• 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 Korea Society for Energy Engineering kosee Conference
• /
• 2010.04a
• /
• pp.44-44
• /
• 2010

대부분의 산업분야에서 사용되는 공업로는 연소 시 $1000^{\circ}C$이상의 고온 배기가스가 배출되고 있기 때문에, 이들 배기가스로부터 폐열을 회수하여 에너지를 절감하고자 열교환기를 사용하고 있다. 하지만 현재 사용되고 있는 금속 소재의 열교환기의 경우 사용온도가 $800^{\circ}C$로 제한되어 있으며, 산화, 부식, creep등의 문제가 대두되고 있다. 본 연구에서는 세라믹 소재의 열교환기를 제작하여 열전달율, 공기누설율등을 평가하는 성능실험을 하였다. 실험방법은 연소시작 후 정상상태에 도달하기까지의 경향을 알아보고 공기유량과 가스유량에 변화를 주었을 때, 열전달율과 공기누설율등을 평가하고 이론적인 데이터와 비교하였다. 실험 결과, $1300^{\circ}C$이상의 환경에서도 정상적으로 작동하였으며, 외부환경조건을 고려하였을 때, 이론적인 데이터와 근접한 데이터를 얻을 수 있었다. 그러나 대류열전달에 의한 열손실과 공기누설등이 문제점으로 지적되었다. 향후 연구에서는 대류열전달에 의한 열손실을 고려한 연구와 효과적인 씰링으로 누설을 줄일 수 있는 실험장치를 개발할 필요가 있다.

• Journal of the Korea Convergence Society
• /
• v.8 no.11
• /
• pp.103-108
• /
• 2017

In the ubiquitous environment, there has been a problem of transmitting data from various sensors at a long distance. Especially, in the process of integrating data arriving at different locations, data having different property values of data or having some loss in data had to be processed. This paper present a method to analyze such data. The core of this method is to define an objective function suitable for the problem and to develop an algorithm that can optimize this objective function. The objective function is used by modifying the OCS function. MFA (Mean Field Annealing), which was able to process only binary data, is extended to be applicable to fields with continuous values. It is called CMFA and used as an optimization algorithm.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.53 no.3
• /
• pp.99-106
• /
• 2016

In this paper, we will discuss about correction method of missing data on noisy observation matrix and uncertainty analysis for the potential noise. In situations without missing data in an observation matrix, this solution is known to be accurately induced by SVD (Singular Value Decomposition). However, usually the several entries of observation matrix have not been observed and other entries have been perturbed by the influence of noise. In this case, it is difficult to find the solution as well as cause the 3D reconstruction error. Therefore, in order to minimize the 3D reconstruction error, above all things, it is necessary to correct reliably the missing data under noise distribution and to give a quantitative evaluation for the corrected results. This paper focuses on a method for correcting missing data using geometrical properties between 2D projected object and 3D reconstructed shape and for estimating a noise level of the observation matrix using ranks of SVD in order to quantitatively evaluate the performance of the correction algorithm.

• Journal of KIISE:Information Networking
• /
• v.28 no.3
• /
• pp.389-405
• /
• 2001

In these days, we have increasing demands on the real-time services, especially for the multimedia data transmission in both of wired and wireless environments and thus efficient and stable ways of transmitting realtime data are needs. Although RTP is widely used for internet-based realtime applications, it cannot avoid packet losses, due to the use of UDP stack and its underlying layers. In the case of mobile computing applications, the packet losses are more frequent and consecutive because of the limited bandwidth. In this paper, we first statistically analyze the characteristics of packet losses in the wired and wireless communications, based on Gilbert model, and a new packet recovery scheme for realtime data transmission is presented. To reflect the transmission characteristics of the present network environment, our scheme makes the sender to dynamically adjust the amount of redundant information, using the current packet loss characteristic parameters reported by the receiver. Additionally, we use relatively large and discontinuous offset values, which enables us to recover from both of the random and consecutive packet losses. Due to these characteristics, our scheme is suitable for the mobile computing environment where packet loss rates are relatively high and varies rapidly in a wide range. Since our scheme is based on the analytic model form statistics, it can also be used for other network environments. We have implemented the scheme with Mobile IP and RTP/RTCP protocols to experimentally verify its efficiency.

