Kernel analysis on Grassmann manifolds for action recognition

Modelling video sequences by subspaces has recently shown promise for recognising human actions. Subspaces are able to accommodate the effects of various image variations and can capture the dynamic properties of actions. Subspaces form a non-Euclidean and curved Riemannian manifold known as a Grassmann manifold. Inference on manifold spaces usually is achieved by embedding the manifolds in higher dimensional Euclidean spaces. In this paper, we instead propose to embed the Grassmann manifolds into reproducing kernel Hilbert spaces and then tackle the problem of discriminant analysis on such manifolds. To achieve efficient machinery, we propose graph-based local discriminant analysis that utilises within-class and between-class similarity graphs to characterise intra-class compactness and inter-class separability, respectively. Experiments on KTH, UCF Sports, and Ballet datasets show that the proposed approach obtains marked improvements in discrimination accuracy in comparison to several state-of-the-art methods, such as the kernel version of affine hull image-set distance, tensor canonical correlation analysis, spatial-temporal words and hierarchy of discriminative space-time neighbourhood features.     

Development of a method for quantification of acrolein-deoxyguanosine adducts in DNA using isotope dilution-capillary LC/MS/MS and its application to human brain tissue

Acrolein is a highly reactive alpha,beta-unsaturated aldehyde and is known to react with DNA forming exocyclic acrolein-deoxyguanosine adducts (Acro-dG). These aldehyde-DNA lesions may play a role in mutagenesis, carcinogenesis, and neurodegenerative diseases. In the present work, we described the development and evaluation of a highly sensitive and selective capillary liquid chromatography nanoelectrospray isotope dilution tandem mass spectrometry method for quantitatively analyzing Acro-dG in DNA hydrolysates. This was achieved by applying a stable isotope-labeled analogue Acro-dG-13C10,15N5 as an internal standard to the DNA to be analyzed and then hydrolyzing the DNA enzymatically to nucleosides. The acrolein-modified nucleosides were separated from normal nucleosides by capillary liquid chromatography and quantified by a high-capacity ion trap mass spectrometer in the MS/MS mode. The developed method achieved attomole-level sensitivity (limit of detection was 10 fg, 31 amol on column) for detection of pure Acro-dG adduct standards. The limit of quantification of Acro-dG adducts obtained in 10 mug of DNA hydrolysates was 1.5 fmol, which corresponded to 50 adducts/10(9) normal nucleosides. Application of this method to the analysis of Acro-dG adducts in acrolein (10-fold)-treated calf thymus DNA showed approximately 830 lesion/10(6) DNA nucleosides using as low as 50 ng of DNA. Application of this method to DNA samples (1-2 mug) isolated from brain tissues from Alzheimer's disease subjects and age-matched controls demonstrated 2800-5100 Acro-dG adducts/10(9) DNA nucleosides. Statistically significant differences (P < 0.05) in levels of Acro-dG between AD subjects and controls were observed in DNA isolated from the hippocampus/parahippocampal gyrus.     

Water soluble polymer/carbon nanotube bulk heterojunction solar cells

We report the characteristics of polymer/quantum dot solar cells fabricated using a water-soluble polymer and carbon nanotubes in a bulk heterojunction configuration. The water-soluble polythiophene polymer showed significant photoresponse and the potential for use in photovoltaics. The addition of carbon nanotubes to the polymer resulted in an order of magnitude increase in the photoconductivity. Improved charge separation and collection was evidenced by the large difference between light and dark conductivities as well as the increase in both open circuit voltage and short circuit current. Finally, photovoltaic cells using aligned nanotubes showed further improvement in the photoconductivity and IV characteristics.     

