Search

The adoption of high-accuracy speech recognition algorithms without an effective evaluation of their impact on the target computational resource is impractical for mobile and embedded systems. In this paper, techniques are adopted to minimise the required computational resource for an effective mobile-based speech recognition system. A Dynamic Multi-Layer Perceptron speech recognition technique, capable of running in real time on a state-of-the-art mobile device, has been introduced. Even though a conventional hidden Markov model when applied to the same dataset slightly outperformed our approach, its processing time is much higher. The Dynamic Multi-layer Perceptron presented here has an accuracy level of 96.94% and runs significantly faster than similar techniques. [ABSTRACT FROM AUTHOR]

In this paper, a speaker verification algorithm, which is inspired by the physiology of hearing in the human auditory system, is proposed. The algorithm uses a modified self-organising map composed of spiking neurons. The architecture of the algorithm imitates the biomechanical mechanism of the human auditory system, which converts the speech signal from sound vibration waves into electrical spikes inside the cochlea. The paper also suggests a spike-based rank order coding input feature vector, which is designed to be representative of the real biological spike trains found within the human auditory nerve. During the training phase, the winner neuron in the proposed Spiking Self-Organising Map is updated only when its activity exceeds a specified threshold. The algorithm is evaluated using 50 speakers from the Centre for Spoken Language Understanding (CSLU2002) speaker verification database and shows a speaker verification performance of 90.1%. This compares favourably with previous non-spiking self-organising map that used Discrete Fourier Transform-based input feature vector with the same dataset. [ABSTRACT FROM AUTHOR]

The Simplified Method as reported in AASHTO and Federal Highway Administration (FHWA) manuals has been demonstrated to give poor predictions of unfactored reinforcement loads and strains, especially for geosynthetic reinforced soil walls. The writers have proposed the K-stiffness Method to improve the load prediction accuracy for walls under working stress (operational) conditions. However, it has also been recognized in recent publications by the writers and others that further improvements to the K-stiffness Method are needed. Furthermore, acceptance of the K-stiffness Method has been hindered due to its perceived complexity and the use of the plane strain friction angle to quantify the strength of the reinforced soil. This paper takes a fresh look at both methods and uses lessons learned from the K-stiffness Method development to improve the accuracy of the AASHTO/FHWA Simplified Method. Key parameters introduced during the development of the K-stiffness Method are applied to the Simplified Method and updated to further improve load prediction accuracy. Additional wall case histories have been added to the database used for the original K-stiffness Method to calibrate the new model and to broaden its utility. An important improvement is a single model that allows for seamless load prediction across a range of walls constructed with relatively extensible geosynthetic reinforcement and inextensible steel reinforcement materials. The quantitative improvement of the new model (Simplified Stiffness Method) compared to the current AASHTO/FHWA Simplified Method is demonstrated through statistical analysis of load bias values (i.e., the ratio of measured to predicted reinforcement load). [ABSTRACT FROM AUTHOR]

This paper presents a multi-level speaker verification system that uses 64 discrete Fourier transform spectrum components as input feature vectors. A speech activity detection technique is used as a pre-processing stage to identify vowel phoneme boundaries within a speech sample. A modified self-organising map (SOM) is then used to filter the speech data by using cluster information extracted from three vowels for a claimed speaker. This SOM filtering stage also provides coarse speaker verification. Finally, a second speaker verification level of three multi-layer perceptron networks classifies the filtered frames provided by the SOMs. These multi-layer perceptrons work as fine-grained vowel-based speaker verifiers. The proposed verification algorithm shows a performance of 94.54% when evaluated using 50 speakers from the Centre for Spoken Language Understanding speaker verification database. In addition, it is shown that the novel discrete Fourier transform spectrum-based linear correlation pre-processing technique, presented here, provides the system with greater robustness against changes in speech volume levels when compared with an equivalent energy frame analysis. [ABSTRACT FROM AUTHOR]

