• Journal of information and communication convergence engineering
• /
• v.10 no.2
• /
• pp.162-167
• /
• 2012

This paper proposes a speech processing system based on a model of the human auditory system and a noise reduction neural network with fast Fourier transform (FFT) amplitude and phase spectrums for noise reduction under background noise environments. The proposed system reduces noise signals by using the proposed neural network based on FFT amplitude spectrums and phase spectrums, then implements auditory processing frame by frame after detecting voiced and transitional sections for each frame. The results of the proposed system are compared with the results of a conventional spectral subtraction method and minimum mean-square error log-spectral amplitude estimator at different noise levels. The effectiveness of the proposed system is experimentally confirmed based on measuring the signal-to-noise ratio (SNR). In this experiment, the maximal improvement in the output SNR values with the proposed method is approximately 11.5 dB better for car noise, and 11.0 dB better for street noise, when compared with a conventional spectral subtraction method.

• Proceedings of the KAIS Fall Conference
• /
• 2009.05a
• /
• pp.768-771
• /
• 2009

현재 음성인식에서 가장 많이 사용하고 있는 특징벡터는 MFCC(Mel-Frequency Cepstral Coefficients)이다. 그러나 MFCC도 잡음이 존재하는 환경에서는 인식 성능이 저하된다. 이러한 MFCC의 단점을 해결하기 위해 mel sub-band 스펙트럼 차감법과 신호대잡음비에 따른 에너지 압축을 이용하는 CMSBS(Compression and Mel Sub-Band Spectral subtraction) 방법을 사용한다. 본 논문에서는 CMSBS 방법 적용 시 음성이 발성되는 구간과 묵음 구간에서 mel sub-band 스펙트럼 차감법이 동일한 조건으로 이루어져 발생하는 중요한 음성정보의 손실을 보완하기 위하여 신호의 주기성을 이용하여 spectral flooring 파라미터를 변형하는 방법을 제안한다. 제안한 방법으로 실험을 한 결과 잡음이 거의 없는 음성신호에 대해서는 기존의 방법과 비슷한 인식률을 가지고, 잡음성분이 많을수록 변형된 mel sub-band 스펙트럼 차감법을 적용한 방법이 인식률에서 보다 높은 성능 향상을 가져왔다.

• The Journal of the Acoustical Society of Korea
• /
• v.38 no.4
• /
• pp.406-413
• /
• 2019

This paper proposes a speech enhancement method through the process of combining the gain function with spectrum subtraction method in the two microphone array with close spacing. A speech enhancement method that uses a gain function estimated by the SNR (Signal-to Noise Ratio) based on the multi frequency band coherence function causes the performance degradation in high correlation between input noises of two channels. A new speech enhancement method is proposed where the weighted gain function is used by combining the gain function from the spectral subtraction. The performance evaluation of the proposed method was shown by comparison with PESQ (Perceptual Evaluation of Speech Quality) value which is an objective quality evaluation test provided by the ITU-T (International Telecommunications Union Telecommunication). In the PESQ tests, the maximum 0.217 of PESQ value is improved in the various background noise environments.

• Journal of Korea Multimedia Society
• /
• v.4 no.1
• /
• pp.46-55
• /
• 2001

In this paper, a new voice activity detection method estimating SNR of enhanced speech with extended spectral subtraction (ESS) is proposed. Voice activity detection is performed by putting an second Wiener filter behind an Wiener filter used in the ESS to estimate speech and noise power of output signal of first Wiener filter. The proposed voice activity detection method does not require many computational loads and performs well under severe input SNR. Boll's spectral substraction algorithm with proposed voice activity detection was compared to ESS under several noise environment having different time-frequency distributions. During speech and non-speech activity, performance of Boll's spectral substraction algorithm with proposed voice activity detection is superior to that of ESS.

• The Journal of the Acoustical Society of Korea
• /
• v.23 no.2
• /
• pp.180-186
• /
• 2004

The presence of noise in speech signals degrades the performance of recognition systems in which there are mismatches between the training and test environments. To make a speech recognizer robust, it is necessary to compensate these mismatches. In this paper, we studied about an improvement of stochastic feature extraction based on band-SNR for robust speech recognition. At first, we proposed a modified version of the multi-band spectral subtraction (MSS) method which adjusts the subtraction level of noise spectrum according to band-SNR. In the proposed method referred as M-MSS, a noise normalization factor was newly introduced to finely control the over-estimation factor depending on the band-SNR. Also, we modified the architecture of the stochastic feature extraction (SFE) method. We could get a better performance when the spectral subtraction was applied in the power spectrum domain than in the mel-scale domain. This method is denoted as M-SFE. Last, we applied the M-MSS method to the modified stochastic feature extraction structure, which is denoted as the MMSS-MSFE method. The proposed methods were evaluated on isolated word recognition under various noise environments. The average error rates of the M-MSS, M-SFE, and MMSS-MSFE methods over the ordinary spectral subtraction (SS) method were reduced by 18.6%, 15.1%, and 33.9%, respectively. From these results, we can conclude that the proposed methods provide good candidates for robust feature extraction in the noisy speech recognition.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.39C no.1
• /
• pp.60-69
• /
• 2014

Footstep detection using seismic sensors for security is a very meaningful task, but readings can easily fluctuate due to noise in outdoor environment. We propose NSSC method based on nonlinear spectral subtraction and cross-correlation using prime footstep model signal as a footstep signal refining process that enhances the signal-to-noise ratio (SNR) and attenuates noise. After de-noising, a detection event classification method is presented as further refining process to ensure that the detection result is a footstep. To validate the proposed algorithm, representative experiments including sunny and rainy-day cases are demonstrated.