Stability of the color-opponent signals under changes of illuminant in natural scenes

Illumination varies greatly both across parts of a natural scene and as a function of time, whereas the spectral reflectance function of surfaces remains more stable and is of much greater relevance when searching for specific targets. This study investigates the functional properties of postreceptoral opponent-channel responses, in particular regarding their stability against spatial and temporal variation in illumination. We studied images of natural scenes obtained in UK and Uganda with digital cameras calibrated to produce estimated L-, M-, and S-cone responses of trichromatic primates (human) and birds (starling). For both primates and birds we calculated luminance and red-green opponent (RG) responses. We also calculated a primate blue-yellow-opponent (BY) response. The BY response varies with changes in illumination, both across time and across the image, rendering this factor less invariant. The RG response is much more stable than the BY response across such changes in illumination for primates, less so for birds. These differences between species are due to the greater separation of bird L and M cones in wavelength and the narrower bandwidth of the cone action spectra. This greater separation also produces a larger chromatic signal for a given change in spectral reflectance. Thus bird vision seems to suffer a greater degree of spatiotemporal "clutter" than primate vision, but also enhances differences between targets and background. Therefore, there may be a trade-off between the degree of chromatic clutter in a visual system versus the degree of chromatic difference between a target and its background. Primate and bird visual systems have found different solutions to this trade-off.     

Graft copolymer separators — development and use

Graft copolymer membranes have been used for a number of years as interelectrode separators in alkaline batteries either singly, laminated to cellophane, or in conjunction with other materials such as felts.The preparation details of some of the Royal Military College of Science's copolymers are summarised with emphasis on the production of commercial quantities having a high degree of uniformity and reproducibility. Important properties in a battery environment are discussed and, where possible, compared with other separator materials; notably cellophane.The use of graft copolymers in a number of primary and secondary alkaline battery systems is reported and test data given. Conclusions are drawn relating the properties of these separators to their performance in cells.     

Dependence of coal liquefaction behaviour on coal characteristics. 2. Role of petrographic composition

The techniques used were the same as those used in Part 1 (p 34). Comparison of the liquefaction behaviour of two lithotypes from a Kentucky bituminous coal indicated that in this process pseudovitrinite is a reactive maceral. The hydrogenation of sets of maceral concentrates obtained from a New Mexico sub-bituminous and a Kentucky bituminous coal showed fair correlations between conversion and the total concentration of the presumed reactive macerals (vitrinite, pseudovitrinite and sporinite). Similar concentrates from a Montana lignite showed no such correlation; the one sample that showed a high conversion was a high-density fraction that had a high mineral-matter content and in which nearly all the pyrite in the coal had accumulated. Two samples that have boghead and cannel characteristics gave quite different results on hydrogenation. Both were highly aliphatic in structure and had unusually high hydrogen contents and volatile matter. One, which contained appreciable proportions of sporinite, alginite and resinite, gave essentially no conversion to oil. The other, predominantly vitrinitic but containing alginite as the second most abundant maceral, gave an excellent yield of an oil of low viscosity and aromaticity. It was concluded that although rank, petrographic composition and perhaps geological history are important factors determining liquefaction behaviour, there are other characteristics of coals that may at times override these basic parameters, and the composition of the inorganic matter may be the most significant of these other characteristics.     

Comparing and evaluating HMM ensemble training algorithms using train and test and condition number criteria

Abstract Hidden Markov Models have many applications in signal processing and pattern recognition, but their convergence-based training algorithms are known to suffer from over-sensitivity to the initial random model choice. This paper describes the boundary between regions in which ensemble learning is superior to Rabiners multiple-sequence Baum-Welch training method, and proposes techniques for determining the best method in any arbitrary situation. It also studies the suitability of the training methods using the condition number, a recently proposed diagnostic tool for testing the quality of the model. A new method for training Hidden Markov Models called the Viterbi Path Counting algorithm is introduced and is found to produce significantly better performance than current methods in a range of trials.An erratum to this article can be found at     

Temporo-spatial model construction using the MML and software framework

Development of complex temporo-spatial biological computational models can be a time consuming and arduous task. These models may contain hundreds of differential equations as well as realistic geometries that may require considerable investment in time to ensure that all model components are correctly implemented and error free. To tackle this problem, the Modeling Markup Languages (MML) and software framework is a modular XML/HDF5-based specification and toolkits that aims to simplify this process. The main goal of this framework is to encourage reusability, sharing and storage. To achieve this, the MML framework utilizes the CellML specification and repository, which comprises an extensive range of curated models available for use. The MML framework is an open-source project available at http://mml.gsbme.unsw.edu.au.