Surface roughness noise prediction for silent aircraft experimental design SAX-40

Surface roughness noise is a potentially important contributor to airframe noise. In this paper, noise assessment due to surface roughness is performed for a conceptual Silent Aircraft design SAX-40 by means of a prediction model developed in previous theoretical work and validated experimentally. Estimates of three idealized test cases show that surface roughness could produce a significant noise level above that due to the trailing edge at high frequencies. Roughness height and roughness density are the two most significant parameters influencing surface roughness noise, with roughness height having the dominant effect. The ratio of roughness height to boundary-layer thickness is the relevant non-dimensional parameter and this decreases in the streamwise direction. The candidate surface roughness is selected for SAX-40 to meet an aggressive noise target and keep surface roughness noise at a negligible level. Copyright © 2008 by Yu Liu and Ann P. Dowling.

The calculation of acoustic shielding of engine noise by the Silent Aircraft Airframe

The Silent Aircraft airframe has a flying wing design with a large wing planform and a propulsion system embedded in the rear of the airframe with intake on the upper surface of the wing. In the present paper, boundary element calculations are presented to evaluate acoustic shielding at low frequencies. Besides the three-dimensional geometry of the Silent Aircraft airframe, a few two-dimensional problems are considered that provide some physical insight into the shielding calculations. Mean flow refraction effects due to forward flight motion are accounted for by a simple time transformation that decouples the mean-flow and acoustic-field calculations. It is shown that significant amount of shielding can be obtained in the shadow region where there is no direct line of sight between the source and observer. The boundary element solutions are restricted to low frequencies. We have used a simple physically-based model to extend the solution to higher frequencies. Based on this model, using a monopole acoustic source, we predict at least an 18 dBA reduction in the overall sound pressure level of forward-propagating fan noise because of shielding.

Modeling the development of pronunciation in infant speech acquisition.

Pronunciation is an important part of speech acquisition, but little attention has been given to the mechanism or mechanisms by which it develops. Speech sound qualities, for example, have just been assumed to develop by simple imitation. In most accounts this is then assumed to be by acoustic matching, with the infant comparing his output to that of his caregiver. There are theoretical and empirical problems with both of these assumptions, and we present a computational model- Elija-that does not learn to pronounce speech sounds this way. Elija starts by exploring the sound making capabilities of his vocal apparatus. Then he uses the natural responses he gets from a caregiver to learn equivalence relations between his vocal actions and his caregiver's speech. We show that Elija progresses from a babbling stage to learning the names of objects. This demonstrates the viability of a non-imitative mechanism in learning to pronounce.

SUB-BAND CODING TECHNIQUES FOR ROBUST 4. 8 TO 8 kb/s SPEECH CODERS.

This paper describes a speech coding technique that has been developed in order to provide a method of digitising speech at bit rates in the range 4. 8 to 8 kb/s, that is insensitive to the effects of acoustic background noise and bit errors on the digital link. The main aim has been to develop a coding scheme which provides speech quality and robustness against noise and errors that is similar to a 16000 b/s continuously variable slope delta (CVSD) coder, but which operates at half its data rate or less. A desirable aim was to keep the complexity of the coding scheme within the scope of what could reasonably be handled by current signal processing chips or by a single custom integrated circuit. Applications areas include mobile radio and small Satcomms terminals.

A new, fast response CO2 sensor

The inability of emissions reduction methods to meet upcoming legislation without an unacceptable increase in vehicle cost is a major problem of automobile manufacturer. This work aims to develop a cost-effective reduction of automobile emissions. A prototype CO2 sensor with 5 msec response time was built and bench tested, then used on an engine. The sensor design was based on standard emissions measurement technology using non-dispersive IR absorption. An improved sensor has now been completed with significant improvements in terms of signal to noise ratio and long-term stability. The improved sensor will be used to measure CO2 concentrations on three different engines. The results will then be used to validate engine and catalyst models and to propose control strategies aimed at reducing overall emissions. A brief description of the sensor itself was presented. Original is an abstract.

Defect and disorder reduction by annealing in hydrogenated tetrahedral amorphous carbon

Hydrogenated tetrahedral amorphous carbon (ta-C:H) is a form of diamond-like carbon with a high sp3 content (>60%), grown here using a plasma beam source. Information on the behaviour of hydrogen upon annealing is obtained from effusion measurements, which show that hydrogen does not effuse significantly at temperatures less than 500 °C in films grown using methane and 700 °C in films grown using acetylene. Raman measurements show no significant structural changes at temperatures up to 300 °C. At higher temperatures, corresponding to the onset of effusion, the Raman spectra show a clustering of the sp2 phase. The density of states of ta-C:H is directly measured using scanning tunnelling spectroscopy. The measured gradients of the conduction and valence band tails increase up to 300 °C, confirming the occurrence of band tail sharpening. Examination of the photoluminescence background in the Raman spectra shows an increase in photoluminescence intensity with decreasing defect density, providing evidence that paramagnetic defects are the dominant non-radiative recombination centres in ta-C:H.

