• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 1995.09a
• /
• pp.349-357
• /
• 1995

본 연구에서 혼성기준 어프로치의 성과가 기존 방법론들의 성과를 선형결합 한 직선보다 상당히 아래에 나타나 기존 방법들보다 우월함을 알 수 있다. 뿐만 아니라 일부의 경우에는 배칭 어프로치보다 짧은 평균 리이드 탐임과 유연 어프로치보다 짧은 총생산지체시간을 나타내고 있다. 혼성기준을 선정 할 때 모든 상황에 대해서 효율적인 혼성기준이 있는 것이 아니라 전체시스 템의 가동률이 높고 낮음에 따라, 즉 기계 유형간 작업시간의 불균형의 정도 에 따라 유동적으로 선택하여야 한다. 시스템의 가동률이 전반적으로 낮을 때는 혼성기준을 낮게 설정 (약 40%에서 65%)하는 것이 바람직하고, 시스템 의 가동률이 높을 때는 (약 65%에서 90%정도)는 혼성기준을 높게 설정하는 것이 바람직하다.

• The Journal of the Acoustical Society of Korea
• /
• v.29 no.1
• /
• pp.41-47
• /
• 2010

The pitch alteration technique for voice conversion is classified in time domain, frequency domain and hybrid domain. The Hybrid domain method has a merit of clearness and natural-ness of pitch altered speech but has the major drawback of long processing time. In this paper, we proposed a new method that can reduce the processing time of pitch alteration in time-frequency hybrid domain. We omitted the bit-reversing process of FFT and IFFT in changing the processing domain. Therefore we can reduce the processing time by 86.26% to the conventional method with same quality.

• The KIPS Transactions:PartB
• /
• v.8B no.6
• /
• pp.699-704
• /
• 2001

ABSTRACT In this paper, a new hybrid modular recurrent TDNN (time-delay neural network)-HMM (hidden Markov model) architecture for speech recognition has been studied. In TDNN, the recognition rate could be increased if the signal window is extended. To obtain this effect in the neural network, a high-level memory generated through a feedback within the first hidden layer of the neural network unit has been used. To increase the ability to deal with the temporal structure of phonemic features, the input layer of the network has been divided into multiple states in time sequence and has feature detector for each states. To expand the network from small recognition task to the full speech recognition system, modular construction method has been also used. Furthermore, the neural network and HMM are integrated by feeding output vectors from the neural network to HMM, and a new parameter smoothing method which can be applied to this hybrid system has been suggested.

• Journal of Korean Society of Transportation
• /
• v.1 no.1
• /
• pp.82-103
• /
• 1983

본 연구는 1980년 Moses-Anas에 의해 제안된 $\ulcorner혼성로-짓모형\lrcorner$의 유효성을 서울 시의 출근통행실태 자료를 이용하여 검증하기 위한 것으로, $\ulcorner혼성로-짓모형\lrcorner$에 의한 추정 결과와 $\ulcorner전통적로-짓모형\lrcorner$에 의한 추정결과를 통계적 시점에서 상호 비교하였다. 이용된 자료는 1981년6월10일부터 6월16일까지 일주일간의 출근통행실태자료이며 조사표본수는 약 2000여 표본이었다. $\ulcorner혼성로-짓모형\lrcorner$에 의한 추정결과는 $\ulcorner전통적로-짓모형\lrcorner$ 에 의한 경 우보다 통계적으로 우월한 것으로 나타났으며, 특히 통행시간 및 통행비용에 대한 선택확률 의 탄력성은 $\ulcorner혼성모형\lrcorner$에 의해 추정된 결과가 더욱 합리적인 것으로 판명되었다.

• Proceedings of the Korea Multimedia Society Conference
• /
• 2003.05b
• /
• pp.279-282
• /
• 2003

패칭은 새로운 비디오 요청들에 대하여 이미 전송중인 동일 비디오 스트림을 공유함으로써 비디오 서버의 네트워크 대역폭 요구를 감소시키며 true VOD를 지원하는 멀티캐스트 기법이다. 본 논문에서는 탐욕 패칭과 점진 패칭의 성능을 향상시키기 위해 두 가지 방법의 장점을 취할 수 있는 혼성 패칭 방법을 제안한다. 혼성 패칭은 탐욕 패칭을 기본으로 하고 있으나 새로운 요청과 동일 비디오에 대한 최근의 정규 멀티캐스트간의 스큐가 패칭 윈도우보다는 크지만 최근의 패칭 멀티캐스트와의 스큐가 패칭 윈도우보다 작은 경우에는 최근의 패칭 멀티캐스트를 정규 멀티캐스트가 되도록 확장함으로써 새로운 요청을 위한 패칭 채널의 시간을 대폭 감소시켜 VOD 시스템의 성능을 향상시키고자 한다.