A high-density polyethylene (HDPE) geogrid soil-reinforced dry-cast concrete block retaining wall 6.3-m high was designed using the K-stiffness method as part of a highway-widening project southeast of Seattle, Washington. The amount of reinforcement needed for the original wall design using the K-stiffness method was approximately 50% of that required using the AASHTO simplified method. This paper describes the construction, instrumentation program, and interpretation of the measurements. Geogrid strains were measured using strain gauges and extensometers attached to reinforcement layers. An extensive materials testing program was conducted to characterize the backfill soil properties and geogrid stiffness properties and to calibrate strain gauge readings. The reinforcement loads deduced from the measured strains are compared with Class A, B, and C1 predictions using the AASHTO simplified and K-stiffness methods. These comparisons demonstrate that the simplified method significantly overestimated reinforcement loads, whereas the K-stiffness method provided estimates that were consistent with the measured results. This paper describes lessons learned, the influence of construction activities on wall performance, and the limitations of both methods in estimating connections loads. [ABSTRACT FROM AUTHOR]

The grade raising associated with the Third Runway Project at Seattle-Tacoma International Airport included construction of tall mechanically stabilized earth (MSE) walls, including the near-vertical, two-tier, 26-m North MSE wall and the near-vertical, four-tier, 46-m tall west MSE wall. Twenty reinforcement strips at critical wall cross sections were instrumented with over 500 strain gauges to monitor strains during and following construction. The reinforcement loads inferred from observed strains are of interest because of their great height and global reinforcement stiffness, which place these walls outside the range in height and stiffness used to calibrate commonly used design methods. This paper presents the development and distribution of reinforcement strains measured during and following the construction of these walls. The reinforcement stresses calculated using the original reinforcement load design methods and design friction angle agreed with those inferred from the measured strains. The accuracy of two standard-of-practice and two alternate design methods is evaluated by comparing the reinforcement loads inferred from measured strains to those calculated using the actual friction angle of the reinforced fill material. Advantages and limitations in these design methods are identified, and recommendations for the improvement of some of these methods are provided. [ABSTRACT FROM AUTHOR]

Silog is a biometric authentication system that extends the conventional PC logon process using voice verification. Users enter their ID and password using a conventional Windows logon procedure but then the biometric authentication stage makes a voice over IP (VoIP) call to a VoiceXML (VXML) server. User interaction with this speech-enabled component then allows the user’s voice characteristics to be extracted as part of a simple user/system spoken dialogue. If the captured voice characteristics match those of a previously registered voice profile, then network access is granted. If no match is possible, then a potential unauthorised system access has been detected and the logon process is aborted. [Copyright &y& Elsevier]

Abstract: The integration of Global Positioning System (GPS) technology into mobile devices provides them with an awareness of their physical location. This geospatial context can be employed in a wide range of applications including locating nearby places of interest as well as guiding emergency services to incidents. In this research, a GPS-enabled Personal Digital Assistant (PDA) is used to create a computerised tree inventory for the management of arboriculture. Using the General Packet Radio Service (GPRS), GPS information and arboreal image data are sent to a web-server. An office-based PC running customised Geographical Information Software (GIS) then automatically retrieves the GPS tagged image data for display and analysis purposes. The resulting application allows an expert user to view the condition of individual trees in greater detail than is possible using remotely sensed imagery. [Copyright &y& Elsevier]

Abstract: The paper provides a synthesis of work by the writers that has the objective of developing a new working stress method for the calculation of reinforcement loads in geosynthetic reinforced soil walls. As a precursor to this objective, careful back-analyses of a database of instrumented and monitored full-scale field and laboratory walls are used to demonstrate that the current American Association of State Highway and Transportation Officials (AASHTO) Simplified Method used in North America results in excessively conservative estimates of the volume of reinforcement required to generate satisfactory long-term wall performance. The new design method captures the essential contributions of the different wall components and properties to reinforcement loads. The method is calibrated against measured in situ wall reinforcement loads using a careful interpretation of reinforcement strains and the conversion of strain to load using a suitably selected reinforcement stiffness value. A novel feature of the method is to design the wall reinforcement so that the soil within the wall backfill is prevented from reaching a failure limit state, consistent with the notion of working stress conditions. [Copyright &y& Elsevier]

