Feature classification criterion for missing features mask estimation in robust speaker recognition

Currently, many speaker recognition applications must handle speech corrupted by environmental additive noise without having a priori knowledge about the characteristics of noise. Some previous works in speaker recognition have used the missing feature (MF) approach to compensate for noise. In most of those applications, the spectral reliability decision step is performed using the signal to noise ratio (SNR) criterion, which attempts to directly measure the relative signal to noise energy at each frequency. An alternative approach to spectral data reliability has been used with some success in the MF approach to speech recognition. Here, we compare the use of this new criterion with the SNR criterion for MF mask estimation in speaker recognition. The new reliability decision is based on the extraction and analysis of several spectro-temporal features from across the entire speech frame, but not across the time, which highlight the differences between spectral regions dominated by speech and by noise. We call it the feature classification (FC) criterion. It uses several spectral features to establish spectrogram reliability unlike SNR criterion that relies only in one feature: SNR. We evaluated our proposal through speaker verification experiments, in Ahumada speech database corrupted by different types of noise at various SNR levels. Experiments demonstrated that the FC criterion achieves considerably better recognition accuracy than the SNR criterion in the speaker verification tasks tested.     

Development of a Simple and Cost-Effective Bead-Milling Method for DNA Extraction from Fish Muscles

In the fish food sector, due to a growing globalization of the market, where intentional and unintentional frauds reach alarming levels, the molecular analysis is increasingly used by both official agencies, to enforce the law on traceability, and private companies, to verify the quality of goods. DNA extraction represents a necessary and critical step for all types of DNA analysis. Among the drawbacks associated with this procedure, there are handling of toxic materials, low DNA yield, and low throughput, due to time-consuming manual procedures. In this work, to overcome some of these problems, we developed an alternative method based on a bead-milling procedure without proteinase K digestion. The new method was then compared with both a salting-out protocol, developed in a previous work, and a commercial kit. Yield, spectrophotometric purity, electrophoretic degradation pattern, and amplificability of the extracted DNA were assessed. In particular, DNA amplificability was evaluated by comparing the band intensity on the gel, after amplification of the 16S rRNA and cytochrome oxidase I genes with a conventional PCR, and the take-off cycles, after amplification of the 16S rRNA gene with a real-time PCR. The results showed that the bead-based method allowed to obtain acceptable amounts of DNA, with good purity and good characteristics of amplificability. Although the salting-out method remains the most effective protocol in terms of pure performances, the bead-milling procedure can be considered a valid alternative, in the light of its lower demand in terms of labor and costs.     

Effect of the cutting process on the fatigue behaviour of press hardened and high strength dual phase steels

Press hardened 22MnB5 steels are steady broadening their applications in vehicles since they allow to meet the increasing demands for weight reduction and safety standards. Press hardened parts have been typically applied as structural reinforcements where high rigidity and crash resistance are required. However, the need for further weight reductions in vehicles leads to potential applications in chassis areas, where parts must face up to cyclic loads. The fatigue behaviour of the press hardened 22MnB5 steel and the effect that post-forming processes (such as trimming or punching that usually follows press hardening) on the fatigue behaviour is scarcely known. Thus, the aim of this work is to analyse the fatigue behaviour of 22MnB5 press hardened steels cut using different strategies. The fatigue limit has been obtained in tensile samples of 22MnB5 with an Al–Si coating. Samples were cut by laser and shearing with two different clearance values. Results are compared to those obtained with a high drawing quality mild steel and a dual phase steel with a tensile strength of 1000 MPa, DP1000. It is shown that the fatigue behaviour of press hardened 22MnB5 steels and DP1000 is governed by the defects introduced in cut edges, while the behaviour of the mild steel is almost independent of the cut edge quality. This finding indicates that high strength steels are markedly sensitive to pre-existent defects, such as burr, cracks at the cut edge or surface cracks, and they can be considered as low damage tolerant steels. On the contrary, mild steels are highly damage tolerant, i.e. the fatigue life is mainly independent of the initial defect size. In press hardened 22MnB5, even when cut edges have a high surface quality, the fatigue limit is still limited by the presence of cracks in the brittle Al–Si coating. Crack propagation due to the low damage tolerance capacity of the press hardened steel has been successfully rationalized through a fracture mechanics approach. Thus, if coated press hardened 22MnB5 steels are to be applied in vehicle components subjected to cyclic loads, they must be designed following fracture mechanics concepts to state the safe loading conditions for adequate fatigue resistance.     

Crystal Growth of Calcium Carbonate in Hydrogels as a Model of Biomineralization

In recent years, the prevalence of hydrogel‐like organic matrices in biomineralization has gained attention as a route to synthesizing a diverse range of crystalline structures. Here, examples of hydrogels in biological, as well as synthetic, bio‐inspired systems are discussed. Particular attention is given to understanding the physical versus chemical effects of a broad range of hydrogel matrices and their role in directing polymorph selectivity and morphological control in the calcium carbonate system. Finally, recent data regarding hydrogel‐matrix incorporation into the growing crystals is discussed and a mechanism for the formation of these single‐crystal composite materials is presented.     

