• ETRI Journal
• /
• v.38 no.2
• /
• pp.366-375
• /
• 2016

This paper describes an algorithm to suppress composite noise in a two-microphone speech enhancement system for robust hands-free speech communication. The proposed algorithm has four stages. The first stage estimates the power spectral density of the residual stationary noise, which is based on the detection of nonstationary signal-dominant time-frequency bins (TFBs) at the generalized sidelobe canceller output. Second, speech-dominant TFBs are identified among the previously detected nonstationary signal-dominant TFBs, and power spectral densities of speech and residual nonstationary noise are estimated. In the final stage, the bin-wise output signal-to-noise ratio is obtained with these power estimates and a Wiener post-filter is constructed to attenuate the residual noise. Compared to the conventional beamforming and post-filter algorithms, the proposed speech enhancement algorithm shows significant performance improvement in terms of perceptual evaluation of speech quality.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.10
• /
• pp.5006-5022
• /
• 2017

Effect and robust detection of targets in infrared images has crucial meaning for many applications, such as infrared guidance, early warning, and video surveillance. However, it is not an easy task due to the special characteristics of the infrared images, in which the background clutters are severe and the targets are weak. The recent literature demonstrates that sparse representation can help handle the detection problem, however, the detection performance should be improved. To this end, in this text, a hybrid method based on local sparse representation and contrast is proposed, which can effectively and robustly detect the infrared targets. First, a residual image is calculated based on local sparse representation for the original image, in which the target can be effectively highlighted. Then, a local contrast based method is adopted to compute the target prediction image, in which the background clutters can be highly suppressed. Subsequently, the residual image and the target prediction image are combined together adaptively so as to accurately and robustly locate the targets. Based on a set of comprehensive experiments, our algorithm has demonstrated better performance than other existing alternatives.

• Structural Engineering and Mechanics
• /
• v.56 no.2
• /
• pp.201-221
• /
• 2015

The fatigue damage problems are frequently encountered in the design of civil engineering structures. A realistic and accurate fatigue life prediction is quite essential to ensure the safety of engineering design. However, constructing a reliable fatigue life prediction model can be quite challenging. The use of traditional deterministic approach in predicting the fatigue life is sometimes too dangerous in the real practical designs as the method itself contains a wide range of uncertain factors. In this paper, a new fatigue life prediction method is going to be proposed where the residual strength is been utilized. Several cumulative damage models, capable of predicting the fatigue life of a structural element, are considered. Based on Miner's rule, a randomized approach is developed from a deterministic equation. The residual strength is used in a one to one transformation methodology which is used for the derivation of the fatigue life. To arrive at more robust results, fuzzy sets are introduced to model the parameter uncertainties. This leads to a convoluted fuzzy based fatigue life prediction model. The developed model is illustrated in an example analysis. The calculated results are compared with real experimental data. The applicability of this approach for a required reliability level is also discussed.

• Journal of Korea Water Resources Association
• /
• v.35 no.4 s.129
• /
• pp.425-442
• /
• 2002

The land reclamation area of Saemangeum(Kunsan) is located between $126^{\circ}10'\;E{\sim}126^{\circ}50'\;E\;and\;35^{\circ}35'\;N{\sim}36^{\circ}05'\;N$ at the western coast of the Korean peninsula. The reclamation area of Saemangeum has a range of 5.6m spring tide and the maximum tidal current speed is about $1.41m\;s^{-1}$. In ordinary spring tide. Most of the sediments deposited on the tidal flats are transported from the Geum river, the Mankyung river and the Dongjin river. The soil in this area consists of silty sand with the depth of 10m to 30m. The wind in winter is strong from the direction of northwest. Saemangeum coastal area is being constructed 33km the sea dike and 40,100ha reclamation area. The purpose of this study Is to find the residual circulations in four seasons before and after the dike construction by a robust diagnostic and prognostic numerical model.

