Skin color-based video segmentation under time-varying illumination

A novel approach for real-time skin segmentation in video sequences is described. The approach enables reliable skin segmentation despite wide variation in illumination during tracking. An explicit second order Markov model is used to predict evolution of the skin-color (HSV) histogram over time. Histograms are dynamically updated based on feedback from the current segmentation and predictions of the Markov model. The evolution of the skin-color distribution at each frame is parameterized by translation, scaling, and rotation in color space. Consequent changes in geometric parameterization of the distribution are propagated by warping and resampling the histogram. The parameters of the discrete-time dynamic Markov model are estimated using Maximum Likelihood Estimation and also evolve over time. The accuracy of the new dynamic skin color segmentation algorithm is compared to that obtained via a static color model. Segmentation accuracy is evaluated using labeled ground-truth video sequences taken from staged experiments and popular movies. An overall increase in segmentation accuracy of up to 24 percent is observed in 17 out of 21 test sequences. In all but one case, the skin-color classification rates for our system were higher, with background classification rates comparable to those of the static segmentation.     

Input Feature Extraction for Multilayered Perceptrons Using Supervised Principal Component Analysis

A method is proposed for constructing salient features from a set of features that are given as input to a feedforward neural network used for supervised learning. Combinations of the original features are formed that maximize the sensitivity of the network's outputs with respect to variations of its inputs. The method exhibits some similarity to Principal Component Analysis, but also takes into account supervised character of the learning task. It is applied to classification problems leading to improved generalization ability originating from the alleviation of the curse of dimensionality problem.     

Semantic content of motorcycle riders’ eye fixations during lane-splitting

A naturalistic field observation study is presented, exploring the experienced riders’ eye fixations on specific traffic objects and their relative s     

On extended P4-reducible and extended P4-sparse graphs

A graph G was defined in [16] as P₄-reducible, if no vertex in G belongs to more than one chordless path on four vertices or P₄. A graph G is defined in [15] as P₄-sparse if no set of five vertices induces more than one P₄, in G. P₄-sparse graphs generalize both P₄-reducible and the well known class of p₄-free graphs or cographs. In an extended abstract in [11] the first author introduced a method using the modular decomposition tree of a graph as the framework for the resolution of algorithmic problems. This method was applied to the study of P₄-sparse and extended P₄-sparse graphs.

In this paper, we begin by presenting the complete information about the method used in [11]. We propose a unique tree representation of P₄-sparse and a unique tree representation of P₄-reducible graphs leading to a simple linear recognition algorithm for both classes of graphs. In this way we simplify and unify the solutions for these problems, presented in [16–19]. The tree representation of an n-vertex P₄-sparse or a P₄-reducible graph is the key for obtaining O(n) time algorithms for the weighted version of classical optimization problems solved in [20]. These problems are NP-complete on general graphs.

Finally, by relaxing the restriction concerning the exclusion of the C₅ cycles from P₄-sparse and P₄-reducible graphs, we introduce the class of the extended P4-sparse and the class of the extendedP₄-reducible graphs. We then show that a minimal amount of additional work suffices for extending most of our algorithms to these new classes of graphs.     

Advanced glycation endproducts stimulate the MAP-kinase pathway in tubulus cell line LLC-PK1

Advanced glycation endproducts (AGEs) are suggested to play an important role in diabetic nephropathy. They induce specific cellular responses such as the release of cytokines in different cell lines. The effect of AGEs on signal transduction pathways was investigated in the renal tubulus cell line LLC-PK1. Using a serine-phosphate-specific antibody AGE-induced cellular responses associated with phosphorylation/dephosphorylation events were demonstrated. In particular, the p42MAP kinase and its downstream target, the AP-1 complex, are shown to be activated by AGE-BSA but not by BSA. In contrast, only partial phosphorylation is observed for the p70S6-kinase. Thus, AGEs appear to induce specific signal transduction pathways.     

Are Phosphatidylcholine and Lysophosphatidylcholine Body Levels Potentially Reliable Biomarkers in Obesity? A Review of Human Studies

Phosphatidylcholines (PCs) are the major components of biological membranes in animals and are a class of phospholipids that incorporate choline as a headgroup. Lysophosphatidylcholines (LPCs) are a class of lipid biomolecules derived from the cleavage of PCs, and are the main components of oxidized low-density lipoproteins (oxLDLs) that are involved in the pathogenesis of atherosclerosis. Since obesity is associated with a state of chronic low-grade inflammation, one can anticipate that the lipidomic profile changes in this context and both PCs and LPCs are gaining attention as hypothetically reliable biomarkers of obesity. Thus, a literature search is performed on PubMed, Latin American and Caribbean Health Science Literature (LILACS), and Excerpta Medica DataBASE (Embase) to obtain the findings of population studies to clarify this hypothesis. The search strategy resulted in a total of 2403 reports and 21 studies were included according to the eligibility criteria. Controversial data on the associations of PCs and LPCs with body mass index (BMI) and body fat parameters have been identified. There is an inverse relationship between BMI and most species of PCs, and a majority of studies exhibited negative associations between BMI and LPCs. Other findings regarding the differences between PCs and LPCs in obesity are presented, and the associated uncertainties are discussed in detail.