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.

Enzyme activity of horseradish peroxidase immobilized in chitosan matrices in alternated layers

The successful immobilization of enzymes such as horseradish peroxidase (HRP) in solid films is essential for applications in sensors and for fundamental studies aimed at identifying possible biotechnological devices. In this study we show that HRP can be immobilized in alternated layers with chitosan as the template material. The activity of HRP in HRP/chitosan films was preserved for several weeks, and could be detected optically upon monitoring the reaction with pyrogallol. The morphology of the film displayed stripes that disappeared after reaction with pyrogallol. Though the activity in the HRP/chitosan film was lower than in a homogeneous solution or in an LB film investigated earlier, the response was linear for a considerable period of time, which may be advantageous for sensing hydrogen peroxide. (C) 2009 Elsevier B.V. All rights reserved.

Immobilization of lutetium bisphthalocyanine in nanostructured biomimetic sensors using the LbL technique for phenol detection

This study describes the development of amperometric sensors based on poly(allylamine hydrochloride) (PAH) and lutetium bisphthalocyanine (LuPc(2)) films assembled using the Layer-by-Layer (LbL) technique. The films have been used as modified electrodes for catechol quantification. Electrochemical measurements have been employed to investigate the catalytic properties of the LuPc(2) immobilized in the LbL films. By chronoamperometry, the sensors present excellent sensitivity (20 nA mu M(-1)) in a wide linear range (R(2) = 0.994) up to 900 mu M and limit of detection (s/n = 3) of 37.5 x 10(-8) M for catechol. The sensors have good reproducibility and can be used at least for ten times. The work potential is +0.3 V vs. saturated calomel electrode (SCE). In voltammetry measurements, the calibration curve shows a good linearity (R(2) = 0.992) in the range of catechol up to 500 mu M with a sensitivity of 90 nA mu M(-1) and LD of 8 mu M. (C) 2011 Elsevier B.V. All rights reserved.

Activated Copper Cathodes as Sensors for Nitrite Analysis

The increased surface area of copper electrodes upon applying a suitable potential protocol was characterized by atomic force microscopy images. Scanning electrochemical microscopy was used to demonstrate the enhanced reactivity of the generated surface. The modified electrode showed excellent catalytic activity towards nitrite reduction in acidic medium (pH 2). This new platform was used in the development of a fast and simple voltammetric method for nitrite determination. Commercial and rainwater spiked samples were analyzed and the data showed an excellent agreement with those obtained with a reference spectrophotometric method (Griess reaction) at a confidence level of 95% (Student`s t-test).

Flow injection analysis of ethyl xanthate by gas diffusion and UV detection as CS(2) for process monitoring of sulfide ore flotation

A sensitive and robust analytical method for spectrophotometric determination of ethyl xanthate, CH(3)CH(2)OCS(2)(-) at trace concentrations in pulp solutions from froth flotation process is proposed. The analytical method is based on the decomposition of ethyl xanthate. EtX(-), with 2.0 mol L(-1) HCl generating ethanol and carbon disulfide. CS(2). A gas diffusion cell assures that only the volatile compounds diffuse through a PTFE membrane towards an acceptor stream of deionized water, thus avoiding the interferences of non-volatile compounds and suspended particles. The CS(2) is selectively detected by UV absorbance at 206 nm (epsilon = 65,000 L mol(-1) cm(-1)). The measured absorbance is directly proportional to EtX(-) concentration present in the sample solutions. The Beer`s law is obeyed in a 1 x 10(-6) to 2 x 10(-4) mol L(-1) concentration range of ethyl xanthate in the pulp with an excellent correlation coefficient (r = 0.999) and a detection limit of 3.1 x 10(-7) mol L(-1), corresponding to 38 mu g L. At flow rates of 200 mu L min(-1) of the donor stream and 100 mu L min(-1) of the acceptor channel a sampling rate of 15 injections per hour could be achieved with RSD < 2.3% (n = 10, 300 mu L injections of 1 x 10(-5) mol L(-1) EtX(-)). Two practical applications demonstrate the versatility of the FIA method: (i) evaluation the free EtX(-) concentration during a laboratory study of the EtX(-) adsorption capacity on pulverized sulfide ore (pyrite) and (ii) monitoring of EtX(-) at different stages (from starting load to washing effluents) of a flotation pilot plant processing a Cu-Zn sulfide ore. (C) 2010 Elsevier By. All rights reserved.

Determination of carbaryl in tomato ""in natura"" using an amperometric biosensor based on the inhibition of acetylcholinesterase activity

This work reports the utilization of two methodologies for carbaryl determination in tomatoes. The measurements were carried out using an amperometric biosensor technique based on the inhibition of acetylcholinesterase activity due to carbaryl adsorption and a HPLC procedure. The electrochemical experiments were performed in 0.1 mol L-1 phosphate buffer solutions at pH 7.4 with an incubation time of 8 min. The analytical curve obtained in pure solutions showed excellent linearity in the 5.0 x 10(-5) to 75 x 10(-5) mol L-1 range, with the limit of detection at 0.4 x 10(-3) gL(-1). The application of such a methodology in tomato samples involved solely liquidising the samples, which were spiked with 6.0 x 10(-6) and 5.0 x 10(-5) mol L-1 carbaryl. Recovery in such samples presented values of 99.0 and 92.4%, respectively. In order to obtain a comparison, HPLC experiments were also conducted under similar conditions. However, the tomato samples have to be manipulated by an extraction procedure (MSPD), which yielded much lower recovery values (78.3 and 84.8%, respectively). On the other hand, the detection limit obtained was much lower than that for the biosensor, i.e., 3.2 x 10(-6) g L-1. Finally, the biosensor methodology was employed to analyze carbaryl directly inside the tomato, without any previous manipulation. In this case, the biosensor was immersed in the tomato pulp, which had previously been spiked with the pesticide for 8 min, removed and inserted in the electrochemical cell. A recovery of 83.4% was obtained, showing very low interference of the matrix constituents. (C) 2007 Elsevier B.V. All rights reserved.