Efficient post-processing techniques for speech enhancement

We address the problem of speech enhancement in real-world noisy scenarios. We propose to solve the problem in two stages, the first comprising a generalized spectral subtraction technique, followed by a sequence of perceptually-motivated post-processing algorithms. The role of the post-processing algorithms is to compensate for the effects of noise as well as to suppress any artifacts created by the first-stage processing. The key post-processing mechanisms are aimed at suppressing musical noise and to enhance the formant structure of voiced speech as well as to denoise the linear-prediction residual. The parameter values in the techniques are fixed optimally by experimentally evaluating the enhancement performance as a function of the parameters. We used the Carnegie-Mellon university Arctic database for our experiments. We considered three real-world noise types: fan noise, car noise, and motorbike noise. The enhancement performance was evaluated by conducting listening experiments on 12 subjects. The listeners reported a clear improvement (MOS improvement of 0.5 on an average) over the noisy signal in the perceived quality (increase in the mean-opinion score (MOS)) for positive signal-to-noise-ratios (SNRs). For negative SNRs, however, the improvement was found to be marginal.

A risk-estimation-based formulation for speech enhancement and its relation to Wiener filtering

The goal of speech enhancement algorithms is to provide an estimate of clean speech starting from noisy observations. The often-employed cost function is the mean square error (MSE). However, the MSE can never be computed in practice. Therefore, it becomes necessary to find practical alternatives to the MSE. In image denoising problems, the cost function (also referred to as risk) is often replaced by an unbiased estimator. Motivated by this approach, we reformulate the problem of speech enhancement from the perspective of risk minimization. Some recent contributions in risk estimation have employed Stein's unbiased risk estimator (SURE) together with a parametric denoising function, which is a linear expansion of threshold/bases (LET). We show that the first-order case of SURE-LET results in a Wiener-filter type solution if the denoising function is made frequency-dependent. We also provide enhancement results obtained with both techniques and characterize the improvement by means of local as well as global SNR calculations.

A generalized Stein's estimation approach to speech enhancement based on perceptual criteria

We address the problem of speech enhancement using a risk- estimation approach. In particular, we propose the use the Stein’s unbiased risk estimator (SURE) for solving the problem. The need for a suitable finite-sample risk estimator arises because the actual risks invariably depend on the unknown ground truth. We consider the popular mean-squared error (MSE) criterion first, and then compare it against the perceptually-motivated Itakura-Saito (IS) distortion, by deriving unbiased estimators of the corresponding risks. We use a generalized SURE (GSURE) development, recently proposed by Eldar for MSE. We consider dependent observation models from the exponential family with an additive noise model,and derive an unbiased estimator for the risk corresponding to the IS distortion, which is non-quadratic. This serves to address the speech enhancement problem in a more general setting. Experimental results illustrate that the IS metric is efficient in suppressing musical noise, which affects the MSE-enhanced speech. However, in terms of global signal-to-noise ratio (SNR), the minimum MSE solution gives better results.

Single electron transfer-driven multi-dimensional signal read-out function of TCNQ as an ``off-the-shelf'' detector for cyanide

Herein we report the first applications of TCNQ as a rapid and highly sensitive off-the-shelf cyanide detector. As a proof-of-concept, we have applied a kinetically selective single-electron transfer (SET) from cyanide to deep-lying LUMO orbitals of TCNQ to generate a persistently stable radical anion (TCNQ(center dot-)), under ambient condition. In contrast to the known cyanide sensors that operate with limited signal outputs, TCNQ(center dot-) offers a unique multiple signaling platform. The signal readability is facilitated through multichannel absorption in the UV-vis-NIR region and scattering-based spectroscopic methods like Raman spectroscopy and hyper Rayleigh scattering techniques. Particularly notable is the application of the intense 840 nm NIR absorption band to detect cyanide. This can be useful for avoiding background interference in the UV-vis region predominant in biological samples. We also demonstrate the fabrication of a practical electronic device with TCNQ as a detector. The device generates multiorder enhancement in current with cyanide because of the formation of the conductive TCNQ(center dot-).

