• The Journal of the Acoustical Society of Korea
• /
• v.13 no.1
• /
• pp.24-31
• /
• 1994

In this paper, we report on the development of a speaker independent continuous speech recognition system using continuous hidden Markov models. The continuous hidden Markov model consists of mean and covariance matrices and directly models speech signal parameters, therefore does not have quantization error. Filter bank coefficients with their 1st and 2nd-order derivatives are used as feature vectors to represent the dynamic features of speech signal. We use the segmental K-means algorithm as a training algorithm and triphone as a recognition unit to alleviate performance degradation due to coarticulation problems critical in continuous speech recognition. Also, we use the one-pass search algorithm that Is advantageous in speeding-up the recognition time. Experimental results show that the system attains the recognition accuracy of $83\%$ without grammar and $94\%$ with finite state networks in speaker-indepdent speech recognition.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2018.10a
• /
• pp.369-372
• /
• 2018

This paper suggests non-parallel-voice-conversion network conversing voice between unmapped voice pair as source voice and target voice. Conventional voice conversion researches used learning methods that minimize spectrogram's distance error. Not only these researches have some problem that is lost spectrogram resolution by methods averaging pixels. But also have used parallel data that is hard to collect. This research uses PPGs that is input voice's phonetic data and a GAN learning method to generate more clear voices. To evaluate the suggested method, we conduct MOS test with GMM based Model. We found that the performance is improved compared to the conventional methods.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.22 no.3
• /
• pp.319-327
• /
• 2012

In this paper, we present a sign language recognition model which does not use any wearable devices for object tracking. The system design issues and implementation issues such as data representation, feature extraction and pattern classification methods are discussed. The proposed data representation method for sign language patterns is robust for spatio-temporal variances of feature points. We present a feature extraction technique which can improve the computation speed by reducing the amount of feature data. A neural network model which is capable of incremental learning is described and the behaviors and learning algorithm of the model are introduced. We have defined a measure which reflects the relevance between the feature values and the pattern classes. The measure makes it possible to select more effective features without any degradation of performance. Through the experiments using six types of sign language patterns, the proposed model is evaluated empirically.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.23 no.12
• /
• pp.1373-1379
• /
• 2012

The interaction of magnetic field in the frequency range of 9~150 kHz radiating from a railway system with wireless systems has been the cause of radio frequency interference. In this paper, the equivalent circuit model of the RFI noise is proposed through source and transfer path analysis, and it is confirmed that the switching noise of several kHz that occurs a vehicle traction drive system and a substation is radiated by forming the loop circuit with a feeder line by a rolling stock. And the validity of the proposed equivalent circuit model is verified by analyzing the effects of RC banks installed in the real railway between Guri and Guksu stations, the RFI noise can be effectively mitigated by loading suitable capacitance between rail and feeding line.

• Plant Journal
• /
• v.17 no.3
• /
• pp.42-46
• /
• 2021

The equation of motion of the drill string along the excavation trajectory was analyzed using the Lagrangian approach together with the finite element method (FEM). A drill string of circular cross section is constructed by combining a plurality of circular axes each having 12 degrees of freedom (DOF). FEM analysis can observe the vibration and dynamic changes of the entire drill string, and it is easy to apply comprehensive boundary conditions to reproduce the simulation of a realistic drill string. In this study, the constructed FEM motel was simulated. In order to apply the FEM program to the actual drill trajectory, the dynamic analysis of the curved beam was verified by comparison with the actual values. The dynamic change over time was observed.

• Journal of the Earthquake Engineering Society of Korea
• /
• v.13 no.5
• /
• pp.11-21
• /
• 2009

Considering that the weight of a biaxial hollow slab system is not increased with an incremental increase in its thickness, and that the flexural stiffness of a biaxial hollow slab is not significantly lower than that of a general solid slab, there has been a growing need for biaxial hollow slab systems, because long span structures are in great demand. In a long span structure, the problem of vibration of floor slabs frequently occurs, and the dynamic characteristics of a biaxial hollow slab system are quite different from the conventional floor systems. Therefore, in this study, the floor vibration of a biaxial hollow slab system subjected to walking load is investigated in comparison with a conventional floor slab system. For the efficiency of time history analysis, an equivalent plate slab model that can precisely represent the dynamic behavior of a biaxial hollow slab system is used. From the analytical results, it was determined that vibration of a biaxial hollow slab system subjected to walking load is evaluated as "office-level vibration," according to the classifications of the architectural institute of Japan and ANSI.

• Journal of KIISE:Computer Systems and Theory
• /
• v.36 no.4
• /
• pp.274-282
• /
• 2009

Various location tracking sensors; such as GPS, INS, radar, and optical equipment; are used for tracking moving targets. In order to effectively track moving targets, it is necessary to develop an effective fusion method for these heterogeneous devices. There have been studies in which the estimated values of each sensors were regarded as different models and fused together, considering the different error characteristics of the sensors for the improvement of tracking performance using heterogeneous multi-sensor. However, the rate of errors for the estimated values of other sensors has increased, in that there has been a sharp increase in sensor errors and the attempts to change the estimated sensor values for the Sensor Probability could not be applied in real time. In this study, the Sensor Probability is obtained by comparing the RMSE (Root Mean Square Error) for the difference between the updated and measured values of the Kalman filter for each sensor. The process of substituting the new combined values for the Kalman filter input values for each sensor is excluded. There are improvements in both the real-time application of estimated sensor values, and the tracking performance for the areas in which the sensor performance has rapidly decreased. The proposed algorithm adds the error characteristic of each sensor as a conditional probability value, and ensures greater accuracy by performing the track fusion with the sensors with the most reliable performance. The trajectory of a UAV is generated in an experiment and a performance analysis is conducted with other fusion algorithms.

