Summarisation of short-term and long-term videos using texture and colour

We present a novel approach to video summarisation that makes use of a Bag-of-visual-Textures (BoT) approach. Two systems are proposed, one based solely on the BoT approach and another which exploits both colour information and BoT features. On 50 short-term videos from the Open Video Project we show that our BoT and fusion systems both achieve state-of-the-art performance, obtaining an average F-measure of 0.83 and 0.86 respectively, a relative improvement of 9% and 13% when compared to the previous state-of-the-art. When applied to a new underwater surveillance dataset containing 33 long-term videos, the proposed system reduces the amount of footage by a factor of 27, with only minor degradation in the information content. This order of magnitude reduction in video data represents significant savings in terms of time and potential labour cost when manually reviewing such footage.

Matching image sets via adaptive multi convex hull

Traditional nearest points methods use all the samples in an image set to construct a single convex or affine hull model for classification. However, strong artificial features and noisy data may be generated from combinations of training samples when significant intra-class variations and/or noise occur in the image set. Existing multi-model approaches extract local models by clustering each image set individually only once, with fixed clusters used for matching with various image sets. This may not be optimal for discrimination, as undesirable environmental conditions (eg. illumination and pose variations) may result in the two closest clusters representing different characteristics of an object (eg. frontal face being compared to non-frontal face). To address the above problem, we propose a novel approach to enhance nearest points based methods by integrating affine/convex hull classification with an adapted multi-model approach. We first extract multiple local convex hulls from a query image set via maximum margin clustering to diminish the artificial variations and constrain the noise in local convex hulls. We then propose adaptive reference clustering (ARC) to constrain the clustering of each gallery image set by forcing the clusters to have resemblance to the clusters in the query image set. By applying ARC, noisy clusters in the query set can be discarded. Experiments on Honda, MoBo and ETH-80 datasets show that the proposed method outperforms single model approaches and other recent techniques, such as Sparse Approximated Nearest Points, Mutual Subspace Method and Manifold Discriminant Analysis.

Automatic classification of human epithelial type 2 cell indirect immunofluorescence images using cell pyramid matching

This paper describes a novel system for automatic classification of images obtained from Anti-Nuclear Antibody (ANA) pathology tests on Human Epithelial type 2 (HEp-2) cells using the Indirect Immunofluorescence (IIF) protocol. The IIF protocol on HEp-2 cells has been the hallmark method to identify the presence of ANAs, due to its high sensitivity and the large range of antigens that can be detected. However, it suffers from numerous shortcomings, such as being subjective as well as time and labour intensive. Computer Aided Diagnostic (CAD) systems have been developed to address these problems, which automatically classify a HEp-2 cell image into one of its known patterns (eg. speckled, homogeneous). Most of the existing CAD systems use handpicked features to represent a HEp-2 cell image, which may only work in limited scenarios. We propose a novel automatic cell image classification method termed Cell Pyramid Matching (CPM), which is comprised of regional histograms of visual words coupled with the Multiple Kernel Learning framework. We present a study of several variations of generating histograms and show the efficacy of the system on two publicly available datasets: the ICPR HEp-2 cell classification contest dataset and the SNPHEp-2 dataset.

Multi-shot person re-identification via relational stein divergence

Person re-identification is particularly challenging due to significant appearance changes across separate camera views. In order to re-identify people, a representative human signature should effectively handle differences in illumination, pose and camera parameters. While general appearance-based methods are modelled in Euclidean spaces, it has been argued that some applications in image and video analysis are better modelled via non-Euclidean manifold geometry. To this end, recent approaches represent images as covariance matrices, and interpret such matrices as points on Riemannian manifolds. As direct classification on such manifolds can be difficult, in this paper we propose to represent each manifold point as a vector of similarities to class representers, via a recently introduced form of Bregman matrix divergence known as the Stein divergence. This is followed by using a discriminative mapping of similarity vectors for final classification. The use of similarity vectors is in contrast to the traditional approach of embedding manifolds into tangent spaces, which can suffer from representing the manifold structure inaccurately. Comparative evaluations on benchmark ETHZ and iLIDS datasets for the person re-identification task show that the proposed approach obtains better performance than recent techniques such as Histogram Plus Epitome, Partial Least Squares, and Symmetry-Driven Accumulation of Local Features.

