Regularized Online Mixture of Gaussians for Background Subtraction

Mixture of Gaussians (MoG) modelling [13] is a popular approach to background subtraction in video sequences. Although the algorithm shows good empirical performance, it lacks theoretical justification. In this paper, we give a justification for it from an online stochastic expectation maximization (EM) viewpoint and extend it to a general framework of regularized online classification EM for MoG with guaranteed convergence. By choosing a special regularization function, l1 norm, we derived a new set of updating equations for l1 regularized online MoG. It is shown empirically that l1 regularized online MoG converge faster than the original online MoG .

Root Cause Analysis by a Combined Sparse Classification and Monte Carlo Approach

Classification methods with embedded feature selection capability are very appealing for the analysis of complex processes since they allow the analysis of root causes even when the number of input variables is high. In this work, we investigate the performance of three techniques for classification within a Monte Carlo strategy with the aim of root cause analysis. We consider the naive bayes classifier and the logistic regression model with two different implementations for controlling model complexity, namely, a LASSO-like implementation with a L1 norm regularization and a fully Bayesian implementation of the logistic model, the so called relevance vector machine. Several challenges can arise when estimating such models mainly linked to the characteristics of the data: a large number of input variables, high correlation among subsets of variables, the situation where the number of variables is higher than the number of available data points and the case of unbalanced datasets. Using an ecological and a semiconductor manufacturing dataset, we show advantages and drawbacks of each method, highlighting the superior performance in term of classification accuracy for the relevance vector machine with respect to the other classifiers. Moreover, we show how the combination of the proposed techniques and the Monte Carlo approach can be used to get more robust insights into the problem under analysis when faced with challenging modelling conditions.

Robust visual tracking using structurally random projection and weighted least squares

Sparse representation based visual tracking approaches have attracted increasing interests in the community in recent years. The main idea is to linearly represent each target candidate using a set of target and trivial templates while imposing a sparsity constraint onto the representation coefficients. After we obtain the coefficients using L1-norm minimization methods, the candidate with the lowest error, when it is reconstructed using only the target templates and the associated coefficients, is considered as the tracking result. In spite of promising system performance widely reported, it is unclear if the performance of these trackers can be maximised. In addition, computational complexity caused by the dimensionality of the feature space limits these algorithms in real-time applications. In this paper, we propose a real-time visual tracking method based on structurally random projection and weighted least squares techniques. In particular, to enhance the discriminative capability of the tracker, we introduce background templates to the linear representation framework. To handle appearance variations over time, we relax the sparsity constraint using a weighed least squares (WLS) method to obtain the representation coefficients. To further reduce the computational complexity, structurally random projection is used to reduce the dimensionality of the feature space while preserving the pairwise distances between the data points in the feature space. Experimental results show that the proposed approach outperforms several state-of-the-art tracking methods.

Cathepsin L1, the major protease involved in liver fluke (Fasciola hepatica) virulence.

The secretion and activation of the major cathepsin L1 cysteine protease involved in the virulence of the helminth pathogen Fasciola hepatica was investigated. Only the fully processed and active mature enzyme can be detected in medium in which adult F. hepatica are cultured. However, immunocytochemical studies revealed that the inactive procathepsin L1 is packaged in secretory vesicles of epithelial cells that line the parasite gut. These observations suggest that processing and activation of procathepsin L1 occurs following secretion from these cells into the acidic gut lumen. Expression of the 37-kDa procathepsin L1 in Pichia pastoris showed that an intermolecular processing event within a conserved GXNXFXD motif in the propeptide generates an active 30-kDa intermediate form. Further activation of the enzyme was initiated by decreasing the pH to 5.0 and involved the progressive processing of the 37 and 30-kDa forms to other intermediates and finally to a fully mature 24.5 kDa cathepsin L with an additional 1 or 2 amino acids. An active site mutant procathepsin L, constructed by replacing the Cys26 with Gly26, failed to autoprocess. However, [Gly26]procathepsin L was processed by exogenous wild-type cathepsin L to a mature enzyme plus 10 amino acids attached to the N terminus. This exogenous processing occurred without the formation of a 30-kDa intermediate form. The results indicate that activation of procathepsin L1 by removal of the propeptide can occur by different pathways, and that this takes place within the parasite gut where the protease functions in food digestion and from where it is liberated as an active enzyme for additional extracorporeal roles.

Norm attaining operators and pseudospectrum

It is shown that if $11$, the operator $I+T$ attains its norm. A reflexive Banach space $X$ and a bounded rank one operator $T$ on $X$ are constructed such that $\|I+T\|>1$ and $I+T$ does not attain its norm.

The level sets of the resolvent norm and convexity properties of Banach spaces

We construct a bounded linear operator on a separable, reflexive and strictly convex Banach space whose resolvent norm is constant in a neighbourhood of zero.

Protection of cattle against a natural infection of Fasciola hepatica by vaccination with recombinant cathepsin L1 (rFhCL1)

The liver fluke, Fasciola hepatica causes liver fluke disease, or fasciolosis, in ruminants such as cattle and sheep. An effective vaccine against the helminth parasite is essential to reduce our reliance on anthelmintics, particularly in light of frequent reports of resistance to some frontline drugs. In our study, Friesian cattle (13 per group) were vaccinated with recombinant F. hepatica cathepsin L1 protease (rFhCL1) formulated in mineral-oil based adjuvants, Montanide (TM) ISA 70VG and ISA 206VG. Following vaccination the animals were exposed to fluke-contaminated pastures for 13 weeks. At slaughter, there was a significant reduction in fluke burden of 48.2% in the cattle in both vaccinated groups, relative to the control non-vaccinated group, at p

Fluorobenzoyl dipeptidyl derivatives as inhibitors of the Fasciola hepatica cysteine protease cathepsin L1

Cathepsins are known to have many important physiological roles and provide a viable target for inhibition. Fluorobenzoyl dipeptide derivatives were synthesized and tested for biological activity in an effort to find an efficient inhibitor of the cysteine protease cathepsin L. Thirty-six novel inhibitors (1-36) were synthesized from protected amino acids via the standard DCC/HOBt coupling protocol, containing a benzyl ester or a nitrile as an electrophilic warhead. The activity of the inhibitors was evaluated against cathepsin L and IC50 values calculated. Modification of both amino acids and terminal groups afforded compounds with single digit micromolar inhibition. Results utilizing the benzoyl-L-leucine-glycine nitrile backbone are comparable to that for the commercially available inhibitor 39.