigate the use of this software for information security compliance and governance activities. Finally, an initial concept for extending the use of this approach across an information infrastructure is presented.

Enabling USDL by tools

Fundamental tooling is required in order to apply USDL in practical settings. This chapter discusses three fundamental types of tools for USDL. First, USDL editors have been developed for expert and casual users, respectively. Second, several USDL repositories have been built to allow editors accessing and storing USDL descriptions. Third, our generic USDL marketplace allows providers to describe their services once and potentially trade them anywhere. In addition, the iosyncrasies of service trading as opposed to the simpler case of product trading. The chapter also presents several deployment scenarios of such tools to foster individual value chains and support new business models across organizational boundaries. We close the chapter with an application of USDL in the context of service engineering.

SOA approaches

As the service-oriented architecture paradigm has become ever more popular, different standardization efforts have been proposed by various consortia to enable interaction among heterongeneous environments through this paradigm. This chapter will overview the most prevalent of these SOA approaches. It will first show how technical services can be described, how they can interact with each other and be discovered by users. Next, the chapter will present different standards to facilitate service composition and to design service-oriented environments in light of a universal understanding of service orientation. The chapter will conclude with a summary and a discussion on the limitations of the reviewed standards along their ability to describe service properties. This paves the way to the next chapters where the USDL standard will be presented, which aims to lift such limitations.

Design overview of USDL

Enabling web-based service networks and ecosystems requires a way of describing services by a "commercial envelope" as discussed in Chapter 1. A uniform conception of services across all walks of life (including technical services) is required capturing business, operational and technical aspects. Therefore, our proposed Unified Service Description Language (USDL) particularly draws from and generalizes the best-of-breed approaches presented in Part I. The following chapter presents the design rationale of USDL where the different aspects are put in a framework of descriptions requirements. This is followed by the subsequent chapters of this part that provide details on specific aspects such as pricing or legal issues.

Integrity, attribution, and exploitation : contractual and normative moral rights protection in open licensing

Over the last twenty years, the use of open content licenses has become increasingly and surprisingly popular. The use of such licences challenges the traditional incentive-based model of exclusive rights under copyright. Instead of providing a means to charge for the use of particular works, what seems important is mitigating against potential personal harm to the author and, in some cases, preventing non-consensual commercial exploitation. It is interesting in this context to observe the primacy of what are essentially moral rights over the exclusionary economic rights. The core elements of common open content licences map somewhat closely to continental conceptions of the moral rights of authorship. Most obviously, almost all free software and free culture licences require attribution of authorship. More interestingly, there is a tension between social norms developed in free software communities and those that have emerged in the creative arts over integrity and commercial exploitation. For programmers interested in free software, licence terms that prohibit commercial use or modification are almost completely inconsistent with the ideological and utilitarian values that underpin the movement. For those in the creative industries, on the other hand, non-commercial terms and, to a lesser extent, terms that prohibit all but verbatim distribution continue to play an extremely important role in the sharing of copyright material. While prohibitions on commercial use often serve an economic imperative, there is also a certain personal interest for many creators in avoiding harmful exploitation of their expression – an interest that has sometimes been recognised as forming a component of the moral right of integrity. One particular continental moral right – the right of withdrawal – is present neither in Australian law or in any of the common open content licences. Despite some marked differences, both free software and free culture participants are using contractual methods to articulate the norms of permissible sharing. Legal enforcement is rare and often prohibitively expensive, and the various communities accordingly rely upon shared understandings of acceptable behaviour. The licences that are commonly used represent a formalised expression of these community norms and provide the theoretically enforceable legal baseline that lends them legitimacy. The core terms of these licences are designed primarily to alleviate risk in sharing and minimise transaction costs in sharing and using copyright expression. Importantly, however, the range of available licences reflect different optional balances in the norms of creating and sharing material. Generally, it is possible to see that, stemming particularly from the US, open content licences are fundamentally important in providing a set of normatively accepted copyright balances that reflect the interests sought to be protected through moral rights regimes. As the cost of creation, distribution, storage, and processing of expression continues to fall towards zero, there are increasing incentives to adopt open content licences to facilitate wide distribution and reuse of creative expression. Thinking of these protocols not only as reducing transaction costs but of setting normative principles of participation assists in conceptualising the role of open content licences and the continuing tensions that permeate modern copyright law.

VIVO as an Institutional Data Registry

The Queensland University of Technology (QUT) in Brisbane, Australia, is involved in a number of projects funded by the Australian National Data Service (ANDS). Currently, QUT is working on a project (Metadata Stores Project) that uses open source VIVO software to aid in the storage and management of metadata relating to data sets created/managed by the QUT research community. The registry (called QUT Research Data Finder) will support the sharing and reuse of research datasets, within and external to QUT. QUT uses VIVO for both the display and the editing of research metadata.