Multiple face detection at different resolutions for perceptual user interfaces

[EN]This paper describes in detail a real-time multiple face detection system for video streams. The system adds to the good performance provided by a window shift approach, the combination of different cues available in video streams due to temporal coherence. The results achieved by this combined solution outperform the basic face detector obtaining a 98% success rate for around 27000 images, providing additionally eye detection and a relation between the successive detections in time by means of detection threads.

Multiple Linear Regression and Artificial Neural Networks to Predict Time and Efficiency of Soil Vapor Extraction

The prediction of the time and the efficiency of the remediation of contaminated soils using soil vapor extraction remain a difficult challenge to the scientific community and consultants. This work reports the development of multiple linear regression and artificial neural network models to predict the remediation time and efficiency of soil vapor extractions performed in soils contaminated separately with benzene, toluene, ethylbenzene, xylene, trichloroethylene, and perchloroethylene. The results demonstrated that the artificial neural network approach presents better performances when compared with multiple linear regression models. The artificial neural network model allowed an accurate prediction of remediation time and efficiency based on only soil and pollutants characteristics, and consequently allowing a simple and quick previous evaluation of the process viability.

Lung mechanics and gas exchange in ventilated preterm infants during treatment of hyaline membrane disease with multiple doses of artificial surfactant (Exosurf)

Eight premature infants ventilated for hyaline membrane disease and enrolled in the OSIRIS surfactant trial were studied. Lung mechanics, gas exchange [PaCO2, arterial/alveolar PO2 ratio (a/A ratio)], and ventilator settings were determined 20 minutes before and 20 minutes after the end of Exosurf instillation, and subsequently at 12-24 hour intervals. Respiratory system compliance (Crs) and resistance (Rrs) were measured by means of the single breath occlusion method. After surfactant instillation there were no significant immediate changes in PaCO2 (36 vs. 37 mmHg), a/A ratio (0.23 vs. 0.20), Crs (0.32 vs. 0.31 mL/cm H2O/kg), and Rrs (0.11 vs. 0.16 cmH2O/mL/s) (pooled data of 18 measurement pairs). During the clinical course, mean a/A ratio improved significantly each time from 0.17 (time 0) to 0.29 (time 12-13 hours), to 0.39 (time 24-36 hours) and to 0.60 (time 48-61 hours), although mean airway pressure was reduced substantially. Mean Crs increased significantly from 0.28 mL/cmH2O/kg (time 0) to 0.38 (time 12-13 hours), to 0.37 (time 24-38 hours), and to 0.52 (time 48-61 hours), whereas mean Rrs increased from 0.10 cm H2O/mL/s (time 0) to 0.11 (time 12-13 hours), to 0.13 (time 24-36 hours) and to (time 48-61 hours) with no overall significance. A highly significant correlation was found between Crs and a/A ratio (r = 0.698, P less than 0.001). We conclude that Exosurf does not induce immediate changes in oxygenation as does the instillation of (modified) natural surfactant preparations. However, after 12 and 24 hours of treatment oxygenation and Crs improve significantly.(ABSTRACT TRUNCATED AT 250 WORDS)

Artificial Gait In Complete Spinal Cord Injured Subjects: How To Assess Clinical Performance.

Objective Adapt the 6 minutes walking test (6MWT) to artificial gait in complete spinal cord injured (SCI) patients aided by neuromuscular electrical stimulation. Method Nine male individuals with paraplegia (AIS A) participated in this study. Lesion levels varied between T4 and T12 and time post injured from 4 to 13 years. Patients performed 6MWT 1 and 6MWT 2. They used neuromuscular electrical stimulation, and were aided by a walker. The differences between two 6MWT were assessed by using a paired t test. Multiple r-squared was also calculated. Results The 6MWT 1 and 6MWT 2 were not statistically different for heart rate, distance, mean speed and blood pressure. Multiple r-squared (r2 = 0.96) explained 96% of the variation in the distance walked. Conclusion The use of 6MWT in artificial gait towards assessing exercise walking capacity is reproducible and easy to apply. It can be used to assess SCI artificial gait clinical performance.

