Feature-domain super-resolution framework for gabor-based face and iris recognition

The low resolution of images has been one of the major limitations in recognising humans from a distance using their biometric traits, such as face and iris. Superresolution has been employed to improve the resolution and the recognition performance simultaneously, however the majority of techniques employed operate in the pixel domain, such that the biometric feature vectors are extracted from a super-resolved input image. Feature-domain superresolution has been proposed for face and iris, and is shown to further improve recognition performance by capitalising on direct super-resolving the features which are used for recognition. However, current feature-domain superresolution approaches are limited to simple linear features such as Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA), which are not the most discriminant features for biometrics. Gabor-based features have been shown to be one of the most discriminant features for biometrics including face and iris. This paper proposes a framework to conduct super-resolution in the non-linear Gabor feature domain to further improve the recognition performance of biometric systems. Experiments have confirmed the validity of the proposed approach, demonstrating superior performance to existing linear approaches for both face and iris biometrics.

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.

The taxation of multinational banks : alternative apportionment through a unitary taxation regime aligning with economic reality

The taxation of multinational banks currently is governed by the general principles of international tax. However, it is arguable that there are characteristics exclusive to multinational banks that may warrant the consideration of a separate taxing regime. This article argues that because of the unique nature of multinational banks, the traditional international tax rules governing jurisdiction to tax and allocation of income do not produce a result which is optimal, as it does not reflect economic reality. That is, the current system does not produce a result that accurately reflects the economic source of the income or the location of the economic activity. The suggested alternative is unitary taxation using global formulary apportionment. Formulary apportionment is considered as an alternative that reflects economic reality by recognising the unique nature of multinational banks and allocating the income to the location of the economic activity. The unique nature of multinational banking is recognised in the fact that formulary apportionment does not attempt to undertake a transactional division of a highly integrated multinational entity. Rather, it allocates income to the jurisdictions based on an economically justifiable formula. Starting from this recognition, the purpose of this article is to demonstrate that formulary apportionment is a theoretically superior (or optimal) model for the taxation of multinational banks. An optimal regime, for the purposes of this article, is considered to be one that distributes the taxing rights in an equitable manner between the relevant jurisdictions, while, simultaneously allowing decisions of the international banks to be tax neutral. In this sense, neutrality is viewed as an economic concept and equity is regarded as a legal concept. A neutral tax system is one in which tax rules do not affect economic choices about commercial activities. Neutrality will ideally be across jurisdictions as well as across traditional and non-traditional industries. The primary focus of this article is jurisdictional neutrality. A system that distributes taxing rights in an equitable manner between the relevant jurisdictions ensures that each country receives its fair share of tax revenue. Given the increase in multinational banking, jurisdictions should be concerned that they are receiving their fair share. Inter-nation equity is concerned with re-determining the proper division of the tax base among countries. Richard and Peggy Musgrave argue that sharing of the tax base by countries of source should be seen as a matter of inter-nation equity requiring international cooperation. The rights of the jurisdiction of residency will also be at issue. To this extent, while it is agreed that inter-nation equity is an essential attribute to an international tax regime, there is no universal agreement as to how to achieve it. The current system attempts to achieve such equity through a combined residency and source regime, with the transfer pricing rules used to apportion income between the relevant jurisdictions. However, this article suggests, that as an alternative to the current regime, equity would be achieved through formulary apportionment. Opposition to formulary apportionment is generally based on the argument that it is not a theoretically superior (or optimal) model because of the implementation difficulties. Yet these are two separate issues. As such, this article is divided into two core parts. The first part examines the theoretical soundness of the formulary apportionment model concluding that it is theoretically superior to the arm’s length pricing requirement of the traditional transfer pricing regime. The second part examines the practical implications of accepting formulary apportionment as an optimal model with a view to disclosing the issues that arise when a formulary apportionment regime is adopted. Prior to an analysis of the theoretical and practical application of formulary apportionment to multinational banks, the unique nature of these banks is considered. The article concludes that, while there are significant implementation, compliance, and enforcement issues to overcome, the unitary taxation model may be theoretically superior to the current arm’s length model which applies to multinational banks. This conclusion is based on the unitary taxation model providing greater alignment with the unique features of these banks.

Improved facial expression recognition via uni-hyperplane classification

Large margin learning approaches, such as support vector machines (SVM), have been successfully applied to numerous classification tasks, especially for automatic facial expression recognition. The risk of such approaches however, is their sensitivity to large margin losses due to the influence from noisy training examples and outliers which is a common problem in the area of affective computing (i.e., manual coding at the frame level is tedious so coarse labels are normally assigned). In this paper, we leverage the relaxation of the parallel-hyperplanes constraint and propose the use of modified correlation filters (MCF). The MCF is similar in spirit to SVMs and correlation filters, but with the key difference of optimizing only a single hyperplane. We demonstrate the superiority of MCF over current techniques on a battery of experiments.

