• Proceedings of the Korea Institute of Convergence Signal Processing
• /
• 2003.06a
• /
• pp.126-129
• /
• 2003

자동 음성 인식(ASR) 시스템에, 끝점 추출과 에너지 정규화는 중요한 역할을 하게 된다 그러나 낮은 SNR이나 nonstationary 환경에서, 기존 방법은 끝점 추출과 에너지 정규화에 있어서 자주 실패하게 되며, ASR을 급격히 열화시키곤 한다. ASR을 수행하기 위해, 최적의 필터에 3상태 천이도를 사용하고, 필터는 정확성과 강인함을 확실히 하기 위해 여러 이론들을 이용하여 설계하였고 여러 가지 잡음이 있는 음성 신호환경에서 거의 일정한 응답을 주었다. 검출된 끝점은 곧바로 에너지 정규화에 적용된다. 실험 결과는 제안된 알고리즘이 낮은 SNR에서 에러율을 크게 감소시키고 있다는 것을 보여준다.

• The Journal of the Acoustical Society of Korea
• /
• v.42 no.4
• /
• pp.364-369
• /
• 2023

Time-delay estimation between two receivers is a technique that has been applied in a variety of fields, from underwater acoustics to room acoustics and robotics. There are two types of time delay estimation techniques: one that estimates the amount of time delay from the correlation between receivers, and the other that parametrically models the time delay between receivers and estimates the parameters by system recognition. The latter has the characteristic that only a small fraction of the system's parameters are directly related to the delay. This characteristic can be exploited to improve the accuracy of the estimation by methods such as Lasso regularization. However, in the case of Lasso regularization, the necessary information is lost. In this paper, we propose a method using Elastic Net that adds Ridge regularization to Lasso regularization to compensate for this. Comparing the proposed method with the conventional Generalized Cross Correlation (GCC) method and the method using Lasso regularization, we show that the estimation variance is very small even for white Gaussian signal sources and colored signal sources.

• Journal of the Institute of Convergence Signal Processing
• /
• v.10 no.1
• /
• pp.72-79
• /
• 2009

This paper describes the performance evaluation of NHD(nonhomogeneity detector) for STAP(space-time adaptive processing) airborne radar according to various normalization methods in the nonhomogeneous clutter environment. In practice, the clutter can be characterized as random variation signals, because it sometimes includes signals with very large magnitude like impulsive signal due to the system environment. The received interference signals are composed of homogeneous and nonhomogeneous data. In this situation, NHB is needed to maintain the STAP performance. The normalization using the NHD result is an effective method for removing the nonhomogeneous data. The optimum normalization can be performed by a representative value considered with a characteristic of the given data, so we propose the K-means clustering algorithm. The characteristic of random variation data due to nonhomogeneous clutters can be considered by the number of clusters, and then the representative value for selecting the homogeneous data is determined in the clustering result. In order to reflect a characteristic of the nonstationary interference data, we also investigate the algorithm for a calculation of the proper number of clusters. Through our simulations, we verified that the K-means clustering algorithm has very superior normalization and target detection performances compared with the previous introduced normalization methods.

• Journal of the Korea Convergence Society
• /
• v.8 no.5
• /
• pp.13-17
• /
• 2017

Existing Speech feature extracting method in speech Signal, there are incorrect recognition rates due to incorrect speech which is not clear threshold value. In this article, the modeling method for improving speech recognition performance that combines the feature extraction for speech and silence characteristics normalized to the non-speech. The proposed method is minimized the noise affect, and speech recognition model are convergence of speech signal feature extraction to each speech frame and the silence feature normalization. Also, this method create the original speech signal with energy spectrum similar to entropy, therefore speech noise effects are to receive less of the noise. the performance values are improved in signal to noise ration by the silence feature normalization. We fixed speech and non speech classification standard value in cepstrum For th Performance analysis of the method presented in this paper is showed by comparing the results with CHMM HMM, the recognition rate was improved 2.7%p in the speech dependent and advanced 0.7%p in the speech independent.

• Proceedings of the IEEK Conference
• /
• 2001.09a
• /
• pp.797-800
• /
• 2001

스펙트럼의 평탄화는 스펙트럼 신호로부터 포만트의 영향이나 천이진폭의 영향을 제거하는 것이다. 따라서 정확한 피치검출과 포만트검출에 적용할 수 있다. 본 논문에서는 새로운 스펙트럼 평탄화 기법을 제안하고 기존의 방법인 LPC법, Cepstrum법과 비교하여 어느 정도의 우수성을 보이는지 평가하였다. 평가 방법은 각각의 평탄화된 신호의 분산을 구하여 평탄화의 정도를 측정하였다. 이때 핑탄화된 신호는 최고점이 영이 되도 록 정규화 시키고 평균이 영인 분산을 계산하였다. 실험 결과는 제안한 방법이 기존의 방법보다 우수함을 보여 준다.