A rigorous probabilistic-based approach for load and resistance factor design (LRFD) calibration of internal stability tensile rupture and pullout limit states for geogrid reinforced mechanically stabilized earth (MSE) walls is presented. Load and resistance factors are provided for different combinations of the load model with pullout model and the tensile strength model used in current practice. The load models are the AASHTO Simplified Method and the Simplified Stiffness Method. Resistance factors are tabulated for use in LRFD foundation practice in Canada which rewards designers with larger resistance factors for greater project level of understanding. Resistance factors are also provided that are amenable to US practice where the notion of level of understanding is not explicitly considered in LRFD practice. The paper demonstrates the application of the load and resistance factors for tensile rupture and pullout internal stability limit states using a design example and shows the reduction in required reinforcement strength using the Simplified Stiffness Method. The paper is of value to practitioners and LRFD code writers in the United States and Canada for future revisions of design specifications in their jurisdictions. [ABSTRACT FROM AUTHOR]

The paper demonstrates load and resistance factor design (LRFD) calibration for tensile strength and pullout strength limit states for steel strip mechanically stabilized earth (MSE) walls using a reliability theory-based approach. The analyses adopt load models found in current North American LRFD specifications in combination with a coupled mechanical-corrosion model for the reinforcement tensile strength and pullout models found in design guidelines and in the research literature. Resistance factors for prescribed load factors and different load and resistance model combinations are compiled in tables. The limitations of the LRFD approach to generate resistance factors that are judged to be acceptable based on current practice are demonstrated. The challenges to generate consistent design outcomes with the same margin of safety expressed probabilistically are identified, and the link to reliability-based design is described. Example designs are provided to show the impact on steel reinforcement quantities and length when using different models in combination with proposed and current specified load and resistance factors. The calibration outcomes will be of value to code writers in Canada and the US for MSE wall design sections in the next editions of national geotechnical foundation LRFD codes. [ABSTRACT FROM AUTHOR]

The paper demonstrates deterministic and reliability-based assessment of strength limit states (tensile resistance and pullout) and the service limit state for soil failure for mechanically stabilized earth (MSE) walls constructed with polyester (PET) strap reinforcement. The general approach considers the accuracy of the load and resistance models that appear in each limit state equation plus uncertainty in the estimate of nominal load and resistance values at time of design. Reliability index is computed using a closed-form solution that is easily implemented in a spreadsheet. Three PET strap MSE wall case studies are used to demonstrate the reliability-based assessment approach and to compare margins of safety using different load and resistance model combinations. In some walls using the Coherent Gravity Method to compute loads, the recommended nominal factors of safety for tensile strength and pullout limit states were not satisfied. However, reliability analyses showed that the walls satisfy recommended minimum target reliability index values for the limit states investigated, usually by large amounts. The most critical limit state is the soil failure limit state which is used in the Simplified Stiffness Method to keep the reinforced soil zone at working stress conditions assumed for geosynthetic MSE walls under operational conditions. • Describes deterministic and reliability-based analyses of internal limit states for PET strap walls. • Reliability index is computed using a closed-form solution that is easily implemented in a spreadsheet. • Three PET strap MSE wall case studies are used to demonstrate the reliability-based assessment approach. • Most critical limit state is the soil failure limit state which is used in the Simplified Stiffness Method. • General approach is applicable to MSE walls with extensible and inextensible reinforcement materials. [ABSTRACT FROM AUTHOR]

The paper reports the details of numerical models used to predict the performance of two 3.6 m-high well-instrumented wrapped-face walls. The walls were nominally identical except that the reinforcement material in one wall was a steel welded wire mesh and in the other a biaxial polypropylene geogrid. The backfill soil was modelled using both linear and nonlinear elastic-plastic constitutive models. A general hyperbolic (nonlinear) axial load-strain-time model was used for the reinforcement. The numerical results show good agreement with measured performance features for the welded wire mesh wrapped-face wall. Agreement between numerical predictions of facing displacements and maximum reinforcement loads was less accurate for the very flexible geogrid wrapped-face wall. The discrepancies are believed to be related to the unusually flexible wrapped face used in the geogrid wall construction. Numerically predicted and measured maximum reinforcement loads are compared to loads using the AASHTO reinforcement strength-based design approach (Simplified Method) and the Simplified Stiffness Method which is an empirical reinforcement stiffness-based method. The paper provides physical test data that can be used to benchmark other numerical models, highlights lessons learned during the development of the models, and identifies reasonable expectations for numerical model accuracy for models of similar complexity used to simulate the performance of mechanically stabilized earth (MSE) wall structures under operational conditions. [ABSTRACT FROM AUTHOR]

