Accurate and efficient face recognition from video

As a problem of high practical appeal but outstanding challenges, computer-based face recognition remains a topic of extensive research attention. In this paper we are specifically interested in the task of identifying a person from multiple training and query images. Thus, a novel method is proposed which advances the state-of-the-art in set based face recognition. Our method is based on a previously described invariant in the form of generic shape-illumination effects. The contributions include: (i) an analysis of computational demands of the original method and a demonstration of its practical limitations, (ii) a novel representation of personal appearance in the form of linked mixture models in image and pose-signature spaces, and (iii) an efficient (in terms of storage needs and matching time) manifold re-illumination algorithm based on the aforementioned representation. An evaluation and comparison of the proposed method with the original generic shape-illumination algorithm shows that comparably high recognition rates are achieved on a large data set (1.5% error on 700 face sets containing 100 individuals and extreme illumination variation) with a dramatic improvement in matching speed (over 700 times for sets containing 1600 faces) and storage requirements (independent of the number of training images).

A manifold approach to face recognition from low quality video across illumination and pose using implicit super-resolution

We consider the problem of matching a face in a low resolution query video sequence against a set of higher quality gallery sequences. This problem is of interest in many applications, such as law enforcement. Our main contribution is an extension of the recently proposed Generic Shape-Illumination Manifold (gSIM) framework. Specifically, (i) we show how super-resolution across pose and scale can be achieved implicitly, by off-line learning of subsampling artefacts; (ii) we use this result to propose an extension to the statistical model of the gSIM by compounding it with a hierarchy of subsampling models at multiple scales; and (iii) we describe an extensive empirical evaluation of the method on over 1300 video sequences – we first measure the degradation in performance of the original gSIM algorithm as query sequence resolution is decreased and then show that the proposed extension produces an error reduction in the mean recognition rate of over 50%.

Bend me, shape me, anyway you want me: constructing teacher identities across different learning paradigms

This article discusses one key finding from a qualitative study that investigated the experiences of overseas-born, ethnic minority early childhood pre-service teachers in New Zealand. Data were collected through interviews with recently graduated Bachelor of Teaching (Early Childhood Education) teachers and early childhood lecturers. Until they were supported to incorporate their cultural knowledges into their new learning, most graduate participants found there was little to which they could relate in the pedagogies and content of their teacher education courses. This article makes recommendations for the planning, preparation and delivery of early childhood teacher education.

The citrate-mediated shape evolution of transforming photomorphic silver nanoparticles

The photoconversion of photomorphic silver nanoparticles from discs to prisms via citrate mediated growth on the twin plane faces of the nanoparticles is demonstrated. This systematic shape evolution from discs to hexagons and then prisms of increasing aspect ratios is a result of the growth process being confined to specific faces of the growing nanoparticles.

Generation of total runoff hydrographs using a method derived from a digital filter algorithm

A reliable prediction of the total runoff hydrograph is necessary for water resources management. This study investigates two approaches to generate total runoff hydrograph by adding baseflow to direct runoff hydrographs. The first approach uses a method, derived from a digital filter algorithm for hydrograph separation, to generate baseflow hydrographs from direct runoff hydrographs. The method appears to perform well in producing the overall shape of the total runoff hydrographs and the acceptable mass balance errors for a year of water cycle. For application, the recession baseflow constant needs to be estimated reliably and the initial baseflow could be approximated to the long-term mean dry weather flow. The second approach assumes a constant baseflow rate. Although this approach is still capable of producing the overall hydrograph shape, it yields high mass balance errors in the total runoff hydrographs for both monthly and long-term periods. Further analysis shows that two-third of the mass balance errors are contributed from periods with direct runoff, implying that the constant baseflow assumption could introduce significant errors into the computations of total runoff hydrograph

Letter from the chair, Bulletin : ASME Victoria

Histogram shape-based robust image watermarking method

Developing a watermarking method that is robust to cropping attack and random bending attacks (RBAs) is a challenging task in image watermarking. In this paper, we propose a histogram-based image watermarking method to tackle with both cropping attack and RBAs. In this method first the gray levels are divided into groups. Secondly the groups for watermark embedding are selected according to the number of pixels in them, which makes this method fully based on the histogram shape of the original image and adaptive to different images. Then the watermark bits are embedded by modifying the histogram of the selected groups. Since histogram shape is insensitive to cropping and independent from pixel positions, the proposed method is robust to cropping attack and RBAs. Besides, it also has high robustness against other common attacks. Experimental results demonstrate the effectiveness of the proposed method. © 2014 IEEE.