Papillomavirus virus-like particles for the delivery of multiple cytotoxic T cell epitopes

Chimeric papillomavirus (PV) virus-like particles (VLPs) based on the bovine papillomavirus type 1 (BPV-1) L1 protein were constructed by replacing the 23-carboxyl-terminal amino acids of the BPV1 major protein L1 with an artificial polytope minigene, containing known CTL epitopes of human PV16 E7 protein, HIV IIIB gp120 P18, Nef, and reverse transcriptase (RT) proteins, and an HPV16 E7 linear B epitope. The CTL epitopes were restricted by three different MHC class 1 alleles (H-2(b), H-2(d), HLA-A*0201). The chimeric L1 protein assembled into VLPs when expressed in SF-9 cells by recombinant baculovirus. After immunization of mice with polytope VLPs in the absence of adjuvant, serum antibodies were detected which reacted with both polytope VLPs and wild-type BPV1L1 VLPs, in addition to the HPV16E7 linear B cell epitope. CTL precursors specific for the HPV16 E7, HIV P18, and RT CTL epitopes were also detected in the spleen of immunized mice. Polytope VLPs can thus deliver multiple B and T epitopes as immunogens to the MHC class I and class II pathways, extending the utility of VLPs as self-adjuvanting immunogen delivery systems. (C) 2000 Academic Press.

Decisions with multiple simultaneous goals and uncertain causal effects

A key aspect of decision-making in a disaster response scenario is the capability to evaluate multiple and simultaneously perceived goals. Current competing approaches to build decision-making agents are either mental-state based as BDI, or founded on decision-theoretic models as MDP. The BDI chooses heuristically among several goals and the MDP searches for a policy to achieve a specific goal. In this paper we develop a preferences model to decide among multiple simultaneous goals. We propose a pattern, which follows a decision-theoretic approach, to evaluate the expected causal effects of the observable and non-observable aspects that inform each decision. We focus on yes-or-no (i.e., pursue or ignore a goal) decisions and illustrate the proposal using the RoboCupRescue simulation environment.

A vertebrate reproductive system involving three ploidy levels : hybrid origin of triploids in a contact zone of diploid and tetraploid green toads (Bufo viridis subgroup)

The rise and consequences of polyploidy in vertebrates, whose origin was associated with genome duplications, may be best studied in natural diploid and polyploid populations. In a diploid/tetraploid (2n/4n) geographic contact zone of Palearctic green toads in northern Kyrgyzstan, we examine 4ns and triploids (3n) of unknown genetic composition and origins. Using mitochondrial and nuclear sequence, and nuclear microsatellite markers in 84 individuals, we show that 4n (Bufo pewzowi) are allopolyploids, with a geographically proximate 2n species (B. turanensis) being their maternal ancestor and their paternal ancestor as yet unidentified. Local 3n forms arise through hybridization. Adult 3n mature males (B. turanensis mtDNA) have 2n mothers and 4n fathers, but seem distinguishable by nuclear profiles from partly aneuploid 3n tadpoles (with B. pewzowi mtDNA). These observations suggest multiple pathways to the formation of triploids in the contact zone, involving both reciprocal origins. To explain the phenomena in the system, we favor a hypothesis where 3n males (with B. turanensis mtDNA) backcross with 4n and 2n females. Together with previous studies of a separately evolved, sexually reproducing 3n lineage, these observations reveal complex reproductive interactions among toads of different ploidy levels and multiple pathways to the evolution of polyploid lineages.

A bi-dimensional genome scan for prolificacy traits in pigs shows the existence of multiple epistatic QTL

Background: Prolificacy is the most important trait influencing the reproductive efficiency of pig production systems. The low heritability and sex-limited expression of prolificacy have hindered to some extent the improvement of this trait through artificial selection. Moreover, the relative contributions of additive, dominant and epistatic QTL to the genetic variance of pig prolificacy remain to be defined. In this work, we have undertaken this issue by performing one-dimensional and bi-dimensional genome scans for number of piglets born alive (NBA) and total number of piglets born (TNB) in a three generation Iberian by Meishan F2 intercross. Results: The one-dimensional genome scan for NBA and TNB revealed the existence of two genome-wide highly significant QTL located on SSC13 (P < 0.001) and SSC17 (P < 0.01) with effects on both traits. This relative paucity of significant results contrasted very strongly with the wide array of highly significant epistatic QTL that emerged in the bi-dimensional genome-wide scan analysis. As much as 18 epistatic QTL were found for NBA (four at P < 0.01 and five at P < 0.05) and TNB (three at P < 0.01 and six at P < 0.05), respectively. These epistatic QTL were distributed in multiple genomic regions, which covered 13 of the 18 pig autosomes, and they had small individual effects that ranged between 3 to 4% of the phenotypic variance. Different patterns of interactions (a × a, a × d, d × a and d × d) were found amongst the epistatic QTL pairs identified in the current work.Conclusions: The complex inheritance of prolificacy traits in pigs has been evidenced by identifying multiple additive (SSC13 and SSC17), dominant and epistatic QTL in an Iberian × Meishan F2 intercross. Our results demonstrate that a significant fraction of the phenotypic variance of swine prolificacy traits can be attributed to first-order gene-by-gene interactions emphasizing that the phenotypic effects of alleles might be strongly modulated by the genetic background where they segregate.