The paper reports numerical model details and predictions of the end-of-construction performance for two instrumented and well-documented mechanically stabilized earth (MSE) walls. The walls were constructed as part of the highway SR-18 approach fills for a bridge near Seattle, Washington. The geogrid reinforced block face walls were modeled using a commercially available two-dimensional (2D) finite-difference program. The paper provides details on how material properties were selected from laboratory testing of wall components and how the computer modeling was carried out. The paper shows that predicted wall deformations and reinforcement strains were in reasonable agreement with measured data using both linear elastic-plastic and nonlinear elastic-plastic constitutive models for the soil. The geogrid reinforcement was simulated using a nonlinear load-strain-time secant stiffness model and cable elements. The paper compares numerical predictions of reinforcement loads at end of construction with measured values and predictions using a AASHTO Simplified Method, K-stiffness Method, and Simplified Stiffness Method. The paper is a useful benchmark for modelers because it demonstrates what the authors believe are reasonable expectations of model accuracy for Class C predictions of deformations and reinforcement strains for these types of walls when high-quality project information is available. [ABSTRACT FROM AUTHOR]

The paper reports the results of load and resistance factor design (LRFD) calibration for pullout and yield limit states for steel strip reinforced soil walls under self-weight loading. An important feature of the calibration method is the use of bias statistics to account for prediction accuracy of the underlying deterministic models for reinforcement load, pullout capacity and yield strength of the steel strips, and random variability in input parameters. To improve the accuracy of reinforcement load predictions, small adjustments to current semiempirical American Association of State Highway and Transportation Officials (AASHTO) load design charts are proposed. Similarly, current empirical-based design charts found in AASHTO and Federal Highway Administration (FHWA) guidance documents for the estimation of the pullout resistance factor for smooth and ribbed steel strips are adjusted to improve the accuracy of pullout capacity predictions. The results of calibration lead to a load factor of 1.35 that is consistent with current practice and resistance factors that together give a consistent probability of failure of 1% for all three limit states considered. Furthermore, comparison with allowable stress design (ASD) past practice (AASHTO simplified method) shows that the operational factors of safety using a rigorous LRFD approach give the same or higher factors of safety and lower probabilities of failure. In this study, data for steel strip reinforced soil walls are used as an example to illustrate rigorous reliability theory-based LRFD calibration concepts. However, the general approach is applicable to other reinforced soil wall technologies and calibration outcomes can be updated as more data become available. [ABSTRACT FROM AUTHOR]

The results of load and resistance factor design (LRFD) calibration are reported for the pullout limit state in geogrid reinforced soil walls under self-weight loading and permanent uniform surcharge. Bias statistics are used to account for the prediction accuracy of the underlying deterministic models for load and pullout capacity and the random variability in the input parameters. The paper shows that the current AASHTO simplified method to calculate reinforcement loads under operational conditions is overly conservative leading to poor prediction accuracy of the underlying deterministic model used in LRFD calibration. Refinements to the load and default pullout capacity models in the AASHTO and Federal Highway Administration guidance documents are proposed. These models generate reasonable resistance factors using a load factor of 1.35 and give a consistent probability of pullout failure of 1%. A comparison with the allowable stress design (ASD) past practice shows that the operational factors of safety using a reliability-based LRFD approach give factors of safety greater than 1.5. Regardless of the design approach (ASD or LRFD), the analysis results demonstrate that the current empirical minimum reinforcement length criteria will likely control the design for pullout. [ABSTRACT FROM AUTHOR]

A response is from the authors to the reviewers of their article “Predicted Loads in Steel Reinforced Soil Walls Using the AASHTO Simplified Method” which is in Volume 135 of the February 2009 issue is presented.