Photoinduced reversible shape conversion of silver nanoparticles assisted by TiO₂.

Silver nanoprisms were transformed into nanodecahedra through photoinduction of ultraviolet (UV) light in the presence of titanium dioxide (TiO2) quantum dots (QDs). Subsequently, the silver nanodecahedra were reconverted to silver nanoprisms under sodium lamp if there was sufficient citrate in the reaction system. The localized surface plasmon resonance (LSPR) optical properties of silver nanoparticles were tuned during photoinduced shape conversion. The photocatalytic activity of TiO2 QDs assisted the conversion of prisms to decahedra upon UV light irradiation. Nevertheless, the presence of TiO2 did not inhibit the photoinduced reconversion from decahedra to prisms by sodium light. It was demonstrated that citrate was indispensable in the photoinduction process. In addition, oxygen in solution played a vital role in the reversible shape conversion of silver nanoparticles. Moreover, simulated sunlight can convert silver nanoprisms to nanodecahedra instead of UV light with assistance of TiO2 QDs, which would promote the photoinduced reaction of silver nanoparticles based on a natural light source.

Microstructure and mechanical properties of silk from different components of the Antheraea pernyi cocoon

Silk fibres from different components of the Antheraea pernyi silkworm cocoon, namely peduncle, outer floss, and cocoon shells (outermost layer and pelade layer) were studied in detail to gain insights into the structure-property-function relationship. Among the fibres from different components, peduncle fibres are the softest with the largest viscoelastic lag, which may reduce the oscillation amplitude when a cocoon hangs on a twig. Fibres from the outermost layer are the toughest and have the largest breaking energy. Outer floss fibres have the highest content of sericin (about 11.98%) but their hardness and elasticity are intermediate. Pelade fibres are shape - preservable and stable with superior hardness and elasticity. The understanding of the properties of different silk fibres is essential for understanding their respective roles in the function of a silk cocoon and will also inspire new designs of protective materials under stringent environmental conditions.

A first step towards a simple in-line shape compensation routine for the roll forming of high strength steel

The roll forming process is increasingly used in the automotive industry for the manufacture of structural and crash components from Ultra High Strength Steel (UHSS). Due to the high strength of UHSS (<1GPa) even small and commonly observed material property variations from coil to coil can result in significant changes in material yield and through that affect the final shape of the roll formed component. This requires the re-adjustment of tooling to compensate for shape defects and maintain part geometry resulting in costly downtimes of equipment. This paper presents a first step towards an in-line shape compensation method that based on the monitoring of roll load and torque allows for the estimation of shape defects and the subsequent re-adjustment of tooling for compensation. For this the effect of material property variation on common shape defects observed in the roll forming process as well as measurable process parameters such as roll load and torque needs to be understood. The effect of yield strength and material hardening on roll load and torque as well as longitudinal bow is investigated via experimental trials and numerical analysis. A regression analysis combined with Analysis of Variance (ANOVA) techniques is employed to establish the relationships between the process and material parameters and to determine their percentage influence on longitudinal bow, roll load and torque. The study will show that the level of longitudinal bow, one of the major shape defects observed in roll forming, can be estimated by variations in roll load and torque.

Nature of crystalline particle assembly in ring shaped colloidal stains from concentrated dispersions

The drying of colloidal droplet suspensions is important in many realms of practical application and has sustained the interest of researchers over two decades. The arrangements of polystyrene and silica beads, both of diameter 1 μm, 10% by volume of solid deposited on normal glass (hydrophilic), and silicone (hydrophobic) surfaces evaporated from a suspension volume of 3 μL, were investigated. Doughnut shape depositions were found, imputing the influence of strong central circulation flows that resulted in three general regions. In the central region which had strong particle build-up, the top most layers of particle arrangement was confirmed to be disordered using power spectrum and radial distribution function analysis. On closer examination, this appeared more like frustrated attempts to crystallize into larger grains rather than beads arranging in a disordered fashion throughout the piling process. With an adapted micro-bulldozing operation to progressively remove layers of particles from the heap, we found that the later efforts to crystallize through lateral capillary inter-particle forces were liable to be undone once the particles contacted the disorganized particles underneath, which were formed out of the jamming of fast particles arriving at the surface. © 2014 Elsevier B.V.