• Journal of Digital Convergence
• /
• v.11 no.6
• /
• pp.169-174
• /
• 2013

Speech recognition, the problem of performance degradation is the difference between the model training and recognition environments. Silence features normalized using the method as a way to reduce the inconsistency of such an environment. Silence features normalized way of existing in the low signal-to-noise ratio. Increase the energy level of the silence interval for voice and non-voice classification accuracy due to the falling. There is a problem in the recognition performance is degraded. This paper proposed a robust speech detection method in noisy environments using a silence feature normalization and voice energy maximize. In the high signal-to-noise ratio for the proposed method was used to maximize the characteristics receive less characterized the effects of noise by the voice energy. Cepstral feature distribution of voice / non-voice characteristics in the low signal-to-noise ratio and improves the recognition performance. Result of the recognition experiment, recognition performance improved compared to the conventional method.

• The Journal of the Acoustical Society of Korea
• /
• v.18 no.5
• /
• pp.35-43
• /
• 1999

In this paper, a new robust adaptive algorithm and a post-processing method are proposed to improve the performance of AEC without computational burden. Its step-size is normalized by the sum of the powers of the reference input signal and the desired signal. When the near-end speaker's speech and noise are applied into the microphone, the step-size becomes small and the misalignment of coefficients are reduced. To reduce the residual echoes, a new post-processing method, which is co-operated with the proposed noise-robust adaptive algorithm, is proposed in this paper. The method is based on the correlation of the desired signal and the estimation error signal. The residual echoes are attenuated as proportional to the correlation normalized with the power of desired signals. The normalized correlation plays a role as Wiener filter for residual echoes. In the double-talk situation, the estimation error signals, that are residual echoes, dominantly include the near-end speaker's speech and the normalized correlation closes to 1. Therefore, the near-end speaker's speech can be transmitted without being attenuated. When the desired signals consists of only the acoustic echoes, the residual echoes are mostly attenuated and canceled by the proposed post-processor. The computation of AEC using the proposed post-processor is comparable to NLMS algorithm.

• Asia-pacific Journal of Multimedia Services Convergent with Art, Humanities, and Sociology
• /
• v.9 no.6
• /
• pp.729-738
• /
• 2019

Recently, we aim to improve the quality of virtual reality contents based on quantitative analysis results of emotions through combination of emotional recognition field and virtual reality field. Emotions are analyzed based on the participant's vital signs. Much research has been done in terms of signal analysis, but the methodology for quantifying emotions has not been fully discussed. In this paper, we propose a normalization function design and expression method to quantify the emotion between various bio - signals. Use the Brute force algorithm to find the optimal parameters of the normalization function and improve the confidence score of the parameters found using the true and false scores defined in this paper. As a result, it is possible to automate the parameter determination of the bio-signal normalization function depending on the experience, and the emotion can be analyzed quantitatively based on this.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.38 no.5
• /
• pp.567-575
• /
• 2001

This paper suggests new method to get fixed size parameter from different length of voice signals. The efficiency of speech recognizer is determined by how to compare the similarity(distance of each pattern) of the parameter from voice signal. But the variation of voice signal and the difference of speech speed make it difficult to extract the fixed size parameter from the voice signal. The method suggested in this paper is to normalize the parameter at fixed size by using the 2 dimension DCT(Discrete Cosine Transform) after representing the parameter by spectrogram. To prove validity of the suggested method, parameter extracted from 32 auditory filter-bank(it estimates auditory nerve firing probabilities) is used for the input of neural network after being processed by 2 dimension DCT. And to compare with conventional methods, we used one of conventional methods which solve time alignment problem. The result shows more efficient performance and faster recognition speed in the speaker dependent and independent isolated word recognition than conventional method.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1994.06c
• /
• pp.175-178
• /
• 1994

본 논문에서는 선형 예측 방법과 다중 대역 여기 방법의 장점을 조합하여 낮은 전송률에서 고품질의 합성음을 제공하는 LP-SMBE 부호기를 제안한다. LP-SMBE 부호기에서는 선형 예측 방법과 단순화된 여기 신호 추정방법을 이용하여 성도 특성 정보와 여기 신호를 분리 추정한다. 제안한 단순화된 여기 신호 추정 방법은 정규화된 스펙트럼 영역에서 원음 스펙트럼과 합성 스펙트럼을 비교하여 여기 신호를 추정한다. 이 방법은 기존 MBE 방법의 여기 신호 추정 방법보다 연산량이 적고, 여기 신호르 F보다 정확히 추정할 수 있다.