• Fire Science and Engineering
• /
• v.21 no.3
• /
• pp.91-96
• /
• 2007

Performance analysis of the butterfly valve in water fire extinguishing has been carried out. Performance analysis of the butterfly valve are investigated for torque characteristics, pressure loss and cavitations. The torque characteristics of disc are corrected for the angles of attack of valve disc by theoretical torque equation, and correction equation is added. The pressure loss coefficient on opening angle of valve has been formulated by applying the Carnot's equations. The torque characteristics, pressure loss and cavitations of the butterfly valve are analyzed for the ratio of disc thickness to the valve diameter. Cavitations are analyzed from the pressure loss coefficient of valve. The analysis of pressure loss and cavitation has been carried out change of the thickness ratio on opening angle of valve. These analysis data are utilize to necessary engineering data to develope of the butterfly valve.

• Journal of the Korea Convergence Society
• /
• v.9 no.12
• /
• pp.9-15
• /
• 2018

The main goal of extended data expression is to develop a data structure suitable for common problems in ubiquitous environments. The greatest feature of this method is that the attribute values can be represented with probability. The next feature is that each event in the training data has a weight value that represents its importance. After this data structure has been developed, an algorithm has been devised that can learn it. In the meantime, this algorithm has been applied to various problems in various fields to obtain good results. This paper first introduces the extended data expression technique, UChoo, and rule refinement method, which are the theoretical basis. Next, this paper introduces some examples of application areas such as rule refinement, missing data processing, BEWS problem, and ensemble system.

• Journal of KIISE:Computer Systems and Theory
• /
• v.36 no.5
• /
• pp.360-370
• /
• 2009

Using solar power in wireless sensor networks requires a different approach to energy optimization from networks with battery-based nodes. Solar energy is an inexhaustible supply which can potentially allow a system to run forever, but there are several issues to be considered such as the uncertainty of energy supply and the constraint of rechargeable battery capacity. In this paper, we present SolarSS: a reliable storage system for solar-powered sensor networks, which provides a set of functions, in separate layers, such as sensory data collection, replication to prevent failure-induced data loss, and storage balancing to prevent depletion-induced data loss. SolarSS adapts the level of layers activated dynamically depending on solar energy availability, and provides an efficient resource allocation and data distribution scheme to minimize data loss.

• Proceedings of the Korea Multimedia Society Conference
• /
• 2003.11a
• /
• pp.429-432
• /
• 2003

최근 TCP와 유사하게 멀티미디어 데이터를 전송하기 위한 rate control 프로토콜에 관한 연구가 활발하게 진행되고 있다. 본 논문에서는 멀티미디어 전송 방식에 있어서 TCP와 공정성을 유지하며 TCP와 유사하게 동작하도록 하는 TRCP 프로토콜을 제안한다. TRCP는 TCP vegas 이론을 응용하였으며 RTT와 패킷손실율에 의해 네트워크 혼잡 상태를 미리 예측하고 TCP의 AIMD 방식을 사용하여 TCP와 유사하게 전송율을.조절하는 프로토콜이다. 이 프로토콜에 대한 TCP와 공정성을 검증하기 위하여 시뮬레이션을 실시하고 그 결과를 분석한다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2003.10a
• /
• pp.160-162
• /
• 2003

색인 파일은 정보 검색 시스템에서 중요한 구성 요소 중에 하나이다. 스테밍을 하여 색인 파일을 구성하게 되면 파일의 크기를 줄일 수 있고 재현율을 높이는데 효과적이다. 하지만, 과도한 스테밍으로 구성이론. 색인 파일은 원형에 대한 데이터 손실을 가지고 오기 때문에 너무 많은 문서가 검색되어 사용자가 문서를 찾는데 많은 시간이 소요되고 정확률도 떨어진다. 본 논문에서는 정보 검색 시스템에서 검색의 효율성을 높이기 위해 사용하는 색인 파일을 스테밍 한 것과 스테밍 하지 않은 파일로 구성하였다. 스테밍 한 색인 파일은 질의어와 문서 사이의 유사도를 계산하기 위하여 이용되며, 스테밍 하지 않은 파일은 스테밍 했을 때 검색된 문서들 중에서 데이터 손실로 인한 잘 못된 문서 순서를 재조정해 주기 위하여 이용된다. 본 논문에서는 높은 검색 효과를 제공하는 기존의 벡터 공간 모델을 검색 성능 평가 척도 중의 하나인 R-정확률을 이용하여 비교 평가하였다. 본 논문에서 제안하는 시스템이 문서 상위 100위까지에 대하여 일반 벡터 모델 보다 최고 21%의 좋은 성능을 보였다.