Non-Linear and Non-Conventional Speech Processing: Alternative Techniques

This special issue aims to cover some problems related to non-linear and nonconventional speech processing. The origin of this volume is in the ISCA Tutorial and Research Workshop on Non-Linear Speech Processing, NOLISP’09, held at the Universitat de Vic (Catalonia, Spain) on June 25–27, 2009. The series of NOLISP workshops started in 2003 has become a biannual event whose aim is to discuss alternative techniques for speech processing that, in a sense, do not fit into mainstream approaches. A selected choice of papers based on the presentations delivered at NOLISP’09 has given rise to this issue of Cognitive Computation.

Blind channel deconvolution of real world signals using source separation techniques

In this paper we present a method for blind deconvolution of linear channels based on source separation techniques, for real word signals. This technique applied to blind deconvolution problems is based in exploiting not the spatial independence between signals but the temporal independence between samples of the signal. Our objective is to minimize the mutual information between samples of the output in order to retrieve the original signal. In order to make use of use this idea the input signal must be a non-Gaussian i.i.d. signal. Because most real world signals do not have this i.i.d. nature, we will need to preprocess the original signal before the transmission into the channel. Likewise we should assure that the transmitted signal has non-Gaussian statistics in order to achieve the correct function of the algorithm. The strategy used for this preprocessing will be presented in this paper. If the receiver has the inverse of the preprocess, the original signal can be reconstructed without the convolutive distortion.

Understanding diffusion MR imaging techniques: from scalar diffusion-weighted imaging to diffusion tensor imaging and beyond.

The complex structural organization of the white matter of the brain can be depicted in vivo in great detail with advanced diffusion magnetic resonance (MR) imaging schemes. Diffusion MR imaging techniques are increasingly varied, from the simplest and most commonly used technique-the mapping of apparent diffusion coefficient values-to the more complex, such as diffusion tensor imaging, q-ball imaging, diffusion spectrum imaging, and tractography. The type of structural information obtained differs according to the technique used. To fully understand how diffusion MR imaging works, it is helpful to be familiar with the physical principles of water diffusion in the brain and the conceptual basis of each imaging technique. Knowledge of the technique-specific requirements with regard to hardware and acquisition time, as well as the advantages, limitations, and potential interpretation pitfalls of each technique, is especially useful.

Validation of tissue modelization and classification techniques in T1-weighted MR brain images

We propose a deep study on tissue modelization andclassification Techniques on T1-weighted MR images. Threeapproaches have been taken into account to perform thisvalidation study. Two of them are based on FiniteGaussian Mixture (FGM) model. The first one consists onlyin pure gaussian distributions (FGM-EM). The second oneuses a different model for partial volume (PV) (FGM-GA).The third one is based on a Hidden Markov Random Field(HMRF) model. All methods have been tested on a DigitalBrain Phantom image considered as the ground truth. Noiseand intensity non-uniformities have been added tosimulate real image conditions. Also the effect of ananisotropic filter is considered. Results demonstratethat methods relying in both intensity and spatialinformation are in general more robust to noise andinhomogeneities. However, in some cases there is nosignificant differences between all presented methods.

Techniques for Estimating Flood-Frequency Discharges for Streams in Iowa, HR-395A, 2001

A statewide study was conducted to develop regression equations for estimating flood-frequency discharges for ungaged stream sites in Iowa. Thirty-eight selected basin characteristics were quantified and flood-frequency analyses were computed for 291 streamflow-gaging stations in Iowa and adjacent States. A generalized-skew-coefficient analysis was conducted to determine whether generalized skew coefficients could be improved for Iowa. Station skew coefficients were computed for 239 gaging stations in Iowa and adjacent States, and an isoline map of generalized-skew-coefficient values was developed for Iowa using variogram modeling and kriging methods. The skew map provided the lowest mean square error for the generalized-skew- coefficient analysis and was used to revise generalized skew coefficients for flood-frequency analyses for gaging stations in Iowa. Regional regression analysis, using generalized least-squares regression and data from 241 gaging stations, was used to develop equations for three hydrologic regions defined for the State. The regression equations can be used to estimate flood discharges that have recurrence intervals of 2, 5, 10, 25, 50, 100, 200, and 500 years for ungaged stream sites in Iowa. One-variable equations were developed for each of the three regions and multi-variable equations were developed for two of the regions. Two sets of equations are presented for two of the regions because one-variable equations are considered easy for users to apply and the predictive accuracies of multi-variable equations are greater. Standard error of prediction for the one-variable equations ranges from about 34 to 45 percent and for the multi-variable equations range from about 31 to 42 percent. A region-of-influence regression method was also investigated for estimating flood-frequency discharges for ungaged stream sites in Iowa. A comparison of regional and region-of-influence regression methods, based on ease of application and root mean square errors, determined the regional regression method to be the better estimation method for Iowa. Techniques for estimating flood-frequency discharges for streams in Iowa are presented for determining ( 1) regional regression estimates for ungaged sites on ungaged streams; (2) weighted estimates for gaged sites; and (3) weighted estimates for ungaged sites on gaged streams. The technique for determining regional regression estimates for ungaged sites on ungaged streams requires determining which of four possible examples applies to the location of the stream site and its basin. Illustrations for determining which example applies to an ungaged stream site and for applying both the one-variable and multi-variable regression equations are provided for the estimation techniques.

