• MALSORI
• /
• no.65
• /
• pp.153-168
• /
• 2008

We propose a new method to apply shape analysis techniques to a computational auditory scene analysis (CASA)-based speech separation system. The conventional CASA-based speech separation system extracts speech signals from a mixture of speech and noise signals. In the proposed method, we complement the missing speech signals by applying the shape analysis techniques such as labelling and distance function. In the speech separation experiment, the proposed method improves signal-to-noise ratio by 6.6 dB. When the proposed method is used as a front-end of speech recognizers, it improves recognition accuracy by 22% for the speech-shaped stationary noise condition and 7.2% for the two-talker noise condition at the target-to-masker ratio than or equal to -3 dB.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.17 no.3
• /
• pp.409-416
• /
• 2022

This paper relates to the blind noise separation method of time-delayed convolutive mixed signals. Since the mixed model of acoustic signals in a closed space is multi-channel, a convolutive blind signal separation method is applied and time-delayed data samples of the two microphone input signals is used. For signal separation, the mixing coefficient is calculated using an inverse model rather than directly calculating the separation coefficient, and the coefficient update is performed by repeated calculations based on secondary statistical properties to estimate the speech signal. Many simulations were performed to verify the performance of the proposed blind signal separation. As a result of the simulation, noise separation using this method operates safely regardless of convolutive mixing, and PESQ is improved by 0.3 points compared to the general adaptive FIR filter structure.

• Proceedings of the KSME Conference
• /
• 2001.11b
• /
• pp.300-305
• /
• 2001

The unsteady flow structure and the related noise generation, which are caused by the separation of a two-dimensional, incompressible, laminar boundary-layer on the flat plate under the influence of local adverse pressure gradient, are numerically examined. The characteristic lines of the wall pressure are examined to understand the unsteady behavior of vortex shedding near the reattachment point of the separation bubble. Also, the generation and propagation of the vortex-induced noise in the separated boundary-layer are calculated by the method of computational aero-acoustics (CAA), and the effects of Reynolds number, Mach number and the strength of the adverse pressure gradient on the unsteady flow and noise characteristics are examined.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2007.05a
• /
• pp.928-933
• /
• 2007

In this research, we investigated active vibration control based on wave-viewpoint using the longitudinal wave separation method. The control strategy is the one of active vibration control technique for generating vibration reduced zone and uses wave information including the directivity as the cost function. In order to get the wave information from the measured values, we proposed and examined the time-domain longitudinal wave separation method proper to real time application like active vibration control. Using the proposed method, we examine the performance and feasibility of active vibration control based wave view-point through the simulation. The related experimental verification and application is going to be expected in a near future.

• Journal of the Korean Society for Precision Engineering
• /
• v.16 no.7
• /
• pp.178-187
• /
• 1999

In order to identify a transfer function model with noise, penalty function method has been widely used. In this method, estimation process for possible model parameters from low to higher order proceeds the model identification process. In this study, based on linear estimation method, a new approach unifying the estimation and the identification of ARMAX model is proposed. For the parameter estimation of a transfer function model with noise, linear estimation method by noise separation is suggested instead of nonlinear estimation method. The feasibility of the proposed model identification and estimation method is verified through simulations, namely by applying the method to time series model. In the case of time series model with noise, the proposed method successfully identifies the transfer function model with noise without going through model parameter identification process in advance. A new algorithm effectively achieving model identification and parameter estimation in unified frame has been proposed. This approach is different from the conventional method used for identification of ARMAX model which needs separate parameter estimation and model identification processes. The consistency and the accuracy of the proposed method has been verified through simulations.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.15 no.3
• /
• pp.294-299
• /
• 2005

This paper presents the separation performance of the linearly mixed image signals with additive noises by using an independent component analyses(ICAs) of the fixed-point(FP) algorithm based on Newton and secant method, respectively. The Newton's FP-ICA uses the slope of objective function, and the secant's FP-ICA also uses the tangent line of objective function. The 2 kinds of ICA have been applied to the 2 dimensional 2-image with $512\times512$ pixels. Then Gaussian noise and Laplacian noise are added to the mixed images, respectively. The experimental results show that the Newton's FP-ICA has better the separation speed than secant FP-ICA and the secant's FP-ICA has also the better separation rate than Newton's FP-ICA. Especially, the Newton and secant method gives relatively larger improvement degree in separation speed and rate as the noise increases.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2005.05a
• /
• pp.322-325
• /
• 2005

The separation mechanism is installed to separate a note one by one from the stacked notes and the overlap type, one of separation mechanism, has been used a lot in financial equipments like ATM. This paper has compared and estimated analysis results using finite element method with experimental results over various parameters such as conditions of note, overlap value, roller shapes, which affect the friction force (resistance) exerting on notes between rollers. Consequently, the effect of various parameters on the performance of overlap type separation mechanism can be known and optimal shape and overlap value can be obtained.

• Smart Structures and Systems
• /
• v.26 no.2
• /
• pp.135-146
• /
• 2020

This paper deals with an application of adaptive blind source separation (BSS) method, equivariant adaptive separation via independence (EASI), and Teager Energy Operator (TEO) for online identification of structural modal parameters. The aim of adaptive BSS methods is recovering a set of independent sources from their unknown linear mixtures in each step when a new sample is received. In the proposed approach, firstly, the EASI method is used to decompose structural responses into independent sources at each instance. Secondly, the TEO based demodulation method with discrete energy separation algorithm (DESA-1) is applied to each independent source, and the instantaneous frequencies and damping ratios are extracted. The DESA-1 method can provide the fast time response and has high resolution so it is suitable for online problems. This paper also compares the performance of DESA-1 algorithm with Hilbert transform (HT) method. Compared to HT method, the DESA-1 method requires smaller amounts of samples to estimate and has a smaller computational complexity and faster adaption due to instantaneous characteristic. Furthermore, due to high resolution of the DESA-1 algorithm, it is very sensitive to noise and outliers. The effectiveness of the proposed approach has been validated using synthetic examples and a benchmark structure.

• MALSORI
• /
• no.48
• /
• pp.81-91
• /
• 2003

In this paper, we propose a speech recognition method based on independent component analysis (ICA) and spectral enhancement techniques. While ICA tris to separate speech signal from noisy speech using multiple channels, some noise remains by its algorithmic limitations. Spectral enhancement techniques can compensate for lack of ICA's signal separation ability. From the speech recognition experiments with instantaneous and convolved mixing environments, we show that the proposed approach gives much improved recognition accuracies than conventional methods.

• Proceedings of the KSPS conference
• /
• 2003.10a
• /
• pp.97-100
• /
• 2003

This paper addresses a method of convolutive source separation that based on SEONS (Second Order Nonstationary Source Separation) [1] that was originally developed for blind separation of instantaneous mixtures using nonstationarity. In order to tackle this problem, we transform the convolutive BSS problem into multiple short-term instantaneous problems in the frequency domain and separated the instantaneous mixtures in every frequency bin. Moreover, we also employ a H infinity filtering technique in order to reduce the sensor noise effect. Numerical experiments are provided to demonstrate the effectiveness of the proposed approach and compare its performances with existing methods.