Improved voice activity detection algorithm using wavelet and support vector machine

This paper proposes an improved voice activity detection (VAD) algorithm using wavelet and support vector machine (SVM) for European Telecommunication Standards Institution (ETS1) adaptive multi-rate (AMR) narrow-band (NB) and wide-band (WB) speech codecs. First, based on the wavelet transform, the original IIR filter bank and pitch/tone detector are implemented, respectively, via the wavelet filter bank and the wavelet-based pitch/tone detection algorithm. The wavelet filter bank can divide input speech signal into several frequency bands so that the signal power level at each sub-band can be calculated. In addition, the background noise level can be estimated in each sub-band by using the wavelet de-noising method. The wavelet filter bank is also derived to detect correlated complex signals like music. Then the proposed algorithm can apply SVM to train an optimized non-linear VAD decision rule involving the sub-band power, noise level, pitch period, tone flag, and complex signals warning flag of input speech signals. By the use of the trained SVM, the proposed VAD algorithm can produce more accurate detection results. Various experimental results carried out from the Aurora speech database with different noise conditions show that the proposed algorithm gives considerable VAD performances superior to the AMR-NB VAD Options 1 and 2, and AMR-WB VAD. (C) 2009 Elsevier Ltd. All rights reserved.

Controls of heavy minerals and grain size in a holocene regressive barrier (Ilha Comprida, southeastern Brazil)

Ilha Comprida is a regressive barrier island located in southeastern Brazil that was formed essentially by Quaternary sandy sediments. Ilha Comprida sediments were analyzed to assess heavy mineral indices and grain size variables. The spatial variation of heavy minerals and grain size was interpreted in terms of the present barrier dynamics and the barrier`s evolution since the Middle Holocene. These analyses allowed for the identification of the main factors and processes that control the variation of heavy minerals and grain size on the barrier. Rutile and zircon (RZi) and tourmaline and hornblende (THi) are significantly sensitive to provenance and exhibit the contributions of the Ribeira de Iguape River sediments, which reach the coast next to the northeastern end of Ilha Comprida. In addition to the influence of provenance, TZi responds mainly to hydraulic sorting processes. This agrees with a sediment transport pattern characterized by a divergence of two resultant net alongshore drifts southwest of the barrier. The sediments from the Ribeira de Iguape River reach the barrier directly through the river mouth and indirectly after temporary storage in the inner shelf. The combination of grain size and heavy mineral analyses is a reliable method for determining sediment transport patterns and provenance. (C) 2010 Elsevier Ltd. All rights reserved.

Grain size and heavy minerals of the Late Quaternary eolian sediments from the Imbituba-Jaguaruna coast, Southern Brazil: Depositional controls linked to relative sea-level changes

The stratigraphic subdivision and correlation of dune deposits is difficult, especially when age datings are not available. A better understanding of the controls on texture and composition of eolian sands is necessary to interpret ancient eolian sediments. The Imbituba-Jaguaruna coastal zone (Southern Brazil, 28 degrees-29 degrees S) stands out due to its four well-preserved Late Pleistocene (eolian generation 1) to Holocene eolian units (eolian generations 2, 3, and 4). In this study, we evaluate the grain-size and heavy-mineral characteristics of the Imbituba-Jaguartma eolian units through statistical analysis of hundreds of sediment samples. Grain-size parameters and heavy-mineral content allow us to distinguish the Pleistocene from the Holocene units. The grain size displays a pattern of fining and better sorting from generation 1 (older) to 4 (younger), whereas the content of mechanically stable (dense and hard) heavy minerals decreases from eolian generation 1 to 4. The variation in grain size and heavy-mineral content records shifts in the origin and balance (input versus output) of eolian sediment supply attributable mainly to relative sea-level changes. Dunefields submitted to relative sea-level lowstand conditions (eolian generation 1) are characterized by lower accumulation rates and intense post-depositional dissection by fluvial incision. Low accumulation rates favor deflation in the eolian system, which promotes concentration of denser and stable heavy minerals (increase of ZTR index) as well as coarsening of eolian sands. Dissection involves the selective removal of finer sediments and less dense heavy minerals to the coastal source area. Under a high rate of relative sea-level rise and transgression (eolian generation 2), coastal erosion prevents deflation through high input of sediments to the coastal eolian source. This condition favors dunefield growth. Coastal erosion feeds sand from local sources to the eolian system. including sands from previous dunefields (eolian generation 1) and from drowned incised valleys. Therefore, dunefields corresponding to transgressive phases inherit the grain-size and heavy-mineral characteristics of previous dunefields, leading to selective enrichment of finer sands and lighter minerals. Eolian generations 3 and 4 developed during a regressive-progradational phase (Holocene relative sea level highstand). The high rate of sediment supply during the highstand phase prevents deflation. The lack of coastal erosion favors sediment supply from distal sources (fluvial sediments rich in unstable heavy minerals). Thus, dunefields of transgressive and highstand systems tracts may be distinguished from dunefields of the lowstand systems tract through high rates of accumulation (low deflation) in the former. The sediment source of the transgressive dunefields (high input of previously deposited coastal sands) differs from that of the highstand dunefields (high input of fluvial distal sands). Based on this case study, we propose a general framework for the relation between relative sea level, sediment supply and the texture and mineralogy of eolian sediments deposited in siliciclastic wet coastal zones similar to the Imbituba-Jaguaruna coast (C) 2009 Elsevier B.V. All rights reserved.