Time Series Input Selection using Multiple Kernel Learning

In this paper we study the relevance of multiple kernel learning (MKL) for the automatic selection of time series inputs. Recently, MKL has gained great attention in the machine learning community due to its flexibility in modelling complex patterns and performing feature selection. In general, MKL constructs the kernel as a weighted linear combination of basis kernels, exploiting different sources of information. An efficient algorithm wrapping a Support Vector Regression model for optimizing the MKL weights, named SimpleMKL, is used for the analysis. In this sense, MKL performs feature selection by discarding inputs/kernels with low or null weights. The approach proposed is tested with simulated linear and nonlinear time series (AutoRegressive, Henon and Lorenz series).

Multiple Kernel Learning of Environmental Data. Case study: Analysis and Mapping of Wind Fields

The paper presents the Multiple Kernel Learning (MKL) approach as a modelling and data exploratory tool and applies it to the problem of wind speed mapping. Support Vector Regression (SVR) is used to predict spatial variations of the mean wind speed from terrain features (slopes, terrain curvature, directional derivatives) generated at different spatial scales. Multiple Kernel Learning is applied to learn kernels for individual features and thematic feature subsets, both in the context of feature selection and optimal parameters determination. An empirical study on real-life data confirms the usefulness of MKL as a tool that enhances the interpretability of data-driven models.

Biometrical analysis and selection of tetraploid progenies of Panicum maximum using mixed model methods

The objectives of this work were to estimate the genetic and phenotypic parameters and to predict the genetic and genotypic values of the selection candidates obtained from intraspecific crosses in Panicum maximum as well as the performance of the hybrid progeny of the existing and projected crosses. Seventy-nine intraspecific hybrids obtained from artificial crosses among five apomictic and three sexual autotetraploid individuals were evaluated in a clonal test with two replications and ten plants per plot. Green matter yield, total and leaf dry matter yields and leaf percentage were evaluated in five cuts per year during three years. Genetic parameters were estimated and breeding and genotypic values were predicted using the restricted maximum likelihood/best linear unbiased prediction procedure (REML/BLUP). The dominant genetic variance was estimated by adjusting the effect of full-sib families. Low magnitude individual narrow sense heritabilities (0.02-0.05), individual broad sense heritabilities (0.14-0.20) and repeatability measured on an individual basis (0.15-0.21) were obtained. Dominance effects for all evaluated characteristics indicated that breeding strategies that explore heterosis must be adopted. Less than 5% increase in the parameter repeatability was obtained for a three-year evaluation period and may be the criterion to determine the maximum number of years of evaluation to be adopted, without compromising gain per cycle of selection. The identification of hybrid candidates for future cultivars and of those that can be incorporated into the breeding program was based on the genotypic and breeding values, respectively. The prediction of the performance of the hybrid progeny, based on the breeding values of the progenitors, permitted the identification of the best crosses and indicated the best parents to use in crosses.

A Robust Multiple Feature Approach To Endpoint Detection In Car Environment Based On Advanced Classifiers

In this paper we propose an endpoint detection system based on the use of several features extracted from each speech frame, followed by a robust classifier (i.e Adaboost and Bagging of decision trees, and a multilayer perceptron) and a finite state automata (FSA). We present results for four different classifiers. The FSA module consisted of a 4-state decision logic that filtered false alarms and false positives. We compare the use of four different classifiers in this task. The look ahead of the method that we propose was of 7 frames, which are the number of frames that maximized the accuracy of the system. The system was tested with real signals recorded inside a car, with signal to noise ratio that ranged from 6 dB to 30dB. Finally we present experimental results demonstrating that the system yields robust endpoint detection.

Determination of the influence of the variation of reducing and non-reducing sugars on coffee quality with use of artificial neural network

The present study aimed at evaluating the use of Artificial Neural Network to correlate the values resulting from chemical analyses of samples of coffee with the values of their sensory analyses. The coffee samples used were from the Coffea arabica L., cultivars Acaiá do Cerrado, Topázio, Acaiá 474-19 and Bourbon, collected in the southern region of the state of Minas Gerais. The chemical analyses were carried out for reducing and non-reducing sugars. The quality of the beverage was evaluated by sensory analysis. The Artificial Neural Network method used values from chemical analyses as input variables and values from sensory analysis as output values. The multiple linear regression of sensory analysis values, according to the values from chemical analyses, presented a determination coefficient of 0.3106, while the Artificial Neural Network achieved a level of 80.00% of success in the classification of values from the sensory analysis.