Improved image set classification via joint sparse approximated nearest subspaces

Existing multi-model approaches for image set classification extract local models by clustering each image set individually only once, with fixed clusters used for matching with other image sets. However, this may result in the two closest clusters to represent different characteristics of an object, due to different undesirable environmental conditions (such as variations in illumination and pose). To address this problem, we propose to constrain the clustering of each query image set by forcing the clusters to have resemblance to the clusters in the gallery image sets. We first define a Frobenius norm distance between subspaces over Grassmann manifolds based on reconstruction error. We then extract local linear subspaces from a gallery image set via sparse representation. For each local linear subspace, we adaptively construct the corresponding closest subspace from the samples of a probe image set by joint sparse representation. We show that by minimising the sparse representation reconstruction error, we approach the nearest point on a Grassmann manifold. Experiments on Honda, ETH-80 and Cambridge-Gesture datasets show that the proposed method consistently outperforms several other recent techniques, such as Affine Hull based Image Set Distance (AHISD), Sparse Approximated Nearest Points (SANP) and Manifold Discriminant Analysis (MDA).

Population variation in biomonitoring data for persistent organic pollutants (POPs) : an examination of multiple population-based datasets for application to Australian pooled biomonitoring data

Background Australian national biomonitoring for persistent organic pollutants (POPs) relies upon age-specific pooled serum samples to characterize central tendencies of concentrations but does not provide estimates of upper bound concentrations. This analysis compares population variation from biomonitoring datasets from the US, Canada, Germany, Spain, and Belgium to identify and test patterns potentially useful for estimating population upper bound reference values for the Australian population. Methods Arithmetic means and the ratio of the 95th percentile to the arithmetic mean (P95:mean) were assessed by survey for defined age subgroups for three polychlorinated biphenyls (PCBs 138, 153, and 180), hexachlorobenzene (HCB), p,p-dichlorodiphenyldichloroethylene (DDE), 2,2′,4,4′ tetrabrominated diphenylether (PBDE 47), perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS). Results Arithmetic mean concentrations of each analyte varied widely across surveys and age groups. However, P95:mean ratios differed to a limited extent, with no systematic variation across ages. The average P95:mean ratios were 2.2 for the three PCBs and HCB; 3.0 for DDE; 2.0 and 2.3 for PFOA and PFOS, respectively. The P95:mean ratio for PBDE 47 was more variable among age groups, ranging from 2.7 to 4.8. The average P95:mean ratios accurately estimated age group-specific P95s in the Flemish Environmental Health Survey II and were used to estimate the P95s for the Australian population by age group from the pooled biomonitoring data. Conclusions Similar population variation patterns for POPs were observed across multiple surveys, even when absolute concentrations differed widely. These patterns can be used to estimate population upper bounds when only pooled sampling data are available.

Multi-action recognition via stochastic modelling of optical flow and gradients

In this paper we propose a novel approach to multi-action recognition that performs joint segmentation and classification. This approach models each action using a Gaussian mixture using robust low-dimensional action features. Segmentation is achieved by performing classification on overlapping temporal windows, which are then merged to produce the final result. This approach is considerably less complicated than previous methods which use dynamic programming or computationally expensive hidden Markov models (HMMs). Initial experiments on a stitched version of the KTH dataset show that the proposed approach achieves an accuracy of 78.3%, outperforming a recent HMM-based approach which obtained 71.2%.