The modified Kanerva model for automatic speech recognition

A parallel processing network derived from Kanerva's associative memory theory Kanerva 1984 is shown to be able to train rapidly on connected speech data and recognize further speech data with a label error rate of 0·68%. This modified Kanerva model can be trained substantially faster than other networks with comparable pattern discrimination properties. Kanerva presented his theory of a self-propagating search in 1984, and showed theoretically that large-scale versions of his model would have powerful pattern matching properties. This paper describes how the design for the modified Kanerva model is derived from Kanerva's original theory. Several designs are tested to discover which form may be implemented fastest while still maintaining versatile recognition performance. A method is developed to deal with the time varying nature of the speech signal by recognizing static patterns together with a fixed quantity of contextual information. In order to recognize speech features in different contexts it is necessary for a network to be able to model disjoint pattern classes. This type of modelling cannot be performed by a single layer of links. Network research was once held back by the inability of single-layer networks to solve this sort of problem, and the lack of a training algorithm for multi-layer networks. Rumelhart, Hinton & Williams 1985 provided one solution by demonstrating the "back propagation" training algorithm for multi-layer networks. A second alternative is used in the modified Kanerva model. A non-linear fixed transformation maps the pattern space into a space of higher dimensionality in which the speech features are linearly separable. A single-layer network may then be used to perform the recognition. The advantage of this solution over the other using multi-layer networks lies in the greater power and speed of the single-layer network training algorithm. © 1989.

On computing the physical sources of jet noise

We derive a closed system of equations that relates the acoustically radiating flow variables to the sources of sound for homentropic flows. We use radiating density, momentum density and modified pressure as the dependent variables which leads to simple source terms for the momentum equations. The source terms involve the non-radiating parts of the density and momentum density fields. These non-radiating components are obtained by removing the radiating wavenumbers in the Fourier domain. We demonstrate the usefulness of this new technique on an axi-symmetric jet solution of the Navier-Stokes equations, obtained by direct numerical simulation (DNS). The dominant source term is proportional to the square of the non-radiating part of the axial momentum density. We compare the sound sources to that obtained by an acoustic analogy and find that they have more realistic physical properties. Their frequency content and amplitudes are consistent with. We validate the sources by computing the radiating sound field and comparing it to the DNS solution. © 2010 by S. Sinayoko, A. Agarwal.

Factor analysis based VTS and JUD noise estimation and compensation

Model based compensation schemes are a powerful approach for noise robust speech recognition. Recently there have been a number of investigations into adaptive training, and estimating the noise models used for model adaptation. This paper examines the use of EM-based schemes for both canonical models and noise estimation, including discriminative adaptive training. One issue that arises when estimating the noise model is a mismatch between the noise estimation approximation and final model compensation scheme. This paper proposes FA-style compensation where this mismatch is eliminated, though at the expense of a sensitivity to the initial noise estimates. EM-based discriminative adaptive training is evaluated on in-car and Aurora4 tasks. FA-style compensation is then evaluated in an incremental mode on the in-car task. © 2011 IEEE.

Speaker and noise factorisation on the AURORA4 task

For many realistic scenarios, there are multiple factors that affect the clean speech signal. In this work approaches to handling two such factors, speaker and background noise differences, simultaneously are described. A new adaptation scheme is proposed. Here the acoustic models are first adapted to the target speaker via an MLLR transform. This is followed by adaptation to the target noise environment via model-based vector Taylor series (VTS) compensation. These speaker and noise transforms are jointly estimated, using maximum likelihood. Experiments on the AURORA4 task demonstrate that this adaptation scheme provides improved performance over VTS-based noise adaptation. In addition, this framework enables the speech and noise to be factorised, allowing the speaker transform estimated in one noise condition to be successfully used in a different noise condition. © 2011 IEEE.

Model-based approaches to handling additive noise in reverberant environments

Model-based approaches to handle additive and convolutional noise have been extensively investigated and used. However, the application of these schemes to handling reverberant noise has received less attention. This paper examines the extension of two standard additive/convolutional noise approaches to handling reverberant noise. The first is an extension of vector Taylor series (VTS) compensation, reverberant VTS, where a mismatch function including reverberant noise is used. The second scheme modifies constrained MLLR to allow a wide-span of frames to be taken into account and projected into the required dimensionality. To allow additive noise to be handled, both these schemes are combined with standard VTS. The approaches are evaluated and compared on two tasks, MC-WSJ-AV, and a reverberant simulated version of AURORA-4. © 2011 IEEE.

Derivative Kernels for Noise Robust ASR

Recently there has been interest in combined gen- erative/discriminative classifiers. In these classifiers features for the discriminative models are derived from generative kernels. One advantage of using generative kernels is that systematic approaches exist how to introduce complex dependencies beyond conditional independence assumptions. Furthermore, by using generative kernels model-based compensation/adaptation tech- niques can be applied to make discriminative models robust to noise/speaker conditions. This paper extends previous work with combined generative/discriminative classifiers in several directions. First, it introduces derivative kernels based on context- dependent generative models. Second, it describes how derivative kernels can be incorporated in continuous discriminative models. Third, it addresses the issues associated with large number of classes and parameters when context-dependent models and high- dimensional features of derivative kernels are used. The approach is evaluated on two noise-corrupted tasks: small vocabulary AURORA 2 and medium-to-large vocabulary AURORA 4 task.