The Use of Remote Sensing Techniques Compared to Traditional Archaeological Phase II Testing Methods: A Cost-Benefit Analysis, HR-524, 1986

Remote sensing was utilized in the Phase II Cultural Resources Investigation for this project in lieu of extensive excavations. The purpose of the present report is to compare the costs and benefits of the use of remote sensing to the hypothetical use of traditional excavation methods for this project. Estimates for this hypothetical situation are based on the project archaeologist's considerable past experience in conducting similar investigations. Only that part of the Phase II investigation involving field investigations is addressed in this report. Costs for literature review, laboratory analysis, report preparation, etc., are not included. The project manager proposed the use of this technique for the fol lowing logistic, safety and budgetary reasons.

Pressure Relief and Other Joint Rehabilitation Techniques, HR-1042, 1987

A study of four major concrete pavement joint rehabilitation techniques has been conducted, including: pressure relief joints, full-depth repairs, partial-depth repairs and joint resealing. The products of this research include the following for each technique: a summary of published research, detailed documentation of the design and performance of the 36 projects, conclusions and recommendations of the state highway engineers panel, "Design and Construction Guidelines" and "Guide Specifications." The latter two products are prepared for use by state highway agencies. The results of this study are based upon a review of literature, extensive field surveys and analysis of 36 rehabilitation projects, and the experience of an expert panel of state highway engineers.

Reduction of Concrete Deterioration by Ettringite Using Crystal Growth Inhibition Techniques: Part II Field Evaluation of Inhibitor Effectiveness, TR-469, 2004

The effects of diethylenetriaminpenta(methylenephosphonic acid) (DTPMP), a phosphonate inhibitor, on the growth of delayed ettringite have been evaluated using concrete in highway US 20 near Williams, Iowa, and the cores of six highways subject to moderate (built in 1992) or minor (built in 1997) deterioration. Application of 0.01 and 0.1 vol. % DTPMP to cores was made on a weekly or monthly basis for one year under controlled laboratory-based freeze-thaw and wet-dry conditions over a temperature range of -15 degrees to 58 degrees C to mimic extremes in Iowa roadway conditions. The same concentrations of phosphonate were also applied to cores left outside (roof of Science I at Iowa State University) over the same period of time. Nineteen applications of 0.1 vol. % DTPMP with added deicing salt solution (about 23 weight % NACL) were made to US 20 during the winters of 2003 and 2004. In untreated samples, air voids, pores, and occasional cracks are lined with acicular ettringite crystals (up to 50 micrometers in length) whereas air voids, pores, and cracks in concrete from the westbound lane of US 20 are devoid of ettringite up to a depth of about 0.5 mm from the surface of the concrete. Ettringite is also absent in zones up to 6 mm from the surface of concrete slabs placed on the roof of Science I and cores subject to laboratory-based freeze-thaw experiments. In these zones, the relatively high concentration of DTPMP caused it to behave as a chelator. Stunted ettringite crystals 5 to 25 micrometers in length, occasionally coated with porlandite, form on the margins of these zones indicating that in these areas DTPMP behaved as an inhibitor due to a reduction in the concentration of phosphonate. Analyses of mixes of ettringite and DTPMP using electrospray mass spectrometry suggests that the stunting of ettringite growth is caused by the adsorption of a Ca2+ ion and a water molecule to deprotonated DTPMP on the surface of the {0001} face of ettringite. It is anticipated that by using a DTPMP concentration of between 0.001 and 0.01 vol. % for the extended life of a highway (i.e. >20 years), deterioration caused by the expansive growth of ettringite will be markedly reduced.