Evaluation of image resolution and super-resolution on face recognition performance

While researchers strive to improve automatic face recognition performance, the relationship between image resolution and face recognition performance has not received much attention. This relationship is examined systematically and a framework is developed such that results from super-resolution techniques can be compared. Three super-resolution techniques are compared with the Eigenface and Elastic Bunch Graph Matching face recognition engines. Parameter ranges over which these techniques provide better recognition performance than interpolated images is determined.

Perceiving and pursuing novelty : a grounded theory of adolescent creativity

Creativity plays an increasingly important role in our personal, social, educational, and community lives. For adolescents, creativity can enable self-expression, be a means of pushing boundaries, and assist learning, achievement, and completion of everyday tasks. Moreover, adolescents who demonstrate creativity can potentially enhance their capacity to face unknown future challenges, address mounting social and ecological issues in our global society, and improve their career opportunities and contribution to the economy. For these reasons, creativity is an essential capacity for young people in their present and future, and is highlighted as a priority in current educational policy nationally and internationally. Despite growing recognition of creativity’s importance and attention to creativity in research, the creative experience from the perspectives of the creators themselves and the creativity of adolescents are neglected fields of study. Hence, this research investigated adolescents’ self-reported experiences of creativity to improve understandings of their creative processes and manifestations, and how these can be supported or inhibited. Although some aspects of creativity have been extensively researched, there were no comprehensive, multidisciplinary theoretical frameworks of adolescent creativity to provide a foundation for this study. Therefore, a grounded theory methodology was adopted for the purpose of constructing a new theory to describe and explain adolescents’ creativity in a range of domains. The study’s constructivist-interpretivist perspective viewed the data and findings as interpretations of adolescents’ creative experiences, co-constructed by the participants and the researcher. The research was conducted in two academically selective high schools in Australia: one arts school, and one science, mathematics, and technology school. Twenty adolescent participants (10 from each school) were selected using theoretical sampling. Data were collected via focus groups, individual interviews, an online discussion forum, and email communications. Grounded theory methods informed a process of concurrent data collection and analysis; each iteration of analysis informed subsequent data collection. Findings portray creativity as it was perceived and experienced by participants, presented in a Grounded Theory of Adolescent Creativity. The Grounded Theory of Adolescent Creativity comprises a core category, Perceiving and Pursuing Novelty: Not the Norm, which linked all findings in the study. This core category explains how creativity involved adolescents perceiving stimuli and experiences differently, approaching tasks or life unconventionally, and pursuing novel ideas to create outcomes that are not the norm when compared with outcomes by peers. Elaboration of the core category is provided by the major categories of findings. That is, adolescent creativity entailed utilising a network of Sub-Processes of Creativity, using strategies for Managing Constraints and Challenges, and drawing on different Approaches to Creativity – adaptation, transfer, synthesis, and genesis – to apply the sub-processes and produce creative outcomes. Potentially, there were Effects of Creativity on Creators and Audiences, depending on the adolescent and the task. Three Types of Creativity were identified as the manifestations of the creative process: creative personal expression, creative boundary pushing, and creative task achievement. Interactions among adolescents’ dispositions and environments were influential in their creativity. Patterns and variations of these interactions revealed a framework of four Contexts for Creativity that offered different levels of support for creativity: high creative disposition–supportive environment; high creative disposition–inhibiting environment; low creative disposition–supportive environment; and low creative disposition–inhibiting environment. These contexts represent dimensional ranges of how dispositions and environments supported or inhibited creativity, and reveal that the optimal context for creativity differed depending on the adolescent, task, domain, and environment. This study makes four main contributions, which have methodological and theoretical implications for researchers, as well as practical implications for adolescents, parents, teachers, policy and curriculum developers, and other interested stakeholders who aim to foster the creativity of adolescents. First, this study contributes methodologically through its constructivist-interpretivist grounded theory methodology combining the grounded theory approaches of Corbin and Strauss (2008) and Charmaz (2006). Innovative data collection was also demonstrated through integration of data from online and face-to-face interactions with adolescents, within the grounded theory design. These methodological contributions have broad applicability to researchers examining complex constructs and processes, and with populations who integrate multimedia as a natural form of communication. Second, applicable to creativity in diverse domains, the Grounded Theory of Adolescent Creativity supports a hybrid view of creativity as both domain-general and domain-specific. A third major contribution was identification of a new form of creativity, educational creativity (ed-c), which categorises creativity for learning or achievement within the constraints of formal educational contexts. These theoretical contributions inform further research about creativity in different domains or multidisciplinary areas, and with populations engaged in formal education. However, the key contribution of this research is that it presents an original Theory and Model of Adolescent Creativity to explain the complex, multifaceted phenomenon of adolescents’ creative experiences.