Extremely High Near Field Enhancement in a Novel Plasmonic Nano Material used for Photovoltage Generation

Enhancement of localized electric field near metal (plasmonic) nanostructures can have various interesting applications in sensing, imaging, photovoltage generation etc., for which significant efforts are aimed towards developing plasmonic systems with well designed and large electromagnetic response. In this paper, we discuss the wafer scale fabrication and optical characterization of a unique three dimensional plasmonic material. The near field enhancement in the visible range of the electromagnetic spectrum obtained in these structures (order of 106), is close to the fundamental limit that can be obtained in this and similar EM field enhancement schemes. The large near field enhancement has been reflected in a huge Raman signal of graphene layer in close proximity to the plasmonic system, which has been validated with FEM simulations. We have integrated graphene photodetectors with this material to obtain record photovoltage generation, with responsivity as high as A/W. As far as we know, this is the highest sensitivity obtained in any plasmonic-graphene hybrid photodetection system till date.

SPEECH ENHANCEMENT USING DISCRIMINATIVE RANDOM FIELDS

Speech enhancement in stationary noise is addressed using the ideal channel selection framework. In order to estimate the binary mask, we propose to classify each time-frequency (T-F) bin of the noisy signal as speech or noise using Discriminative Random Fields (DRF). The DRF function contains two terms - an enhancement function and a smoothing term. On each T-F bin, we propose to use an enhancement function based on likelihood ratio test for speech presence, while Ising model is used as smoothing function for spectro-temporal continuity in the estimated binary mask. The effect of the smoothing function over successive iterations is found to reduce musical noise as opposed to using only enhancement function. The binary mask is inferred from the noisy signal using Iterated Conditional Modes (ICM) algorithm. Sentences from NOIZEUS corpus are evaluated from 0 dB to 15 dB Signal to Noise Ratio (SNR) in 4 kinds of additive noise settings: additive white Gaussian noise, car noise, street noise and pink noise. The reconstructed speech using the proposed technique is evaluated in terms of average segmental SNR, Perceptual Evaluation of Speech Quality (PESQ) and Mean opinion Score (MOS).

Enhancement of fish production in a reservoir after partitioning by dikes through community participation

A reservoir of 70 acres was portioned by dikes into four manageable big ponds to get more production of fishes at Basurhat, Noakhali, Bangladesh under the supervision of local community through a society of 40 people ownership. Pangus (Pangasius hypophthalmus) @ 20,000/acre, and then fry and fingerlings of different types of fishes such as catla (Catla catla), rohu (Labeo rohita), mrigal (Cirrhina mrigala), grass carp (Ctenophmyngodon idella), bighead (Aristichthys nobili), silver carp (Hypophthalmichthys molitrix), common carp (Cyprinus cmpio) and rajpunti (Puntius gonionatus) @ 500/acre were stocked. Feed containing 25% protein was used two times daily and feed was adjusted fortnightly. After 8 months, all the fishes were weighed 0.80-2.10 kg except rajpunti (150-200 g) and tilapia (150-220 g), and a total of 25 ton of fish was harvested which was five times higher than the previous production under signal ownership. The production of fishes were increased after partitioning the lake with dikes due to proper management and control.

The enhancement of TiO2 photocatalytic activity by hydrogen thermal treatment

In this study, conventional TiO2 powder was heated in hydrogen (H-2) gas at a high temperature as pretreatment. The photoactivity of the treated TiO2 samples was evaluated in the photodegradation of sulfosalicylic acid (SSA) in aqueous suspension. The experimental results demonstrated that the photodegradation rates of SSA were significantly enhanced by using the H-2-treated TiO2 catalysts and an optimum temperature for the H-2 treatment was found to be of 500-600 degreesC. The in situ electron paramagnetic resonance (EPR) signal intensity of oxygen vacancies (OV) and trivalent titanium (Ti3+) associated with the photocatalytic activity was studied. The results proved the presence of OV and Ti3+ in the lattice of the H2-treated TiO2 and indicated that both were contributed to the enhancement of photocatalytic activity. Moreover, the experimental results presented that the EPR signal intensity of OV and Ti3+ in the H-2-treated TiO2 samples after 10 months storage was still significant higher than that in the untreated TiO2 catalyst. The experiment also demonstrated that the significant enhancement occurred in the photodegradation of phenol using the H-2-treated TiO2. (C) 2002 Elsevier Science Ltd. All rights reserved.