Robust histogram shape-based method for image watermarking

Cropping and random bending are two common attacks in image watermarking. In this paper we propose a novel image-watermarking method to deal with these attacks, as well as other common attacks. In the embedding process, we first preprocess the host image by a Gaussian low-pass filter. Then, a secret key is used to randomly select a number of gray levels and the histogram of the filtered image with respect to these selected gray levels is constructed. After that, a histogram-shape-related index is introduced to choose the pixel groups with the highest number of pixels and a safe band is built between the chosen and nonchosen pixel groups. A watermark-embedding scheme is proposed to insert watermarks into the chosen pixel groups. The usage of the histogram-shape-related index and safe band results in good robustness. Moreover, a novel high-frequency component modification mechanism is also utilized in the embedding scheme to further improve robustness. At the decoding end, based on the available secret key, the watermarked pixel groups are identified and watermarks are extracted from them. The effectiveness of the proposed image-watermarking method is demonstrated by simulation examples.

UV protection performance of textiles affected by fiber cross-sectional shape

Abstract
In this study, a model was set up for simulating the effects of varying fiber cross-sectional shapes on ultraviolet protection of fibers. The fiber diameter and fiber type were also involved in the model setting. Experiments of diffuse reflectance spectra measurement on natural (wool, cotton, silk), regenerated (bamboo viscose) and synthetic (polyester, nylon) fibers were conducted to verify the model predicted results. When a more complex shape was assumed as the fiber cross-section for model calculation, the predicted results have a better agreement with the actual results. The effects on ultraviolet absorption from fibers with different cross-sectional shapes were investigated at a single fiber, fiber bundle and yarn levels. With the same material, when the fiber cross-sectional area and the areal coverage of a single fiber were constant, the triangular shape had the lowest ultraviolet transmittance and the highest ultraviolet reflectance for a single fiber and also for a fiber bundle. The difference of fiber cross-sectional shapes was also significant in the ultraviolet protection of a single yarn.

Managerial ownership, capital structure and firm value: Evidence from China’s civilian-run firms

This paper examines the influence of managerial ownership on firm performance through capital-structure choices, using a sample of China’s civilian-run firms listed on the Chinese stock market between 2002 and 2007. The empirical results demonstrate a nonlinear relationship between managerial ownership and firm value. Managerial ownership drives the capital structure into a nonlinear shape, but in an opposite direction to the effect of managerial ownership on firm value. The results of simultaneous regressions suggest that managerial ownership affects capital structure, which in turn affects firm value. Our findings imply that the “interest convergence” and “entrenchment” effects of managers’ behaviour in terms of managerial ownership can also explain the agency-relevant situation of China’s civilian-run firms.

Legal liability, government intervention, and auditor behavior : evidence from structural reform of audit firms in China

This paper investigates how legal liability influences audit quality and audit fees, particularly in the presence of government intervention. Since 2010, all Chinese audit firms were required to transform from a structure of limited liability company (LLC) to limited liability partnership (LLP), which removes the cap on the liability exposure of negligent auditors. By adopting this natural experiment, we document the following findings: first, after audit firms reorganize as LLPs, auditors are more likely to (1) issue modified audit opinions and going-concern opinions, (2) constrain clients’ earnings management, and (3) charge a premium in audit fees, which suggest that exerting unlimited legal liability on negligent auditors improves both audit quality and audit fees. Second, the effect of the LLP adoption is more pronounced when auditors are from local audit firms, and clients are controlled by local governments. Further analyses suggest that the stock prices of clients positively react to the reform event, which indicates that LLP adoption improves the overall value of audits. In summary, our empirical findings are consistent with the argument that legal liability is able to effectively shape auditor behavior in emerging markets where the other institutional mechanisms are relatively weaker and government intervention is heavy.