• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1989년도 추계학술대회 논문집 학회본부
• /
• pp.481-483
• /
• 1989

This paper is concentrated on the recognition of Korean Digits. The speech signals of each of digits are fed into computer through the 18 bandpass filters, AD converter. Spectrum input data are analyzed and used. BASIC program language is used for recognition performance and the result of recognition is outputed to computer screen and printer. In this paper, the strength and weakness of filter-bank analysis method is described and the technique of real-time recognition is argued. In this experiment, Ratio of recognition for speaker dependent recognition was about 97% and recognition time was also satisfied. Therefore, A way of speaker independent recognition will be presented and using for special communication in the future.

• 전기학회논문지
• /
• 제64권1호
• /
• pp.149-158
• /
• 2015

This paper proposes a subband filtering technique to the MUSIC and the ESPRIT algorithm for estimating the magnitude and frequency of the harmonics of power signal. In proposed method, the input power signal is decomposed to the odd harmonics and the even harmonics respectively by the filter bank system. The amplitude and the frequency estimation of the decomposed harmonics are carried out using the MUSIC and the ESPRIT method. Subband filtering can reduce the autocorrelation matrix size of input data, and spectrum leakage between adjacent harmonics. Therefore, this subband technique has advantage in computational cost and estimation accuracy compared to fullband MUSIC and ESPRIT. To demonstrate the performance of the method, computer simulations are performed to the synthesized input signal, and experiment results are compared in subband and fullband cases.

• 대한음성학회:학술대회논문집
• /
• 대한음성학회 2007년도 한국음성과학회 공동학술대회 발표논문집
• /
• pp.123-126
• /
• 2007

This paper suggests a method to improve the performance of the pathological/normal voice classification. The effectiveness of the mel frequency-based filter bank energies using the fisher discriminant ratio (FDR) is analyzed. And mel frequency cepstrum coefficients (MFCCs) and the feature vectors through the linear discriminant analysis (LDA) transformation of the filter bank energies (FBE) are implemented. This paper shows that the FBE LDA-based GMM is more distinct method for the pathological/normal voice classification than the MFCC-based GMM.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1987년도 한국자동제어학술회의논문집; 한국과학기술대학, 충남; 16-17 Oct. 1987
• /
• pp.473-478
• /
• 1987

The purpose of this study is to design a phoneme-class recognition system for Korean language using filter-bank analysis and zero crossing rate method. First, the speech signals are separated in 16 bandpass filters to obtain short-time spectrum of speech signals, and digitized by 16-ch A/D converter. And then, with the set of features which extracted from patterns of ratios of each channel energy level to overall energy level, the decision rules are made for recognize unknown speech signal. In this experiment, the recognition rate was about 93.1 percent for 7 vowels under multitalker environment and 74.4 percent for 10 initial sounds at single speaker.

• 한국기계가공학회지
• /
• 제11권4호
• /
• pp.13-19
• /
• 2012

There are many dynamic properties in measured end-milling force. So, it is difficult to predict the real static property of end-milling force. Also the behavior of end-milling force is very complex to predict with the measured one. To extract the static property from measured force, it must be filtered and its problem is closely related to a de-noising one. Also this paper presents alternative de-noising method of end-milling force using wavelet filter bank, based on the wavelet transform and its inverse one. In this paper, by comparing the measured force and its wavelet filtered one, the fundamental end-milling force property after wavelet transform is well reviewed and analyzed. This result of wavelet filtering with filter bank shows the static force of end-milling which has severe dynamic properties occurring in entry and exit state of edge emersion into the workpiece.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• 제11권2호
• /
• pp.49-57
• /
• 2007

In fingerprints, singular regions including core or delta points have different directional characteristics from non-singular regions. Generally, the ridges of singular regions change more abruptly than those of nonsingular areas, thus in order to effectively enhance fingerprint images regardless of region, local ridge curvature information needs to be used. In this paper, we present an improved Directional Filter Bank (DFB)-based fingerprint image enhancement method that effectively takes advantage of such ridge curvature information. The proposed method first decomposes a fingerprint image into 8 directional subbands using the DFB and then classifies the image into background, low curvature, and high curvature regions using the directional energy estimates calculated from the subbands. Thereafter, the weight values for directional subband processing are determined using classification information and directional energy estimates. Finally, the enhanced image is obtained by synthesizing the processed subbands. The experimental results show that the proposed approach is effective in enhancing both singular and non-singular regions.

• 대한의용생체공학회:의공학회지
• /
• 제27권3호
• /
• pp.101-109
• /
• 2006

It has been known that there is strong relationship between human voices and the movements of the articulatory facial muscles. In this paper, we utilize this knowledge to implement an automatic speech recognition scheme which uses solely surface electromyogram (EMG) signals. The EMG signals were acquired from three articulatory facial muscles. Preliminary, 10 Korean digits were used as recognition variables. The various feature parameters including filter bank outputs, linear predictive coefficients and cepstrum coefficients were evaluated to find the appropriate parameters for EMG-based speech recognition. The sequence of the EMG signals for each word is modelled by a hidden Markov model (HMM) framework. A continuous word recognition approach was investigated in this work. Hence, the model for each word is obtained by concatenating the subword models and the embedded re-estimation techniques were employed in the training stage. The findings indicate that such a system may have a capacity to recognize speech signals with an accuracy of up to 90%, in case when mel-filter bank output was used as the feature parameters for recognition.

• 전기학회논문지
• /
• 제59권2호
• /
• pp.476-483
• /
• 2010

If the data are pre-whitened, then gradient adaptive algorithms which are simpler than higher order statistics algorithms can be used in adaptive blind signal estimation. In this paper, we propose a blind subband affine projection algorithm for multiple-input multiple-output adaptive equalization in the blind environments. All of the adaptive filters in subband affine projection equalization are decomposed to polyphase components, and the coefficients of the decomposed adaptive sub-filters are updated by defining the multiple cost functions. An infinite impulse response filter bank is designed for the data pre-whitening. Pre-whitening procedure through subband filtering can speed up the convergence rate of the algorithm without additional computation. Simulation results are presented showing the proposed algorithm's convergence rate, blind equalization and blind signal separation performances.

• 반도체디스플레이기술학회지
• /
• 제9권4호
• /
• pp.57-61
• /
• 2010

The wavelet transform was adopted as the transform for JPEG2000. However, wavelet has weakness about smoothness along the contours and limited directional information. So we use to other transform, called contourlet transform in compression. Objective quality assessment methods currently used Peak signal to noise Ratio(PSNR). But that is not very well matched to perceived visual quality. So new image quality assessment is required. In this paper, we propose a new method for image compression based on the contourlet transform, which has been recently introduced. In addition we evaluated compression image quality using PSNR and SSIM. Finally contourlet transform has a good result about images with smooth contours and SSIM is good method for image evaluation compared to PSNR.

• 한국통신학회논문지
• /
• 제26권11A호
• /
• pp.1841-1847
• /
• 2001

본 논문에서는 주파수 선택적 다중 경로 페이딩 채널에서 필터 뱅크 기반의 직교 주파수 분할 다중 데이터 전송 시스템을 위한 새로운 DFE 등화 기법을 제안하였다. 기존의 OFDM 시스템과 비교하여 제안된 시스템의 복잡도가 증가하지만 비트 오류율에 있어서 나은 성능을 얻을 수 있었다. 시뮬레이션 결과 특히 낮은 SNR 태널 환경에서 제안한 기법이 우수한 성능을 보였다.