• The Journal of the Acoustical Society of Korea
• /
• v.16 no.5
• /
• pp.112-121
• /
• 1997

In this paper, we study a model parameter compensation method for noise-robust speech recognition. We study model parameter compensation on a sentence by sentence and no other informations are used. Parallel model combination(PMC), well known as a model parameter compensation algorithm, is implemented and used for a reference of performance comparision. We also propose a modified PMC method which tunes model parameter with an association factor that controls average variability of gaussian mixtures and variability of single gaussian mixture per state for more robust modeling. We obtain a re-estimation solution of environmental variables based on the expectation-maximization(EM) algorithm in the cepstral domain. To evaluate the performance of the model compensation methods, we perform experiments on speaker-independent isolated word recognition. Noise sources used are white gaussian and driving car noise. To get corrupted speech we added noise to clean speech at various signal-to-noise ratio(SNR). We use noise mean and variance modeled by 3 frame noise data. Experimental result of the VTS approach is superior to other methods. The scheme of the zero order VTS approach is similar to the modified PMC method in adapting mean vector only. But, the recognition rate of the Zero order VTS approach is higher than PMC and modified PMC method based on log-normal approximation.

• The Journal of the Acoustical Society of Korea
• /
• v.21 no.3
• /
• pp.323-329
• /
• 2002

This paper proposes a frame reliability weighting method to compensate for a time-selective noise that occurs at random positions of speech signal contaminating certain parts of the speech signal. Speech frames have different degrees of reliability and the reliability is proportional to SNR (signal-to noise ratio). While it is feasible to estimate frame Sl? by using the noise information from non-speech interval under a stationary noisy situation, it is difficult to obtain noise spectrum for a time-selective noise. Therefore, we used statistical models of clean speech for the estimation of the frame reliability. The proposed MFR (model-based frame reliability) approximates frame SNR values using filterbank energy vectors that are obtained by the inverse transformation of input MFCC (mal-frequency cepstral coefficient) vectors and mean vectors of a reference model. Experiments on various burnt noises revealed that the proposed method could represent the frame reliability effectively. We could improve the recognition performance by using MFR values as weighting factors at the likelihood calculation step.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2004.05a
• /
• pp.774-779
• /
• 2004

This paper addresses the way in which we can find where impact noise sources are. Specifically, we have an interest in the case that the signal is embedded in noise. We propose a signal processing method that can identify impulsive sources’location. The method is robust with respect to noise; spatially distributed noise. This has been achieved by a beamforming method with regard to cepstrum domain is used. It is noteworthy that the cepstrum has the ability to detect periodic pulse signal in noise. Numerical simulation and experiments are performed to verify the method. Results show that the proposed technique is quite powerful for localizing the faults in noisy environments. The method also required less microphones than conventional beamforming method.

• Proceedings of the Korean Magnestics Society Conference
• /
• 2000.09a
• /
• pp.546-560
• /
• 2000

This paper describes the design and evaluation of a noise suppressing hydrophone that is robust to external noise without sacrificing its performance as a receiver. To increase robustness of the receiver to the external noise, first, effects of location of external noise on its performance are analyzed with the finite element method (FEM). Based on the results, geometrical variations are implemented on the structure with additional air pockets and damping layers that work as acoustic shields or scatterers of the noise, and fourteen trial models are developed for the noise suppressing hydrophone structures. The results show that the effect of the external noise is most significant when it is applied to near the mid-side surface of the hydrophone housing. The external noise is isolated most efficiently when two thin damping layers combined with five air pockets are inserted to the circumference of the hydrophone housing. Overall, of the fourteen structural variations of the hydrophone, the best one shows about 87% reduction in the response of the original structure to external noise.

• International Journal of Contents
• /
• v.8 no.2
• /
• pp.1-6
• /
• 2012

Recently, content-based music information retrieval (MIR) systems for mobile devices have attracted great interest. However, music retrieval systems are greatly affected by background noise when music is recorded in noisy environments. Therefore, we evaluated various pre-processing methods using the Philips method to determine the one that performs most robust music retrieval in such environments. We found that dynamic noise reduction (DNR) is the best pre-processing method for a music retrieval system in noisy environments.

