• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.36 no.7C
• /
• pp.413-420
• /
• 2011

This paper proposes a new adaptive algorithm to cancel the acoustic feedback and noise signals in the binaural hearing aids. The convergence performances of the proposed algorithm are improved by updating coefficients of the feedback canceller after the speech signal is cancelled from the residual signal with dual microphones. The feedback canceller firstly cancels the feedback signal from the microphone signal, and then the noise canceller reduces the noise by the beamforming method. To assure that binaural hearing aids converge stably, the left-sided hearing aid only is converged firstly, next the right-sided hearing aid only is converged. To verify performances of the proposed algorithm, simulations were carried out for a speech. As the results of simulations, it was proved that we can advance 14.43dB SFR(Signal to Feedback Ratio) on the average for the feedback canceller, 10.19dB SNR(Signal to Noise Ratio) improvement on the average for the noise canceller, in case that this algorithm is used.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2019.11a
• /
• pp.119-121
• /
• 2019

부가적 백색 잡음 모델(additive white Gaussian noise, AWGN에서 학습된 깊은 신경만 (deep neural networks)을 이용한 잡음 제거기는 제거하려는 잡음이 AWGN인 경우에는 뛰어난 성능을 보이지만 실제 카메라 잡음에 대해서 잡음 제거를 시도하였을 때는 성능이 크게 저하된다. 본 논문은 U-Net 구조의 깊은 인공신경망 모델에 residual block을 결합함으로서 실제 카메라 영상에서 기존 알고리즘보다 뛰어난 성능을 지니는 신경망을 제안하다. 제안한 방법을 통해 Darmstadt Noise Dataset에서 PSNR과 SSIM 모두 CBDNet 대비 향상됨을 확인하였다.

• The Journal of the Acoustical Society of Korea
• /
• v.12 no.5
• /
• pp.54-63
• /
• 1993

적응 능동 잡음제거를 위한 새로운 방식을 기술했으며, 잡음원이 존재하는 환경에서 물리적인 잡음 제거를 하는데 목표를 두었다. 기존의 시스템 식별 방식과 달리 적응 예측기를 사용하였으며 신호를 검출하는데 쓰이는 마이크를 하나로 구성하였다. 잡음 신호 자체를 예측함으로써 적극적인 대처를 할 수 있으며 적응 알고리즘은 한번만 사용되고 적응 제어형 시스템 식별 방식에서 필요로 하는 피드백 항을 제거시킴으로써 시스템을 간략화시켰다. 컴퓨터 모의 실험 결과를 통하여 제안한 방식이 주기성을 갖거나 혹은 대역 제한된 잡음에 대해서 매우 유용함을 보였다.

• Proceedings of the IEEK Conference
• /
• 2000.09a
• /
• pp.19-22
• /
• 2000

최근 이동전화의 사용이 급격히 확산됨에 따라 핸즈프리 단말기를 이용한 전화통신의 필요성이 대두되고 있다. 차량내 핸즈프리 통신상황의 경우 근거리에 위치한 스피커와 마이크로폰의 커플링에 의해 발생하는 음향반향과 차량내에 존재하는 배경잡음은 통화 품질을 크게 저하시킨다. 본 논문에서는 이동통신에 적합한 음향반향제거기와 잡음제거기의 결합시스템을 제안하고, 이를 고정 소수점 DSP를 이용하여 실시간 구현하였다. 실시간 구현을 위하여 음향반향제거기에는 NLMS 알고리즘에 의해 구동되는 제한된 차수의 적응반향제거기법을 사용하였다. 잔여반향 및 배경잡음제거를 위해 CDMA방식의 셀룰라 이동통신에 사용되는IS-127 EVRC음성 부호화기의 표준안에 포함된 잡음제거방식을 사용하였다. 제안된 시스템을 16 비트 고정소수점DSP인 OAK DSP Core를 이용하여 약 18.6MIPS의 연산량으로 실시간 구현되었다.