• The Journal of the Acoustical Society of Korea
• /
• v.26 no.6
• /
• pp.250-258
• /
• 2007

Current speech recognition technology s achieved high performance with the development of hardware devices, however it is insufficient for some applications where high reliability is required, such as voice control of powered wheelchairs for disabled persons. For the system which aims to operate powered wheelchairs safely by voice in real environment, we need to consider that non-voice commands such as user s coughing, breathing, and spark-like mechanical noise should be rejected and the wheelchair system need to recognize the speech commands affected by disability, which contains specific pronunciation speed and frequency. In this paper, we propose non-voice rejection method to perform voice/non-voice classification using both YIN based fundamental frequency(F0) extraction and reliability in preprocessing. We adopted a multi-template dictionary and acoustic modeling based speaker adaptation to cope with the pronunciation variation of inarticulately uttered speech. From the recognition tests conducted with the data collected in real environment, proposed YIN based fundamental extraction showed recall-precision rate of 95.1% better than that of 62% by cepstrum based method. Recognition test by a new system applied with multi-template dictionary and MAP adaptation also showed much higher accuracy of 99.5% than that of 78.6% by baseline system.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.45 no.11
• /
• pp.15-22
• /
• 2008

The cdma2000 1xEV-DO system controls the data rates of mobile terminals based on a binary overload indicator from the base station and a simple probabilistic model. However, this control scheme has difficulty in predicting the future behavior of mobile terminals due to a probabilistic uncertainty and has no reliable means of suppressing the traffic overload, which may result in performance degradation of CDMA systems that have interference-limited capacity. This Paper proposes a new traffic control scheme that controls the data rates of mobile terminals effectively by predicting the future traffic load and adjusting the forward-link control channel. The proposed scheme is analyzed by modeling it as a multi-dimensional Markov process and compared with conventional schemes. The numerical results show that the maximum cell throughput of the proposed scheme is much higher than those of the conventional schemes.

• Journal of the korean Society of Automotive Engineers
• /
• v.18 no.6
• /
• pp.23-32
• /
• 1996

차량에서 엔진은 가장 큰 질량 집중체(concentrated mass)이다. 만약 엔진이 적절하게 구속되지 않거나 절연되어 있지 않으면, 차체에 진동을 일으키는 원인이 된다. 엔진은 다양한 진동 교란을 받는데 엔진 마운트는 이러한 모든 것들을 고립시키는 역할을 해야 하며, 엔진은 정적인 장착 하중에 대한 지지와 전후, 좌우 및 수직 방향의 운동에 대해 적절한 강성을 가져야 한다. 또한 정숙성을 향상시키기 위해서는 엔진 마운트의 재료인 고무의 강성계수를 낮추는 것이 필요한데 이는 일반적으로 내구성의 저하를 가져온다. 따라서 개발과정에서 강성계수를 낮추는 변경을 하면 부품의 내구성을 보정함에 따르는 재평가 또한 필요하게 된다. 엔진 마운트에 쓰이는 고무부품의 해석은 엔진 마운트 시스템에 대한 진동 해석 및 내구수명의 예측과 병행해야 하며, 진동해석으로부터 얻은 하중 지지 능력 등의 모든 요구 특성을 만족하기 위해서는 고무 재료의 특성에 대한 지식, 엔진 마운트의 장착 위치에 대한 결정 능력과 함께 주어진 조건에 대한 형상의 최적 설계 능력 등이 요구된다. 본 연구에서는 기본적인 형상을 파라미터화하여 엔진 마운트의 형상을 최적화 하는 절차를 제안하였다. 현재 승용차에 널리 사용되고 있는 부시형(bush type) 엔진마운트를 적용 모델로 선택하였으며, 엔진 마운트의 기본적인 형상을 몇개의 파라미터를 사용하여 정의하고 설계 사양으로 주어지는 강성값과 각 파라미터들의 조합으로 구성되는 형상이 갖는 강성값의 차이가 최소가 되도록 파라미터 값들을 최적화하였다. 최적화된 파라미터 값들로 구성되는 형상을 내구 성능, 성형성등을 고려하여 최종 형상으로 결정한다. 내구성능의 예측은 금속부품의 내구수명 예측에 널리 이용되고 있는 방법이 방진 고무부품의 경우에도 적용 가능한지를 검토하고, 방진 고무부품에도 일반적으로 적용될수 있는 내구수명 예측방안의 개발 가능성을 타진해 보았다. 본 연구의 목표는 시제품을 제작하기 이전에 설계된 부품에 대한 스프링 상수 및 내구특성을 체계적으로 규명하여 제품 시험의 횟수를 줄이고, 보다 정밀한 제품을 제작할 수 있도록 하기 위한 것이다.