• The Journal of the Acoustical Society of Korea
• /
• v.20 no.3
• /
• pp.56-62
• /
• 2001

In this paper, we adopt a missing data theory to speech recognition. It can be used in order to maintain high performance of speech recognizer when the missing data occurs. In general, hidden Markov model (HMM) is used as a stochastic classifier for speech recognition task. Acoustic events are represented by continuous probability density function in continuous density HMM(CDHMM). The missing data theory has an advantage that can be easily applicable to this CDHMM. A marginalization method is used for processing missing data because it has small complexity and is easy to apply to automatic speech recognition (ASR). Also, a spectral subtraction is used for detecting missing data. If the difference between the energy of speech and that of background noise is below given threshold value, we determine that missing has occurred. We propose a new method that examines the reliability of detected missing data using voicing probability. The voicing probability is used to find voiced frames. It is used to process the missing data in voiced region that has more redundant information than consonants. The experimental results showed that our method improves performance than baseline system that uses spectral subtraction method only. In 452 words isolated word recognition experiment, the proposed method using the voicing probability reduced the average word error rate by 12％ in a typical noise situation.

• Proceedings of the KIEE Conference
• /
• 2005.11a
• /
• pp.9-11
• /
• 2005

본 연구는 GIS 진단신뢰성 향상기술 개발을 목적으로, 16개의 인위적 결함을 이용하여 부분방전 신호를 발생시키고 검출하여 그 패턴인식 확률을 높이기 위하여 신경망에 Genetic Algorithm (GA) 을 적용하였다. 이를 위하여 다음과 같은 5가지 서로 다른 신경망 모델을 선택하였다: Back Propagation (BP), Jordan-Elman Network (JEN), Principal Component Analysis (PCA), Self-Organizing Feature Map (SOFM) 및 Support Vector Machine (SVM). 이와 같이 선택된 모델에 동일한 데이터를 학습 시키고 패턴인식 확률을 비교 및 분석하였다. 실험 결과에 의하면, BP의 인식률이 가장 높고 다음으로 JEN의 인식률이 높이 나타났으며, 후자의 경우 모든 결함에 대하여 정확한 패턴분류를 한 반면에 전자의 경우 1.8% 의 분류 오차가 발생하였다. 따라서 인식률이 높은 신경망이 더 정확한 패턴분류를 보장하지 못한다는 실험적 결과를 고려 할 때, 인식률이 높은 두 개의 모델을 선정하여 각각의 출력에 일정한 가중치를 주고 합산하여 새로운 출력을 얻는 방법을 제안한다.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.33 no.7A
• /
• pp.733-742
• /
• 2008

In 3GPP LTE downlink system, initial cell search is essential for mobile station to connect to base station. In order to obtain information of the base station, the mobile station detects frame timing, frequency offset, and cell identification using primary synchronization channel(PSC) and secondary synchronization channel(SSC), which are defined in downlink OFDMA specification. In this paper, we analyze various detection algorithms in practical environment of inter-cell-interference, frequency offset, and multi-path fading channel and propose the optimal algorithm. Simulation results show that partial correlation method (for PSC acquisition) and interference cancellation method (for SSC detection) are the most superior algorithms among the applicable algorithms. Employ these two algorithms for receiver design, initial cell search is performed with 99% probability within 70ms in the channel environment considered.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.39C no.11
• /
• pp.1192-1200
• /
• 2014

This paper proposes a cooperative spectrum sensing algorithm in a maritime cognitive radio network and evaluates the performance of proposed algorithm. In the proposed algorithm, the secondary ships decide whether to transmit the feedback information or not on the basis of the threshold when the number of available feedback information of the ships is limited. The fusion ship detects whether the channel is available or not on the basis of the feedback information. This paper evaluates the proposed algorithm in terms of the detection probability and the number of secondary ships that are fed back in the maritime cognitive radio network. The simulation results show the proposed algorithm significantly reduces the feedback overhead even though the detection probability is somewhat declined.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.2A
• /
• pp.147-156
• /
• 2004