• MALSORI
• /
• no.62
• /
• pp.85-96
• /
• 2007

In this paper, we carried out recognition experiments for noisy speech having various levels of car noise and output of an audio system using the speech interface. The speech interface consists of three parts: pre-processing, acoustic echo canceller, post-processing. First, a high pass filter is employed as a pre-processing part to remove some engine noises. Then, an echo canceller implemented by using an FIR-type filter with an NLMS adaptive algorithm is used to remove the music or speech coming from the audio system in a car. As a last part, the MMSE-STSA based speech enhancement method is applied to the out of the echo canceller to remove the residual noise further. For recognition experiments, we generated test signals by adding music to the car noisy speech from Aurora 2 database. The HTK-based continuous HMM system is constructed for a recognition system. Experimental results show that the proposed speech interface is very promising for robust speech recognition in a noisy car environment.

• ETRI Journal
• /
• v.36 no.1
• /
• pp.167-170
• /
• 2014

In this letter, we propose an unsupervised framework for speech noise reduction based on the recent development of low-rank and sparse matrix decomposition. The proposed framework directly separates the speech signal from noisy speech by decomposing the noisy speech spectrogram into three submatrices: the noise structure matrix, the clean speech structure matrix, and the residual noise matrix. Evaluations on the Noisex-92 dataset show that the proposed method achieves a signal-to-distortion ratio approximately 2.48 dB and 3.23 dB higher than that of the robust principal component analysis method and the non-negative matrix factorization method, respectively, when the input SNR is -5 dB.

• Proceedings of the KIEE Conference
• /
• 1999.11c
• /
• pp.555-557
• /
• 1999

This paper implements an active noise controller in a headset by solving Robust $H_{\infty}$ control problem. In the $H_{\infty}$ control framework, we can suppose the noise in a headset as disturbance and the noise control problem is cast on the well-known $H_{\infty}$ regulation problem. By representing the system as LFT(Linear Fractional Transformaion) form, the controller is obtained using D-K iteraitons. The designed controller was implemented with operational amplifiers and it produced the desired noise reduction performance.

• Proceedings of the KIEE Conference
• /
• 2000.11d
• /
• pp.703-705
• /
• 2000

In this paper, we design a robust adaptive controller for a nonlinear systems with uncertainties to be rejected via disturbance adaptation law. The nonlinear system considered in this paper has unknown nonlinear functions being influenced by external disturbance. The upper bounds of unknown nonlinear functions at each time is estimated by using disturbance adaptation law. The estimated nonlinear functions are used to design stabilizing function and control of input. Tuning function is used to estimate unknown system parameter without overparametrization. A set-point regulation error converges to a residual set close to zero asymptotically fast.

• 제어로봇시스템학회:학술대회논문집
• /
• 1996.10b
• /
• pp.28-31
• /
• 1996

We propose an estimator design method of Stewart platform, which gives the 6DOF, positions and velcities of Stewart platform from the measured cylinder length. The solution of forward kinematics is not solved yet as a useful realtime application tool because of the complexity of the equation with multiple solutions. Hence we suggest an nonlinear estimator for the forward kinematics solution using Luenberger observer with nonlinear error correction term. But the way of residual gain selection of the estimator is not clear, so we suggest an algebraic Riccati equation for gain matrix using Lyapunov method. This algorithm gives the sufficient condition of the stability of error dynamics and can be extended to general nonlinear system.

• Speech Sciences
• /
• v.13 no.1
• /
• pp.129-139
• /
• 2006

In this paper, we propose two effective energy feature normalization methods for robust speech recognition in noisy environments. In the first method, we estimate the noise energy and remove it from the noisy speech energy. In the second method, we propose a modified algorithm for the Log-energy Dynamic Range Normalization (ERN) method. In the ERN method, the log energy of the training data in a clean environment is transformed into the log energy in noisy environments. If the minimum log energy of the test data is outside of a pre-defined range, the log energy of the test data is also transformed. Since the ERN method has several weaknesses, we propose a modified transform scheme designed to reduce the residual mismatch that it produces. In the evaluation conducted on the Aurora2.0 database, we obtained a significant performance improvement.