• The Journal of the Acoustical Society of Korea
• /
• v.25 no.7
• /
• pp.325-332
• /
• 2006

In this paper. we Propose the noise suppressor with the Perceptual model and wavelet packet transform. The objective is to enhance speech corrupted colored or non-stationary noise. If corrupted noise is colored. subband approach would be more efficient than whole band one. To avoid serious residual noise and speech distortion, we must adjust the Wavelet Coefficient Threshold (WCT). In this Paper. the subband is designed matching with the critical band and WCT is adapted noise masking threshold (NMT) and segmental signal to noise ratio (seg_SNR). Consequently. it has similar Performance with EVRC in PESQ-MOS. But it's better than wavelet packet transform using universal threshold about 0.289 in PESQ-MOS. The important thing is that it's more useful than EVRC in coded speech. In coded speech. PESQ-MOS is higher than EVRC about 0.23.

• Proceedings of the Acoustical Society of Korea Conference
• /
• spring
• /
• pp.109-112
• /
• 1999

본 연구에서는 신호원에 대한 사전 정보 없이 혼합된 신호로부터 잡음 신호만을 선택적으로 제거할 수 있는 새로운 형태의 능동 소음 제거기(Active noise canceller)를 제안한다. 음성신호와 같은 독특성을 갖는 신호의 분리에 효과적으로 사용되는 동적 재귀 신경망 (Dynamic recurrent neural network)을 원하는 신호원에 섞인 잡음신호를 분리하여 선택적으로 제거하기 위한 능동소음제거기의 전처리기로 미용한다. 능동 소음 제거기는 분리된 잡음 신호에 대한 역 위상 신호를 적응적으로 발생함으로써 특정 위치에서 원하는 신호만을 선택적으로 남길 수 있도록 한다. 컴퓨터를 이용한 시뮬레이션에서는 제안된 시스템이 선택적인 소음제거에 효과적임을 보인다.

• The Journal of the Acoustical Society of Korea
• /
• v.29 no.6
• /
• pp.402-409
• /
• 2010

In this paper, we propose an efficient integrated suppression algorithm based on combined power of acoustic echo and background noise. The proposed method combines the acoustic echo and noise power by the weighting parameter derived from the decision rule based on the estimated echo to noise power ratio. Therefore, in the proposed approach, the acoustic echo and noise signal are able to be reduced through only one suppression filter based on the estimated combined power. The proposed unified structure improves the problems of the residual echo and noise resulted from the conventional unified structure where the noise suppression (NS) operation is placed after the acoustic echo suppression (AES) algorithm or vice versa. The performance of the proposed algorithm is evaluated by the objective test under various environments and yields better results compared with the conventional scheme.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.36 no.6C
• /
• pp.343-348
• /
• 2011

This paper introduces a design for real time wireless headset using ANC (active noise control) system based on NFxLMS adaptive filter algorithm. The training time of the proposed system is significantly reduced by using the RMS delay spread of a channel as an error correction parameter, and convergence rate of the FxLMS filter has been improved with updating the coefficients of the NFxLMS filter, which we have got during the training process. Our system has shorter training time and better convergence rate at the same noise reduction level than the conventional system under real noisy environment.

• Journal of the Institute of Convergence Signal Processing
• /
• v.3 no.2
• /
• pp.30-36
• /
• 2002

This paper describes a nonlinear adaptive noise canceler(ANC) using neural networks(NN) based on filter to make up for the drawback of the conventional ANC with the linear adaptive filter. The proposed ANC was tested its noise rejection performance using broadband time-varying noise signal and compared with the ANC of TDL linear filter. Experimental results show that in cases of nonlinear correlations between the noise of primary input and reference input, the neural network based ANC outperforms the linear ANC with respect to mean square error It is also verified that the recurrent NN adaptive filter is superior to the feedforward NN filter. Thus, we identify that the NN adaptive filter is more effective than the linear adaptive filter for rejection of broadband time-varying noise in the ANC.

• Journal of Broadcast Engineering
• /
• v.14 no.4
• /
• pp.501-508
• /
• 2009

This paper proposes an efficient transform-domain noise reduction scheme in an H.264 video encoder, where the generalized Wiener filtering is performed in a quantization process by multiplying each transform block with its adaptive multiplication factor. In practice, the computational complexity of the proposed scheme is negligible by replacing the multiplication operation with a simple look-up table method. Also, experimental results show that the proposed scheme provides outstanding noise reduction performance in an H.264 video encoder.