• Journal of Korean Society for Quality Management
• /
• v.26 no.2
• /
• pp.17-26
• /
• 1998

Taguchi's parameter design is a method for quality improvement by making the performance fo a system robust to noise. Parameter design with dynamic characteristics has been recently the subject of much interest. This paper is concerned with a review and a generalization of the Signal-to-Noise (SN) ratio, a quality measure for parameter design with dynamic characteristics, proposed by Taguchi. We present a method for determination and analysis of the generalized SN ratio and illustrate its implementation by example.

• 한국HCI학회:학술대회논문집
• /
• 2008.02a
• /
• pp.190-194
• /
• 2008

This paper presents a speaker identification technique which is one of the basic techniques of the ubiquitous robot companion. Though the conventional mel-frequency cepstral coefficients guarantee high performance of speaker identification in clean condition, the performance is degraded dramatically in noise condition. To overcome this problem, we employed the relative autocorrelation sequence mel-frequency cepstral coefficient which is one of the noise robust features. However, there are two problems in relative autocorrelation sequence mel-frequency cepstral coefficient: 1) the limited information problem. 2) the residual noise problem. In this paper, to deal with these drawbacks, we propose a multi-streaming method for the limited information problem and a hybrid method for the residual noise problem. To evaluate proposed methods, noisy speech is used in which air conditioner noise, classic music, and vacuum noise are artificially added. Through experiments, proposed methods provide better performance of speaker identification than the conventional methods.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.38C no.2
• /
• pp.189-195
• /
• 2013

In this paper, a land and maritime unified tourism information guide system is proposed which employs robust speech recognition in ship noise environments. Most of conventional front-ends for speech recognition have used a Wiener filter to compensate for stationary noise such as car or babble noises. However, such the conventional front-ends have limitation in reducing non-stationary noise that are occurred inside the ship on voyage. To overcome such a limitation, the proposed system incorporates nonlinear multi-band spectral subtraction to provide highly accurate tourism route recognition. It is shown from the experiment that compared to a conventional system the proposed system achieves relative improvement of a tourism route recognition rate by 5.54% under a noise condition of 10 dB signal-to-noise ratio (SNR).

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.37 no.4A
• /
• pp.269-277
• /
• 2012

When there are mismatches between the beamforming steering vector and the array response vector for the desired signal, the performance can be severely degraded as the adaptive array attempts to suppress the desired signal as well as interferences. In this paper, an robust method is proposed for the adaptive array in the presence of both direction errors and random errors in the steering vector. The proposed method first finds a signal-plus-interference subspace (SIS) from the correlation matrix, which in turn is exploited to extract an interference subspace based on the structure of a uniform linear array (ULA), the effect of the desired signal direction vector being reduced as much as possible. Then, the weight vector is attained to be orthogonal to the interference subspace. Simulation shows that the proposed method, in terms of signal-to-interference plus noise ratio (SINR), outperforms existing ones such as the doubly constrained robust Capon beamformer (DCRCB).

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.14 no.5
• /
• pp.1093-1102
• /
• 2010

Principal component analysis (PCA) is a well-known method for dimension reduction while maintaining most of the variation in data. Although PCA has been applied to many areas successfully, it is sensitive to outliers. Several variants of PCA have been proposed to resolve the problem and, among the variants, robust fuzzy PCA (RF-PCA) demonstrated promising results. RF-PCA uses fuzzy memberships to reduce the noise sensitivity. However, there are also problems in RF-PCA and the convergence property is one of them. RF-PCA uses two different objective functions to update memberships and principal components, which is the main reason of the lack of convergence property. The difference between two functions also slows the convergence and deteriorates the solutions of RF-PCA. In this paper, a variant of RF-PCA, called RF-PCA2, is proposed. RF-PCA2 uses an integrated objective function both for memberships and principal components. By using alternating optimization, RF-PCA2 is guaranteed to converge on a local optimum. Furthermore, RF-PCA2 converges faster than RF-PCA and the solutions found are more similar to the desired solutions than those of RF-PCA. Experimental results also support this.