Measuring time series predictability using support vector regression

Most studies involving statistical time series analysis rely on assumptions of linearity, which by its simplicity facilitates parameter interpretation and estimation. However, the linearity assumption may be too restrictive for many practical applications. The implementation of nonlinear models in time series analysis involves the estimation of a large set of parameters, frequently leading to overfitting problems. In this article, a predictability coefficient is estimated using a combination of nonlinear autoregressive models and the use of support vector regression in this model is explored. We illustrate the usefulness and interpretability of results by using electroencephalographic records of an epileptic patient.

UNIT VECTOR FIELDS ON ANTIPODALLY PUNCTURED SPHERES: BIG INDEX, BIG VOLUME

We establish in this paper a lower bound for the volume of a unit vector field (v) over right arrow defined ou S(n) \ {+/-x}, n = 2,3. This lower bound is related to the sum of the absolute values of the indices of (v) over right arrow at x and -x.

Periodic geodesics and geometry of compact Lorentzian manifolds with a Killing vector field

We study the geometry and the periodic geodesics of a compact Lorentzian manifold that has a Killing vector field which is timelike somewhere. Using a compactness argument for subgroups of the isometry group, we prove the existence of one timelike non self-intersecting periodic geodesic. If the Killing vector field is nowhere vanishing, then there are at least two distinct periodic geodesics; as a special case, compact stationary manifolds have at least two periodic timelike geodesics. We also discuss some properties of the topology of such manifolds. In particular, we show that a compact manifold M admits a Lorentzian metric with a nowhere vanishing Killing vector field which is timelike somewhere if and only if M admits a smooth circle action without fixed points.

A short proregion of trialysin, a pore-forming protein of Triatoma infestans salivary glands, controls activity by folding the N-terminal lytic motif

Triatoma infestans (Hemiptera: Reduviidae) is a hematophagous insect that transmits the protozoan parasite Trypanosoma cruzi, the etiological agent of Chagas` disease. Its saliva contains trialysin, a protein that forms pores in membranes. Peptides based on the N-terminus of trialysin lyse cells and fold into alpha-helical amphipathic segments resembling antimicrobial peptides. Using a specific antiserum against trialysin, we show here that trialysin is synthesized as a precursor that is less active than the protein released after saliva secretion. A synthetic peptide flanked by a fluorophore and a quencher including the acidic proregion and the lytic N-terminus of the protein is also less active against cells and liposomes, increasing activity upon proteolysis. Activation changes the peptide conformation as observed by fluorescence increase and CD spectroscopy. This mechanism of activation could provide a way to impair the toxic effects of trialysin inside the salivary glands, thus restricting damaging lytic activity to the bite site.