• The Journal of the Acoustical Society of Korea
• /
• v.4 no.2
• /
• pp.13-25
• /
• 1985

광대역 혹은 협대역 잡음이 섞인 음성의 음질을 향상시키기 위하여 빠른 수렴속도를 갖고 잇는 UFBLMS 알고리즘을 음성처리에 응용한다. 광대역 잡음이 섞인 음성인 경우에는, 입력음성의 SNR 이 0 dB에서 10 dB 사이일 때, UFBLMS 알고리즘의 성능이 spectral subtraction 방법이나 Wiener filtering 방법보다도 FWSNR\sub SEG\ 척도로 약 3 dB 더 좋음을 알 수 있다. 협대역 잡음이 섞인 음 성인 경우에는 UFBLMS 알고리즘의 spectral subtraction 방법보다 FWSNR\sub SEG\ 척도로 약 2 dB 정도 성능이 더 좋다. 여러 음질 향상 알고리즘의 계산상의 복잡도와 음질 향상도 및 인식도를 고려해 보면 frequency weighting 기능을 갖고 있는 UFBLMS 알고리즘을 사용하는 것이 바람직함을 알 수 있다.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1985.10a
• /
• pp.87-94
• /
• 1985

In this paper, enhancement of speech corrupted by additive white or colored noise is stuided. The nuconstrained frequency-domain block least-mean-square (UFBLMS) adaptation algorithm and its frequency-weighted version are newly applied to speech enhancement. For enhancement of speech degraded by white noise, the performance of the UFBLMS algorithm is superior to the spectral subtraction method or Wiener filtering technique by more than 3 dB in segmented frequency-weighted signal-to-noise ratio(FWSNERSEG) when SNR of speech is in the range of 0 to 10 dB. As for enhancement of noisy speech corrupted by colored noise, the UFBLMS algorithm is superior to that of the spectral subtraction method by about 3 to 5 dB in FWSNRSEG. Also, it yields better performance by about 2 dB in FWSNR and FWSNRSEG than that of time-domain least-mean-square (TLMS) adaptive prediction filter(APF). In view of the computational complexity and performance improvement in speech quality and intelligibility, the frequency-weighted UFBLMS algorithm appears to yield the best performance among various algorithms in enhancing noisy speech corrupted by white or colored noise.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1998.06e
• /
• pp.411-414
• /
• 1998

음성의 음질 향상(Speech Enhancement)을 위한 여러 가지 방법 중에서 주파수 차감법(Spectral Subtraction)은 계산량이 적기 때문에 현재 실시간으로 Speech Enhancement를 할 수 있는 가장 적절한 방법이다. 그러나, 이 방법은 원래의 입력음성에 없던 새로운 잡음을 만들어내는 큰 단점이 있는데, 이를 제거하기 위해 많은 연구가 되어오고 있다. 이러한 연구의 방향은 대부분 주변프레임 또는 주변의 주파수 성분과의 평균을 통해 피크값을 무디게 해 줌으로써 새로 생긴 튀는 잡음을 감소시키는 것이다. 이런 방법은 음성자체의 정보 또한 평균이 되어버리게 하는 새로운 단점을 낳는데, 이런 현상은 무성음구간에서 특히 심각해진다. 본 논문에서는 입력음성의 LPC 분석으로 백색필터(Whitening Filter)를 구성하여 이를 통과시킨 잔류신호(Residual)를 주파수 차감하여 얻은 새로운 잔류신호를 역 필터링하여(Synthesis Filter) 개선된 음성을 얻는 방법을 제안하였다. 제안된 알고리듬은, 주파수 차감시 포만트(Formant)의 정보가 더 유지 될 수 있기 때문에 잔류잡음을 줄일 수 있다. 청취 테스트 결과 제안한 방법이 기존의 방법보다 잔류잡음을 더 줄이는 사실을 확인할 수 있었다.

• The Journal of the Acoustical Society of Korea
• /
• v.5 no.2
• /
• pp.48-57
• /
• 1986

광대역 또는 협대역잡음이 섞인 음성의 음질을 개선하기 위해 KAK 필터를 사용하는 방법을 제 안한다. KAK 필터는 그 구조가 간단하지만, 잡음이 섞인 음성의 음질을 개선하는데 있어서 객관적인 음질척도로 볼 때 spectral subtraction 방법과 성능이 비슷하다. 또한 귀로 들어봐도 kak 필터를 사용한 경우와 spectral subtraction 방법을 이용한 경우의 개선된 음질이 거의 비슷하다. 그런데 이 kak 필터는 구조가 다른 기존방법보다 훨씬 간단하며, 다른 음질개선 알고리즘과는 달리 음성과 묵음의 판별이 필 요하지 않다. 또한 kak 필터는 ADPCM과 같은 파형 부호화기와 결합하는 것이 용이하다. 따라서 깨끗 한 음성뿐만 아니라 잡음이 섞인 음성을 부호화하는데 있어서 제안한 kak 필터를 ADPCM과 같은 파형 부호화기에 결합하여 사용하는 것이 적합하다.