http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
장승관,김창석 명지대학교 대학원 1999 대학원논문집 Vol.3 No.-
This paper proposed a adaptive dividing method that divided a speech signal into an intial and a middle and a final sound as the form of utterance. A speech signal that composited a consonant and a vowel and a consonant was divided into three period, we analyzed a features of samples by LPC in each periods. As the results that were carried out with spoken words, it was found that the initial and middle period had the spectrum features of the a consonant and vowel respectively, that end period had the both of them. Also we shown that all kinds of words were adaptively divided to 3 period by using the proposed method, that words, which a initial sound are same and a middle and a final sound are differency, had nearly the same spectrum characteristics of the initial period, the spectrum of the middle and the final period are different.
장승관,최성연,김창석 한국음성과학회 1999 음성과학 Vol.5 No.1
In this paper, an adaptive method of dividing a speech signal into an initial, a medial and a final sound of the form of utterance utilized by evaluating extreme limits of shor term energy and autocorrelation functions. By applying this method into speech signal composed of a consonant, a vowel and a consonant, it was divided into an initial, a medial and a final sound and its feature analysis of sample by LPC were carried out. As a result of spectrum analysis in each period, it was observed that there existed spectrum features of a consonant and a vowel in the initial and medial periods respectively and features of both in a final sound. Also, when all kinds of words were adaptively divided into 3 periods by using the proposed method, it was found that the initial sounds of the same consonant and the medial sounds of the same vowels have the same spectrum characteristics respectively, but the final sound showed different spectrum characteristics even if it had the same consonant as the initial sound. Keywords: extreme limits, speech signal, transient period, short-term energy
음성신호의 최적특징을 적응적으로 추출하는 방법에 관한 연구
장승관,차태호,최웅세,김창석 한국통신학회 1994 韓國通信學會論文誌 Vol.19 No.2
본 논문에서는 음성신호를 일정한 크기로 적응시켜 최적의 특징을 추출할 수 있는 방법을 제안하였다. 음성신호의 특징을 추출하기 위하여 고속선형예측 알고리즘인 FRLS 적용할 때 음성신호를 일정한 크기로 분할한 후 각 프레임 마다 제안한 균등사기상관함수를 가지고 최적특징을 추출하였다. In this paper, we proposed a method of extracting optimum features of speech signal to adjust signal level. For extracting features of speech signal we used FRLS(Fast Recursive Least Square) algorithm, we adjusted each frames of equal to constant level, and extracted optimum features of speech signal by using equalized autocorrelation function proposed in this paper.
단구간에너지와 변곡점 평가에 의한 음성의 천이구간 특징추출
최일홍,장승관,김창석 명지대학교 대학원 1997 대학원논문집 Vol.1 No.-
This paper proposed a discriminating method by estimating the turning points and the average magnitude energy in speech signal. This method solved a problem of zero-crossing rate. Transient samples before signal does not reach steady state from the beginning point are discriminated and also features of transient data are estimated by this method. As results that carried out with speech signals, it was found that even through speech samples included in D.C level, the beginning and end point of the speech signals was exactly discriminated by this method, a length of transient period and a short time energy and frequency features were extracted from transient samples and these parameters were compared to each spoken words.
변곡점 및 단구간 에너지평가에 의한 음성의 천이구간 특징분석
최일홍,장승관,차태호,최웅세,김창석 한국음성과학회 1998 음성과학 Vol.3 No.-
In this paper, I would like to propose a dividing method by estimating the inflection points and the average magnitude energy in speech signals. The method proposed in this paper gave not only a satisfactory solution for the problems on dividing method by zero-crossing rate, but could estimate the feature of the transient period after dividing the starting point and transient period in speech signals before steady state. In the results of the experiment carried out with monosyllable speech, it ws found that even through speech samples indicated in D.C. level, the staring and ending point of the speech signals were exactly divided by the method. In addition to the results, I could compare with the features, such as the length of transient period, the short term energy, the frequency characteristics, in each speech signal. Keywords: inflection points, speech signal, transient period, short-term energy
음성인식을 위한 복합형잡음제거필터와 최적특징추출에 관한 연구
차태호,장승관,최웅세,최일홍,김창석 한국음성과학회 1998 음성과학 Vol.4 No.2
In this paper, a novel method of noise reduction of speech based on a complex adaptive noise canceler and method of optimal feature extraction are proposed. this complex adaptive noise canceler needs simply the noise detection, and LMS algorithm used to calculate the adaptive filter coefficient. The method of optimal feature extraction requires the variance of noise. The experimental results have shown that the proposed method effectively reduced noise in noisy speech. Optimal feature extraction has shown similar characteristics in noise-free speech. Keywords : adaptive noise canceler, feature extraction, LMS algorithm
추정평면에서 평가한 데이터와 인공신경망에 의한 숫자음 인식
최일홍,장승관,차태호,최웅세,김창석,Choi, Il-Hong,Jang, Seung-Kwan,Cha, Tae-Hoo,Choi, Ung-Se,Kim, Chang-Seok 한국음향학회 1996 韓國音響學會誌 Vol.15 No.4
본 논문은 추정평면의 데이터로부터 특징파라미터의 평가와 인공신경망에 의한 음성인식방법을 제안한다. 각 프레임에서 평가한 LPC는 매핑함수를 이용하여 추정평면으로 매핑시켰으며, 본 논문에서는 이 추정평면의 데이터로부터 C-LPC, 최대값, 최소값, 3등분할 파워 특징값을 평가하였다. 추정평면에서 평가한 특징 파라미터는 인공신경망에 입력한 음성인식 실험으로부터 원 음성신호의 시간변화에 따른 특징을 포함하고 있음을 확인하였고, 제안한 방법에 의한 인식으로부터 인식율이 약 96.3%이었다. This paper was proposed the recognition method by using parameters which was estimated from the data on the estimated plane and a neural network. After the LPC estimated in each frame algorithm was mapped to the estimated plane by the optimum feature mapping function, we estimated the C-LPC and the maximum and minimum value and 3 divided power from the mapping data on the estimated plane. As a result of the experiment of the speech recognition that those parameters were applied to the input of a neural network, it was found that those parameters estimated from the estimated plane have the features of the original speech for a change in the time scale and that the recongnition rate by the proposed methods was 96.3 percent.
최웅세,장승관,김창석 明知大學校 産業技術硏究所 1994 産業技術硏究所論文集 Vol.13 No.-
This thesis is proposed the adaptive overlap function for absorbing the time variation of speech signal. In this paper, the time variation which is known to the problem on speech recognition is absorbed by diving the speech signal into the equal number of frames regardless of lenght of signal and the frames of the signal are divided by the adaptive overlap function. Equal number of the features of the speech signal extracted from the frame data by using FRLS. As the results that the proposed algorithm applied to the numeral speech, it was found that the adaptive overlap function was absorbed the time variation by dividing the speech signal into the constant frame number, equal number of the features of the speech signal were extracted from each frame.