Base Excision Repair Activities Differ in Human Lung Cancer Cells and Corresponding Normal Controls

Oxidative damage to DNA is thought to play a role in carcinogenesis by causing Mutations, and indeed accumulation of oxidized DNA bases has been observed in samples obtained from tumors but not from surrounding tissue within the same patient. Base excision repair (BER) is the main pathway for the repair of oxidized modifications both in nuclear and mitochondrial, DNA. In order to ascertain whether diminished BER capacity might account for increased levels of oxidative DNA damage in cancer cells, the activities of BER enzymes in three different lung cancer cell lines and their non-cancerous counterparts were measured using oligonucleotide substrates with single DNA lesions to assess specific BER enzymes. The activities of four BER enzymes, OGG1, NTH1, UDG and APE1, were compared in mitochondrial and nuclear extracts. For each specific lesion, the repair activities were similar among the three cell lines used. However, the specific activities and cancer versus control comparison differed significantly between the nuclear and mitochondrial compartments. OGG1 activity, as measured by 8-oxodA incision, was upregulated in cancer cell mitochondria but down-regulated in the nucleus when compared to control cells. Similarly, NTH1 activity was also up-regulated in mitochondrial extracts from cancer cells but did not change significantly in the nucleus. Together, these results support the idea that alterations in BER capacity are associated with carcinogenesis.

Arginine vasopressin controls p27(KiP1) protein expression by PKC activation and irreversibly inhibits the proliferation of K-Ras-dependent mouse Y1 adrenocortical malignant cells

The neurohypophyseal hormone arginine vasopressin (AVP) is a classic mitogen in many cells. In K-Ras-dependent mouse Y1 adrenocortical malignant cells, AVP elicits antagonistic responses such as the activation of the PKC and the ERK1/2 mitogenic pathways to down-regulate cyclin D1 gene expression, which induces senescence-associated beta-galactosidase (SA-beta Gal) and leads to cell cycle arrest. Here, we report that in the metabolic background of Y1 cells, PKC activation either by AVP or by PMA inhibits the PI3K/Akt pathway and stabilises the p27(Kip1) protein even in the presence of the mitogen fibroblast growth factor 2 (FGF2). These results suggest that p27(Kip1) is a critical signalling node in the mechanisms underlying the survival of the Y1 cells. In Y1 cells that transiently express wild-type p27(Kip1), AVP caused a severe reduction in cell survival, as shown by clonogenic assays. However, AVP promoted the survival of Y1 cells transiently expressing mutant p27-S10A or mutant p27-T187A, which cannot be phosphorylated at Ser10 and Thr187, respectively. In addition, PKC activation by PMA mimics the toxic effect caused by AVP in Y1 cells, and inhibition of PKC completely abolishes the effects caused by both PMA and AVP in clonogenic assays. The vulnerability of Y1 cells during PKC activation is a phenotype conditioned upon K-ras oncogene amplification because K-Ras down-regulation with an inducible form of the dominant-negative mutant H-RasN17 has resulted in Y1 cells that are resistant to AVP`s deleterious effects. These data show that the survival destabilisation of K-Ras-dependent Y1 malignant cells by AVP requires large quantities of the p27(Kip1) protein as well as phosphorylation of the p27(Kip1) protein at both Ser10 and Thr187. (C) 2011 Elsevier B.V. All rights reserved.

Identification of an Operon, Pil-Chp, That Controls Twitching Motility and Virulence in Xylella fastidiosa

Xylella fastidiosa is an important phytopathogenic bacterium that causes many serious plant diseases, including Pierce`s disease of grapevines. Disease manifestation by X. fastidiosa is associated with the expression of several factors, including the type IV pili that are required for twitching motility. We provide evidence that an operon, named Pil-Chp, with genes homologous to those found in chemotaxis systems, regulates twitching motility. Transposon insertion into the pilL gene of the operon resulted in loss of twitching motility (pilL is homologous to cheA genes encoding kinases). The X. fastidiosa mutant maintained the type IV pili, indicating that the disrupted pilL or downstream operon genes are involved in pili function, and not biogenesis. The mutated X. fastidiosa produced less biofilm than wild-type cells, indicating that the operon contributes to biofilm formation. Finally, in planta the mutant produced delayed and less severe disease, indicating that the Pil-Chp operon contributes to the virulence of X. fastidiosa, presumably through its role in twitching motility.