Abstract: North American design codes are now committed to the development of load and resistance factor design (LRFD) for reinforced soil structures including internal stability limit states. Reliability-based analysis is required to carry out these calibrations. A framework for LRFD calibration has been proposed by the writers that requires bias statistics for load and resistance terms for each limit state function. In this paper the formulation of the limit state for reinforcement tensile rupture is developed and the component strength-reduction bias statistics identified. The paper describes how to compute bias statistics from project-specific installation damage trials for use in reliability-based design for the reinforcement rupture limit state or using data from multiple sources for LRFD calibration. A database of results from field installation damage trials on 103 different geosynthetic products was collected from 20 different sources. A total of 799 and 2248 in-air tensile test results were reviewed for undamaged and damaged geosynthetic specimens, respectively. This database is used to compute installation damage bias statistics for six different categories of geosynthetic and four categories of backfill soils classified according to the D 50 particle size. A practical outcome is that for analysis purposes, bias statistics can be grouped into two ranges for each geosynthetic type based on D 50 of the soil greater than or less than 19 mm. The paper shows how bias statistics together with load and resistance factors for the geosynthetic rupture limit state function recommended by can be used to calculate probability of failure using Monte Carlo simulation and demonstrates the sensitivity of probability of failure to magnitude of installation damage bias statistics. The installation damage data is valuable for future LRFD calibration to select resistance factors for use in design codes for the geosynthetic rupture limit state in reinforced soil structures. [Copyright &y& Elsevier]

The paper investigates the accuracy of the AASHTO simplified method by using load measurements reported in a large database of full-scale instrumented walls for bar mat, welded wire, and steel strip soil reinforced walls. The accuracy of the AASHTO simplified method is quantified by computing the mean and coefficient of variation of the ratio (bias) of measured loads under operational conditions to predicted loads. The paper shows that for steel strip walls, the AASHTO simplified method is reasonably accurate for granular backfill soils with friction angles less than 45°. For bar mat walls, the method is demonstrated to be slightly conservative. The simplified method underpredicts reinforcement loads at shallow overburden depths for steel strip walls with backfill friction angles greater than 45° due to compaction-related effects. It is concluded that these compaction-induced loads near the wall top do not contribute to internal instability due to pullout. [ABSTRACT FROM AUTHOR]

The paper investigates the accuracy of the coherent gravity method by using measurements reported in a large database of full-scale instrumented walls that was not available at the time the original method was developed. The new database includes data for bar mat, welded wire and steel strip soil reinforced walls. Measured reinforcement loads under operational conditions are compared with predicted values for bar mat and steel strip reinforced walls. The accuracy of the coherent gravity method as presented in the BS 8006 design standard is quantified by computing the mean and coefficient of variation of the ratio (bias) of measured to predicted loads. The paper shows that for steel strip walls the coherent gravity method is reasonably accurate for soils with friction angles less than 458. For granular soils with higher friction angles and bar mat walls, the current coherent gravity method is shown to be less accurate and, on average, non-conservative for design. Modifications to the method as currently described in BS 8006 are proposed to improve the accuracy of the method for bar mat reinforced soil walls and steel strip reinforced soil walls with high friction angle backfill soils. [ABSTRACT FROM AUTHOR]

• Buried steel grids used in mechanically stabilized earth (MSE) walls can corrode. • Probability of failure is computed for limit states of galvanized steel grids. • Three-stage corrosion model is coupled with reinforcement load and resistance models. • Probability of steel grid tensile and pullout failure with time is computed. • Pullout failure of coated steel members due to corrosion is shown to be unlikely. Design for internal limit states of steel grid reinforcement in mechanically stabilized earth (MSE) walls must provide for loss of steel thickness due to corrosion. The paper combines key features of zinc and steel corrosion models used in North American practice with corrosion rate statistics reported in the literature for the zinc coating on buried galvanized steel strips and for steel rebar. Examples of loss of strength and pullout capacity of steel grid elements constructed with different member diameters and up to 120 years after construction are presented in probabilistic terms. Results of reliability-based analysis for the tensile strength and pullout limit state at design lives of 75 and 120 years are demonstrated for a typical wall under operational conditions using tensile load and resistance models in combination with non-aggressive soils recommended in North American design codes. The results of calculations show that for the wall example the influence of different zinc cover thickness on probability of tensile failure of longitudinal members is detectable at the end of design life. However, the typical minimum specified zinc thickness of 0.086 mm recommended in North American practice is likely sufficient to keep probabilities of failure to 1% or less (i.e., reliability index of 2.33 or greater). The influence of a range of initial zinc thickness reported in the literature on probability of pullout failure of the same steel grid elements was found to be negligible. [ABSTRACT FROM AUTHOR]