Effect of hydrothermal ageing on the performance of Ce-promoted PdO/ZrO2 for methane combustion

The performance of different Ce-modified PdO/ZrO₂ catalysts for methane oxidation in lean mixtures (5000 ppm of CH₄) in presence of external water has been studied in this work. Deactivation experiments carried out in presence of 20,000 ppm of external water showed that water reversibly inhibits the reaction. However, it was observed that these catalysts can increase their activity in presence of water at low temperature (350 °C).In order to explain this behaviour, different samples of this catalyst were treated with wet air (20,000 ppm of H₂O for 30 h). After this pre-treatment, their activity and stability for methane combustion were studied by recording light-off curves for the fresh catalysts and the catalyst after 50 h on stream for the oxidation of methane at 500 °C. As general trend, the hydro-ageing at the lowest temperature (300 °C) leaded to a very active catalyst (similar activity than the parent one), but it was more markedly deactivated. Hydro-ageing of the catalyst at higher temperatures enhanced its thermal stability.     

Implications of nitrogen compounds during alcoholic fermentation from some grape varieties at different maturation stages and cultivation systems

The aim of this work was to study the influence of grape variety, cultivation system and stage of grape maturation on nitrogen compounds evolution during alcoholic fermentation. To do this, four grape varieties, Monastrell, Merlot, Syrah and Petit Verdot, traditionally cultivated and Monastrell cultivated using organic agriculture, which were collected in two different stages of maturation, were used. The results showed that, regardless of grape variety, cultivation system and stage of grape maturation, the consumption of nitrogen compounds was directly proportional (R² > 0.7) to their concentration in natural musts. This is very important as it is the first time that these results were obtained with natural musts without external addition of nitrogen compounds.     

Interactions between yeast autolysates and volatile compounds in wine and model solution

The addition of a commercial yeast autolysate to a model solution of five typical wine aroma compounds (ethyl octanoate, linalool, 2-phenylethanol, β-ionone and octanoic acid) was investigated considering different variables, such as temperature, pH and the presence of highly concentrated natural volatile substances in wine (e.g. 3-methyl-1-butanol). The interactions of such compounds with both yeast walls and released colloids were studied using gas chromatography, with liquid–liquid extraction and solid-phase microextraction. The results were compared with those obtained by adding the commercial product to a white table wine, spiked with the five standard compounds. The data confirmed that yeast walls mainly bind less polar molecules: their loss in synthetic medium seemed to increase at higher pH values. Temperature and pH affected differently the interactions between yeast colloids and volatile compounds in wine and model solution: in complex solutions (as the addition of 3-methyl-1-butanol demonstrated) the interaction mechanisms could be influenced by competitive or other matrix-related effects, which can reduce the binding of single compounds, or even enhance their volatility.     

Reactive blending of poly(butylene succinate) and poly(triethylene succinate): characterization of the copolymers obtained

Poly[butylene‐co‐(triethylene succinate)] block copolymers (PBSPTES), prepared by reactive blending of the parent homopolymers (PBS and PTES) in the presence of Ti(OBu)₄, were analysed by nuclear magnetic resonance (NMR), wide angle X‐ray diffraction, thermogravimetric analysis and differential scanning calorimetry techniques, in order to investigate the effects of the transesterification reactions on the molecular structure and thermal properties. ¹H‐NMR analysis evidenced the formation of copolymers whose degree of randomness increased with mixing time. After melt quenching, all the copolymers were found to be semicrystalline with the exception of a random copolymer. The results obtained indicated that the block size has a fundamental role in determining the rate of crystallization and therefore the phase behavior of the block copolymers. Copyright © 2012 Society of Chemical Industry     

Internal corrosion studies in hydrocarbons production pipelines located at Venezuelan Northeastern

The demand of fossil energy has boosted the construction of new oil facilities and the preservation of the physical and mechanical integrity of the already existing infrastructure. Corrosion is the main causes of failures in the hydrocarbons industry and half of them are produced by acid gases (CO₂ and H₂S) (Kermany and Harrop, 1996). Within this framework, a monitoring device was developed by PDVSA-Intevep, in order to study the internal corrosion mechanism in a real hydrocarbons production system. The results show the advantages offered by this tool, which allowed studying thoroughly the internal corrosion mechanism present in the system and their immediate causes. In this case, the main corrosion mechanism present is under deposit corrosion, which causes pitting damage at different positions along the pipeline. According to microscopic analyses, the pitting are related to the presence of sand and solids (iron carbonates and sulphides); and the location of the damage depends basically on the internal diameter of pipelines (the hydrodynamics of the system).     

Spin Wave Excitations and Winter’s Magnons in Vertically Coupled Vortex State Permalloy Dots

Circular soft magnetic dots are the main elements of many proposed novel spintronics devices, capable of fascinating spin-based electronics applications, from extremely sensitive magnetic field sensors, to current-tunable microwave vortex oscillators. Here, we investigate static and broadband dynamic magnetization responses of vertically coupled Permalloy (Py) magnetic dots in the vortex state in layered nanopillars (experiment and simulations), which were explored as a function of in-plane magnetic field and interlayer separation. Under reduction of magnetic field from saturation for the field range just above vortex-vortex ground state. We observe a metastable double vortex state for each of the dots. In this state, novel kinds of spin waves (Winter’s magnons along domain walls between vortex cores and half-edge antivortex) are excited. For dipolarly coupled circular Py(25 nm)/Cu(20 nm)/Py(25 nm) trilayer nanopilars of diameter 600 nm, a small in-plane field splits the eigenfrequencies of azimuthal spin wave modes inducing an abrupt transition between acoustic (in-phase) and optic (out-of-phase) kinds of the low-lying coupled spin wave modes. Qualitatively similar changes (although more gradual and at higher values of in-plane fields) occur in the exchange coupled Py(25 nm)/Cu(1 nm)/Py(25 nm) trilayer nanopillars. These findings are in qualitative agreement with micromagnetic dynamic simulations.