On robust face recognition via sparse coding: The good, the bad and the ugly

In the field of face recognition, sparse representation (SR) has received considerable attention during the past few years, with a focus on holistic descriptors in closed-set identification applications. The underlying assumption in such SR-based methods is that each class in the gallery has sufficient samples and the query lies on the subspace spanned by the gallery of the same class. Unfortunately, such an assumption is easily violated in the face verification scenario, where the task is to determine if two faces (where one or both have not been seen before) belong to the same person. In this study, the authors propose an alternative approach to SR-based face verification, where SR encoding is performed on local image patches rather than the entire face. The obtained sparse signals are pooled via averaging to form multiple region descriptors, which then form an overall face descriptor. Owing to the deliberate loss of spatial relations within each region (caused by averaging), the resulting descriptor is robust to misalignment and various image deformations. Within the proposed framework, they evaluate several SR encoding techniques: l1-minimisation, Sparse Autoencoder Neural Network (SANN) and an implicit probabilistic technique based on Gaussian mixture models. Thorough experiments on AR, FERET, exYaleB, BANCA and ChokePoint datasets show that the local SR approach obtains considerably better and more robust performance than several previous state-of-the-art holistic SR methods, on both the traditional closed-set identification task and the more applicable face verification task. The experiments also show that l1-minimisation-based encoding has a considerably higher computational cost when compared with SANN-based and probabilistic encoding, but leads to higher recognition rates.

Genome-wide association study of CNVs in 16,000 cases of eight common diseases and 3,000 shared controls

Copy number variants (CNVs) account for a major proportion of human genetic polymorphism and have been predicted to have an important role in genetic susceptibility to common disease. To address this we undertook a large, direct genome-wide study of association between CNVs and eight common human diseases. Using a purpose-designed array we typed 19,000 individuals into distinct copy-number classes at 3,432 polymorphic CNVs, including an estimated 50% of all common CNVs larger than 500 base pairs. We identified several biological artefacts that lead to false-positive associations, including systematic CNV differences between DNAs derived from blood and cell lines. Association testing and follow-up replication analyses confirmed three loci where CNVs were associated with diseaseIRGM for Crohns disease, HLA for Crohns disease, rheumatoid arthritis and type 1 diabetes, and TSPAN8 for type 2 diabetesalthough in each case the locus had previously been identified in single nucleotide polymorphism (SNP)-based studies, reflecting our observation that most common CNVs that are well-typed on our array are well tagged by SNPs and so have been indirectly explored through SNP studies. We conclude that common CNVs that can be typed on existing platforms are unlikely to contribute greatly to the genetic basis of common human diseases. © 2010 Macmillan Publishers Limited. All rights reserved.

Are we there yet? Advancing women in Canada and Australia: Similar goals, different policies

Differences in opportunities and outcomes in the workplace are inherent in a free and competitive market. However when differences between individuals and groups are identified as resulting from particular policies, behaviours or attitudes, any resulting inequality may be identified as unfair. Increasingly, unfair disparities in societies and their workplaces are regularly challenged. Many of the unfair disparities are recognised as caused through unfair discrimination (Anker 1997). When defining discrimination, the International Labour Organization Convention (ILO) No. 111 defines it as “any distinction, exclusion or preference made on the basis of race, colour, sex, religion, political opinion, national extraction, or social origin, which has the effect of nullifying or impairing equality of opportunity or treatment in employment or occupation” (ILO, 1958). Yet, the argument for addressing this ideal of ‘equality of opportunity’ is complex. Ekmekci (2013) identifies the difficulties as the determination of whether any process should be based on equality of opportunity or equality of outcome. In addition, there is the difficulty of determining what exactly constitutes a process for addressing unfair disparity due to the haziness of what constitutes discrimination and controversy in the meaning as well as policy implications of equality (Tomei, 2003).

Optimisation techniques for planning automatic under frequency load shedding in New Zealand’s power system

Electricity generation is vital in developed countries to power the many mechanical and electrical devices that people require. Unfortunately electricity generation is costly. Though electricity can be generated it cannot be stored efficiently. Electricity generation is also difficult to manage because exact demand is unknown from one instant to the next. A number of services are required to manage fluctuations in electricity demand, and to protect the system when frequency falls too low. A current approach is called automatic under frequency load shedding (AUFLS). This article proposes new methods for optimising AUFLS in New Zealand’s power system. The core ideas were developed during the 2015 Maths and Industry Study Group (MISG) in Brisbane, Australia. The problem has been motivated by Transpower Limited, a company that manages New Zealand’s power system and transports bulk electricity from where it is generated to where it is needed. The approaches developed in this article can be used in electrical power systems anywhere in the world.