Electrochemical Metalloimmunoassay Based on Enlargement of Gold Nanoparticle Label Treated with Ag Enhancement System

A novel sensitive electrochemical immunoassay with colloidal gold as the antibody labeling tag and subsequent signal amplification by silver enhancement is described. Colloidal gold was treated by a light-sensitive silver enhancement system which made silver deposit on the surface of colloidal gold(form Au/Ag core-shell structure), followed by the release of the metallic silver atoms anchored on the antibody by oxidative dissolution of them in an acidic solution and the indirect determination of the dissolved Ag+ ions by anodic stripping voltammetry(ASV) at a carbon fiber microelectrode. The electrochemical signal is directly proportional to the amount of analyte(goat IgG) in the standard or a sample. The method was evaluated by means of a noncompetitive heterogeneous immunoassay of immunoglobulin G(IgG) with a concentration as low as 0.2 ng/ mL. The high performance of the method is related to the sensitive ASV determination of silver(I) at a carbon fiber microelectrode and to the release of a large number of Ag+ ions from each silver shell anchored on the analyte(goat IgG).

Leveraging nanoscale plasmonic modes to achieve reproducible enhancement of light.

The strongly enhanced and localized optical fields that occur within the gaps between metallic nanostructures can be leveraged for a wide range of functionality in nanophotonic and optical metamaterial applications. Here, we introduce a means of precise control over these nanoscale gaps through the application of a molecular spacer layer that is self-assembled onto a gold film, upon which gold nanoparticles (NPs) are deposited electrostatically. Simulations using a three-dimensional finite element model and measurements from single NPs confirm that the gaps formed by this process, between the NP and the gold film, are highly reproducible transducers of surface-enhanced resonant Raman scattering. With a spacer layer of roughly 1.6 nm, all NPs exhibit a strong Raman signal that decays rapidly as the spacer layer is increased.

Enhancement of cardiac function after adenoviral-mediated in vivo intracoronary beta2-adrenergic receptor gene delivery.

Exogenous gene delivery to alter the function of the heart is a potential novel therapeutic strategy for treatment of cardiovascular diseases such as heart failure (HF). Before gene therapy approaches to alter cardiac function can be realized, efficient and reproducible in vivo gene techniques must be established to efficiently transfer transgenes globally to the myocardium. We have been testing the hypothesis that genetic manipulation of the myocardial beta-adrenergic receptor (beta-AR) system, which is impaired in HF, can enhance cardiac function. We have delivered adenoviral transgenes, including the human beta2-AR (Adeno-beta2AR), to the myocardium of rabbits using an intracoronary approach. Catheter-mediated Adeno-beta2AR delivery produced diffuse multichamber myocardial expression, peaking 1 week after gene transfer. A total of 5 x 10(11) viral particles of Adeno-beta2AR reproducibly produced 5- to 10-fold beta-AR overexpression in the heart, which, at 7 and 21 days after delivery, resulted in increased in vivo hemodynamic function compared with control rabbits that received an empty adenovirus. Several physiological parameters, including dP/dtmax as a measure of contractility, were significantly enhanced basally and showed increased responsiveness to the beta-agonist isoproterenol. Our results demonstrate that global myocardial in vivo gene delivery is possible and that genetic manipulation of beta-AR density can result in enhanced cardiac performance. Thus, replacement of lost receptors seen in HF may represent novel inotropic therapy.

Tunable Enhancement of High Harmonic Emission from Laser Solid Interactions

Coherent wake emission is a unique source of extreme ultraviolet radiation and has been recently shown to provide the basis for intense attosecond light. Here we present a novel scheme, supported by particle-in-cell simulations, demonstrating that enhancement and spectral control of the coherent wake emission signal can be achieved by modifying the interaction plasma density ramp. Significant tunable enhancement of harmonic emission is verified experimentally, with factors of > 50 in relative signal increase achieved in a narrow band of harmonics at the cutoff frequency.