This paper presents an implementation and performance analysis of an access channel decoder which exploits a diversity gain due to the independent magnitude of received signals energy at each of antenna elements of a smart antenna BTS (Base-station Transceiver Subsystem) operating in CDMA2000 1X signal environment. Proposed access channel decoder consists of a searcher supporting 4 fingers, Walsh demodulator, and demodulator controller. They have been implemented with 5 of 1 million-gate FPGA's (Field Programmable Gate Array) Altera's APEX EP20K1000EBC652 and TMS320C6203 DSP (digital signal processing). The objective of the proposed access channel decoders is to enhance the data retrieval at co]1-site during the access period, for which the optimal weight vector of the smart antenna BTS is not available. Through experimental tests, we confirmed that the proposed access channel decoder exploitng the diversity technique outperforms the conventional one, which is based on a single antenna channel, in terms of detection probability of access probe, access channel failure probability, and $E_{b/}$ $N_{o}$ in Walsh demodulator.r.r.

• Journal of the Institute of Convergence Signal Processing
• /
• v.9 no.1
• /
• pp.1-9
• /
• 2008

One of the problems for implementing the spot detection phase in the 2-DE gel image analysis program is the eliminating noises in the image. Remained noises after the preprocessing phase cause the over-segmented regions by the segmentation phase. To identify and exclude the over-segmented background regions, if we use the fixed thresholding method that is choosing an intensity value for the threshold, the spots that is invisible by the eyes but mean a very small amount proteins which have important role in the biological samples could be eliminated. This paper propose an adaptive thresholding method that come from an idea that is got on statistical analysing for the prominences of the peaks. The adaptive thresholding method works as following. Firstly we calculate an average prominence value curve and fit it to exponential function curve, as a result we get parameters for the exponential function. And then we calculate a threshold value by using the parameters and probability distribution of errors. Lastly we apply the threshold value to the region for determining the region is a noise or not. According to the probability distribution of errors, the reliability is 99.85% and we show the correctness of the proposed method by representing experiment results.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.41 no.6
• /
• pp.221-229
• /
• 2004

We present new digital watermarking scheme that embeds a watermark according to the characteristics of the image or video. The scheme is compatible with established image compression standard. We define a weighting function using a parent-child structure of the DCT coefficients in a block to embed a maximum watermark. The spatio-frequency localization of the DCT coefficients can be achieved with this structure. In the detection stage, we present an optimum a posteriori threshold with a given false detection error probability based on the statistical analysis. Simulation results show that the proposed algorithm is efficient and robust against various signal processing techniques. Especially, they are robust against widely used coding standards, such as JPEG and MPEG.

• The Journal of the Acoustical Society of Korea
• /
• v.29 no.2
• /
• pp.133-140
• /
• 2010

We improve the performance of cardiac disorder classification by adding new temporal features extracted from continuous heart sound signals. We add three kinds of novel temporal features to a conventional feature based on mel-frequency cepstral coefficients (MFCC): Heart sound envelope, murmur probabilities, and murmur amplitude variation. In cardiac disorder classification and detection experiments, we evaluate the contribution of the proposed features to classification accuracy and select proper temporal features using the sequential feature selection method. The selected features are shown to improve classification accuracy significantly and consistently for neural network-based pattern classifiers such as multi-layer perceptron (MLP), support vector machine (SVM), and extreme learning machine (ELM).

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.46 no.1
• /
• pp.83-88
• /
• 2009

Selection relaying(SR) is usually based on signal-to-noise ratio(SNR) to decide whether or not to forward recovered symbols. However, instantaneous noise at relay is ignored, leading to the risk of erroneous retransmission induced by the relay that can be detrimental to the eventual detection of symbols at destination. To overcome this problem, we propose using new symbol error probability(SEP) related directly to reliability of received symbols instead of SNR. Simulation results show that the proposed SEP-based SR is considerably better than the conventional SNR-based SR under any relay position and threshold.

• Journal of the Korean Society of Propulsion Engineers
• /
• v.20 no.2
• /
• pp.56-66
• /
• 2016

Robustness is essential for model based FDI (Fault Detection and Isolation) and it is inevitable to have modeling errors and sensor signal noises during the process of FDI. This study suggests an improved method by applying NARX (Nonlinear Auto Regressive eXogenous) model and Kalman estimator in order to cope with problems caused by linear model errors and sensor signal noises in the process of fault diagnoses. Fault decision is made by the probability of the trend of gradually accumulated errors applying Fuzzy logic, which are robust to instantaneous sensor signal noises. Reliability of fault diagnosis is verified under various fault simulations.