Investigation of Improved Utility Cut Repair Techniques to Reduce Settlement in Repaired Areas, Phase II, TR-566, 2010

Trench maintenance problems are caused by improper backfill placement and construction procedures. This report is part of a multiphase research project that aims to improve long-term performance of utility cut restoration trenches. The goal of this research is to improve pavement patch life and reduce maintenance of the repaired areas. The objectives were to use field-testing data, laboratory-testing data, and long-term monitoring (elevation survey and falling weight deflectometer testing) to suggest and modify recommendations from Phase I and to identify the principles of trench subsurface settlement and load distribution in utility cut restoration areas by using instrumented trenches. The objectives were accomplished by monitoring local agency utility construction from Phase I, constructing and monitoring the recommended trenches from Phase I, and instrumenting trenches to monitor changes in temperature, pressure, moisture content, and settlement as a function of time to determine the influences of seasonal changes on the utility cut performance.

Identification of Laboratory Techniques to Optimize Superpave HMA Surface Friction Characteristics, TR-450, 2010

Wet pavement friction is known to be one of the most important roadway safety parameters. In this research, frictional properties of flexible (asphalt) pavements were investigated. As a part of this study, a laboratory device to polish asphalt specimens was refined and a procedure to evaluate mixture frictional properties was proposed. Following this procedure, 46 different Superpave mixtures, one stone matrix asphalt (SMA) mixture and one porous friction course (PFC) mixture were tested. In addition, 23 different asphalt and two concrete field sections were also tested for friction and noise. The results of both field and laboratory measurements were used to develop an International Friction Index (IFI)-based protocol for measurement of the frictional characteristics of asphalt pavements for laboratory friction measurements. Based on the results of the study, it appears the content of high friction aggregate should be 20% or more of the total aggregate blend when used with other, polish susceptible coarse aggregates; the frictional properties increased substantially as the friction aggregate content increased above 20%. Both steel slag and quartzite were found to improve the frictional properties of the blend, though steel slag had a lower polishing rate. In general, mixes containing soft limestone demonstrated lower friction values than comparable mixes with hard limestone or dolomite. Larger nominal maximum aggregate size mixes had better overall frictional performance than smaller sized mixes. In addition, mixes with higher fineness moduli generally had higher macrotexture and friction.

Similarity in food cleaning techniques within matrilines in wild vervet monkeys.

Social learning and the formation of traditions rely on the ability and willingness to copy one another. A central question is under which conditions individuals adapt behaviour to social influences. Here, we demonstrate that similarities in food processing techniques emerge on the level of matrilines (mother-offspring) but not on the group level in an experiment on six groups of wild vervet monkeys that involved grapes covered with sand. Monkeys regularly ate unclean grapes but also used four cleaning techniques more similarly within matrilines: rubbing in hands, rubbing on substrate, open with mouth, and open with hands. Individual cleaning techniques evolved over time as they converged within matrilines, stabilised at the end and remained stable in a follow-up session more than one year later. The similarity within matrilines persisted when we analyzed only foraging events of individuals in the absence of other matriline members and matriline members used more similar methods than adult full sisters. Thus, momentary conversion or purely genetic causation are unlikely explanations, favouring social learning as mechanism for within matriline similarities. The restriction of traditions to matriline membership rather than to the group level may restrict the development of culture in monkeys relative to apes or humans.

Laser treatment of 13 benign oral vascular lesions by three different surgical techniques

Objectives: Benign Oral Vascular Lesions (BOVLs) are a group of vascular diseases characterized by congenital, inflammatory or neoplastic vascular dilations clinically evidenced as more or less wide masses of commonly dark bluish color. If traumatized BOVLs are characterized by a great risk of hemorrhage and their treatment usually requires great caution to prevent massive bleeding. In the last decades lasers have dramatically changed the way of treatment of BOVLs permitting the application of even peculiar techniques that gave interesting advantages in their management reducing hemorrhage risks. The aim of this study was to evaluate the capabilities and disadvantages of three laser assisted techniques in the management of BOVLs. Study design: In this study 13 BOVLs were treated by three different laser techniques: the traditional excisional biopsy (EB), and two less invasive techniques, the transmucosal thermocoagulation (TMT) and the intralesional photocoagulation (ILP). Two different laser devices were adopted in the study: a KTP laser (DEKA, Florence, Italy, 532nm) and a GaAlAs laser (Laser Innovation, Castelgandolfo, Italy, 808nm) selected since their great effectiveness on hemoglobin. Results: In each case, lasers permitted safe treatments of BOVLs without hemorrhages, both during the intervention and in the post-operative period. The minimally invasive techniques (TMT and ILP) permitted even the safe resolution of big lesions without tissue loss. Conclusions: Laser devices confirm to be the gold standard in BOVLs treatment, permitting even the introduction of minimal invasive surgery principles and reducing the risks of hemorrhage typical of these neoplasms. As usual in laser surgery, it is necessary a clear knowledge of the devices and of the laser-tissue interaction to optimize the results reducing risks and disadvantages