Background: Oestrogen receptor (ER) positive (luminal) tumours account for the largest proportion of females with breast cancer. Theirs is a heterogeneous disease presenting clinical challenges in managing their treatment. Three main biological luminal groups have been identified but clinically these can be distilled into two prognostic groups in which Luminal A are accorded good prognosis and Luminal B correlate with poor prognosis. Further biomarkers are needed to attain classification consensus. Machine learning approaches like Artificial Neural Networks (ANNs) have been used for classification and identification of biomarkers in breast cancer using high throughput data. In this study, we have used an artificial neural network (ANN) approach to identify DACH1 as a candidate luminal marker and its role in predicting clinical outcome in breast cancer is assessed. Materials and methods: A reiterative ANN approach incorporating a network inferencing algorithm was used to identify ER-associated biomarkers in a publically available cDNA microarray dataset. DACH1 was identified in having a strong influence on ER associated markers and a positive association with ER. Its clinical relevance in predicting breast cancer specific survival was investigated by statistically assessing protein expression levels after immunohistochemistry in a series of unselected breast cancers, formatted as a tissue microarray. Results: Strong nuclear DACH1 staining is more prevalent in tubular and lobular breast cancer. Its expression correlated with ER-alpha positive tumours expressing PgR, epithelial cytokeratins (CK)18/19 and ‘luminal-like’ markers of good prognosis including FOXA1 and RERG (p<0.05). DACH1 is increased in patients showing longer cancer specific survival and disease free interval and reduced metastasis formation (p<0.001). Nuclear DACH1 showed a negative association with markers of aggressive growth and poor prognosis. Conclusion: Nuclear DACH1 expression appears to be a Luminal A biomarker predictive of good prognosis, but is not independent of clinical stage, tumour size, NPI status or systemic therapy. [ABSTRACT FROM AUTHOR]

Background: While alcohol-related health and social problems amongst youths are increasing internationally, both consumption and associated harms are particularly high in British youth. Youth drinking patterns, including bingeing, frequent drinking and drinking in public spaces, are associated with increased risks of acute (e.g. violence) and long-term (e.g. alcohol-dependence) health problems. Here we examine economic, behavioural and demographic factors that predict these risky drinking behaviours among 15-16 year old schoolchildren who consume alcohol. A crosssectional survey was conducted among schoolchildren in North West England (n = 10,271) using an anonymous questionnaire delivered in school settings. Analysis utilised logistic regression to identify independent predictors of risky drinking behaviour. Results: Of all respondents, 87.9% drank alcohol. Of drinkers, 38.0% usually binged when drinking, 24.4% were frequent drinkers and 49.8% drank in public spaces. Binge, frequent and public drinking were strongly related to expendable income and to individuals buying their own alcohol. Obtaining alcohol from friends, older siblings and adults outside shops were also predictors of risky drinking amongst drinkers. However, being bought alcohol by parents was associated with both lower bingeing and drinking in public places. Membership of youth groups/teams was in general protective despite some association with bingeing. Conclusion: Although previous studies have examined predictors of risky drinking, our analyses of access to alcohol and youth income have highlighted eradicating underage alcohol sales and increased understanding of children's spending as key considerations in reducing risky alcohol use. Parental provision of alcohol to children in a family environment may also be important in establishing child-parent dialogues on alcohol and moderating youth consumption. However, this will require supporting parents to ensure they develop only moderate drinking behaviours in their children and only when appropriate. [ABSTRACT FROM AUTHOR]