Field-of-view, detection range, and false alarm trade-offs in vision-based aircraft detection

This paper presents a survey of previously presented vision based aircraft detection flight test, and then presents new flight test results examining the impact of camera field-of view choice on the detection range and false alarm rate characteristics of a vision-based aircraft detection technique. Using data collected from approaching aircraft, we examine the impact of camera fieldof-view choice and confirm that, when aiming for similar levels of detection confidence, an improvement in detection range can be obtained by choosing a smaller effective field-of-view (in terms of degrees per pixel).

PLDA based speaker recognition on short utterances

This paper investigates the effects of limited speech data in the context of speaker verification using a probabilistic linear discriminant analysis (PLDA) approach. Being able to reduce the length of required speech data is important to the development of automatic speaker verification system in real world applications. When sufficient speech is available, previous research has shown that heavy-tailed PLDA (HTPLDA) modeling of speakers in the i-vector space provides state-of-the-art performance, however, the robustness of HTPLDA to the limited speech resources in development, enrolment and verification is an important issue that has not yet been investigated. In this paper, we analyze the speaker verification performance with regards to the duration of utterances used for both speaker evaluation (enrolment and verification) and score normalization and PLDA modeling during development. Two different approaches to total-variability representation are analyzed within the PLDA approach to show improved performance in short-utterance mismatched evaluation conditions and conditions for which insufficient speech resources are available for adequate system development. The results presented within this paper using the NIST 2008 Speaker Recognition Evaluation dataset suggest that the HTPLDA system can continue to achieve better performance than Gaussian PLDA (GPLDA) as evaluation utterance lengths are decreased. We also highlight the importance of matching durations for score normalization and PLDA modeling to the expected evaluation conditions. Finally, we found that a pooled total-variability approach to PLDA modeling can achieve better performance than the traditional concatenated total-variability approach for short utterances in mismatched evaluation conditions and conditions for which insufficient speech resources are available for adequate system development.

A view from the edge of the forest : recent progress in understanding the relationships of the insect orders

Research over the last two decades has significantly increased our understanding of the evolutionary position of the insects among other arthropods, and the relationships among the insect Orders. Many of these insights have been established through increasingly sophisticated analyses of DNA sequence data from a limited number of genes. Recent results have established the relationships of the Holometabola, but relationships among the hemimetabolous orders have been more difficult to elucidate. A strong consensus on the relationships among the Palaeoptera (Ephemeroptera and Odonata) and their relationship to the Neoptera has not emerged with all three possible resolutions supported by different data sets. While polyneopteran relationships generally have resisted significant resolution, it is now clear that termites, Isoptera, are nested within the cockroaches, Blattodea. The newly discovered order Mantophasmatodea is difficult to place with the balance of studies favouring Grylloblattodea as sister-group. While some studies have found the paraneopteran orders (Hemiptera, Thysanoptera, Phthiraptera and Psocoptera) monophyletic, evidence suggests that parasitic lice (Phthiraptera) have evolved from groups within the book and bark lice (Psocoptera), and may represent parallel evolutions of parasitism within two major louse groups. Within Holometabola, it is now clear that Hymenoptera are the sister to the other orders, that, in turn are divided into two clades, the Neuropteroidea (Coleoptera, Neuroptera and relatives) and the Mecopterida (Trichoptera, Lepidoptera, Diptera and their relatives). The enigmatic order Strepsiptera, the twisted wing insects, have now been placed firmly near Coleoptera, rejecting their close relationship to Diptera that was proposed some 15years ago primarily based on ribosomal DNA data. Phylogenomic-scale analyses are just beginning to be focused on the relationships of the insect orders, and this is where we expect to see resolution of palaeopteran and polyneopteran relationships. Future research will benefit from greater coordination between intra and inter-ordinal analyses. This will maximise the opportunities for appropriate outgroup choice at the intraordinal level and provide the background knowledge for the interordinal analyses to span the maximum phylogenetic scope within groups.

Text mining in negative relevance feedback

It is a big challenge to clearly identify the boundary between positive and negative streams. Several attempts have used negative feedback to solve this challenge; however, there are two issues for using negative relevance feedback to improve the effectiveness of information filtering. The first one is how to select constructive negative samples in order to reduce the space of negative documents. The second issue is how to decide noisy extracted features that should be updated based on the selected negative samples. This paper proposes a pattern mining based approach to select some offenders from the negative documents, where an offender can be used to reduce the side effects of noisy features. It also classifies extracted features (i.e., terms) into three categories: positive specific terms, general terms, and negative specific terms. In this way, multiple revising strategies can be used to update extracted features. An iterative learning algorithm is also proposed to implement this approach on RCV1, and substantial experiments show that the proposed approach achieves encouraging performance.