Road and Traffic Signs Recognition using Vector Machines

Intelligent Transportation System (ITS) is a system that builds a safe, effective and integrated transportation environment based on advanced technologies. Road signs detection and recognition is an important part of ITS, which offer ways to collect the real time traffic data for processing at a central facility.This project is to implement a road sign recognition model based on AI and image analysis technologies, which applies a machine learning method, Support Vector Machines, to recognize road signs. We focus on recognizing seven categories of road sign shapes and five categories of speed limit signs. Two kinds of features, binary image and Zernike moments, are used for representing the data to the SVM for training and test. We compared and analyzed the performances of SVM recognition model using different features and different kernels. Moreover, the performances using different recognition models, SVM and Fuzzy ARTMAP, are observed.

Power Management Controls

In an office building in the US the office equipment uses about 7% of the total electricity use. Eventhough this is a low number, there is still a reason to save more energy, especially since one third of theenergy savings are lost when power management is not enabled.The core in the project ”Power Management Controls” is to develop a voluntary standard, The UserInterface Standard, that manufacturers of office equipment can use as a reference when they developnew equipment and design new interfaces. The interface is an important part of the use of powermanagement and doing this should increase the use of power management and save more energy. Theinterfaces are ofter hidden or inconsistent and confusing, which makes it harder for the user tounderstand power management. A more consistent interface makes it easier for the user to understandthe meaning of an interface and power management itselfThe standard consists of six different parts, which describe what can be done to achieve a consistentinterface. The standard also describes the part of the project called Dynamic Behavior. This part isconcentrating on the interfaces and the behavior of the device over time, which is important for the userto understand.The purpose of this degree project is to study and participate in the project ”Power ManagementControls”, and to understand what is being done to save more energy.

Color Segmentation using LVQ-Learning Vector Quantization

This thesis aims to present a color segmentation approach for traffic sign recognition based on LVQ neural networks. The RGB images were converted into HSV color space, and segmented using LVQ depending on the hue and saturation values of each pixel in the HSV color space. LVQ neural network was used to segment red, blue and yellow colors on the road and traffic signs to detect and recognize them. LVQ was effectively applied to 536 sampled images taken from different countries in different conditions with 89% accuracy and the execution time of each image among 31 images was calculated in between 0.726sec to 0.844sec. The method was tested in different environmental conditions and LVQ showed its capacity to reasonably segment color despite remarkable illumination differences. The results showed high robustness.

Locus of Controls inverkan på strategier för konflikthantering i kärleksrelationer

Syftet var att undersöka om personers Locus of Control påverkar deras konflikthanteringsstrategier i kärleksrelationer. Hypoteserna var: H1: Det finns skillnader mellan personer med extern och intern Locus of Control när det gäller vilka typer av strategier de använder för att hantera kärleksrelationsproblem. H2: Personer med intern Locus of Control har större benägenhet att använda sig av relationskonflikthanteringsstrategin Kompromiss i sina kärleksrelationer. H3: Kvinnor påverkas mer än män av Locus of Control när det gäller vilka typer av strategier de använder för att hantera kärleksrelationsproblem. Deltagarna var 172 studenter. Data insamlades med en webbenkät innehållande svenska versioner av Rotters I-E skala och RPCS. Resultatet visade en signifikant skillnad mellan personer med intern och extern Locus of Control i tre konflikthanteringsstrategier: Kompromiss, Underkastelse och Undvikande. Vidare använde personer med intern Locus of Control Kompromiss mer. Personer med extern Locus of Control använde Underkastelse och Undvikande mer. Skillnaderna fanns bara bland kvinnorna. Samtliga hypoteser kunde därmed godtas. Undersökningen visade att Locus of Control är en möjlig anledning till varför personer använder sig av vissa typer av konflikthanteringsstrategier i kärleksrelationer.

Testing common nonlinear features in vector nonlinear autoregressive models

This paper studies a special class of vector smooth-transition autoregressive (VSTAR) models that contains common nonlinear features (CNFs), for which we proposed a triangular representation and developed a procedure of testing CNFs in a VSTAR model. We first test a unit root against a stable STAR process for each individual time series and then examine whether CNFs exist in the system by Lagrange Multiplier (LM) test if unit root is rejected in the first step. The LM test has standard Chi-squared asymptotic distribution. The critical values of our unit root tests and small-sample properties of the F form of our LM test are studied by Monte Carlo simulations. We illustrate how to test and model CNFs using the monthly growth of consumption and income data of United States (1985:1 to 2011:11).