• Proceedings of the Acoustical Society of Korea Conference
• /
• autumn
• /
• pp.107-110
• /
• 2004

본 논문에서는 시간-주파수 혼성 영역 피치 검출법을 제안하였다. 음성신호를 주파수 영역으로 변환하고 주파수 영역에서 위상 성분을 조절하여 시간영역으로 역변환 함으로써 피치 피크가 최대가 되도록 하여 용이한 피치검출이 가능하였다. 또한 처리시간을 단축하기 위하여 FFT와 IFFT의 비트 재정렬을 생략하여 처리할 수 있는 방법을 제안하였다. 성능 측정 결과 기존의 켑스트럼 검출법에 비하여 검출성능은 우수하면서도 처리시간은 $84.8\%$로 단축됨을 알 수 있었다.

• Journal of the Korea Fashion and Costume Design Association
• /
• v.15 no.1
• /
• pp.59-74
• /
• 2013

Post-modern society leads us to accept diversity and variability instead of pursuit of the absolute truth, beauty or classical value systems, thus leading to hybrid phenomena. The purpose of this study is to analyze characteristics and expressive effects of hybrid expressions through which to provide conceptual bases for interpreting expanded meanings of fashion illustrations that express aesthetic concepts of hybrid culture. Hybrid refers to a condition on which diverse elements are mixed with each other, so any one element can not dominate others. It is often used to create something unique and new by a combination of unprecedented things. Hybrid can be classified into four categories: temporal hybrid, spatial hybrid, morphological hybrid and hybrid of different genres. Temporal hybrid from a combination of past and present in fashion illustration includes temporal blending by repetition and juxtaposition. Spatial hybrid shows itself in the form of inter-penetration and interrelationship by means of projection, overlapping, juxtaposition and multiple space. Morphological hybrid expresses itself through combination of heterogenous forms and restructuring of deformed forms. Hybrid of different genres in fashion illustration applies various graphic elements or photos within the space, and represents blending of arts and daily living. Such hybrid expressions in fashion illustrations reflect the phenomena of diversity and variability of post-modern society. Hybrid expressions in fashion illustrations predict endless possibility of expressing new images through combining various forms or casual elements and can develop toward a new creative technique.

• Investigative Magnetic Resonance Imaging
• /
• v.6 no.1
• /
• pp.64-72
• /
• 2002

Purpose : To introduce and demonstrate the advantages of the new hybrid two-dimensional (2D) proton spectroscopic imaging (SI) over the single voxel spectroscopy (SVS) and conventional 2D SI in the clinical application of spectroscopy for pediatric cerebral disease. Materials and Methods : Eighty-one hybrid 2D proton spectroscopic imaging was performed in 79 children (36 normal infants and children, 10 with hypoxic-ischemic injury, 20 with toxic-metabolic encephalopathy, seven with brain tumor, three with meningoencephalitis, one with neurofibromatosis, one with Sturge-Weber syndrome and one with lissencephaly) ranging in age from the third day of life to 15 years. In adult volunteers (n=5), all three techniques including hybrid 2D proton SI, SVS using PRESS sequence, and conventional 2D proton SI were performed. Both hybrid 2D proton SI and SVS using PRESS sequence were performed in clinical cases (n=). All measurements were performed with a 1.5-T scanner using standard head quadrature coil. The 16$\times$16 phase encoding steps were set on variable field of view (FOV) depending on the size of the brain. The hybrid volume of interest inside FOV was set as $75{\times}75{\times}15{\;}\textrm{mm}^3$ or smaller to get rid of unwanted fat signal. Point-resolved spectroscopy (TR/TE=1,500 msec/135 or 270msec) was employed with standard chemical shift selective saturation (CHESSI pulses for water suppression. The acquisition time and spectral quality of hybrid 2D proton SI were compared with those of SVS and conventional 2D proton SI. Results : The hybrid 2D proton SI was successfully conducted upon all patients.

• Journal of the korean academy of Pediatric Dentistry
• /
• v.30 no.2
• /
• pp.263-271
• /
• 2003

The purpose of the present study was to evaluate the patterns of hybrid layer according to the application time and the frequency and its effects on the shear bonding strength of All-In-One system in primary teeth. A single bonding agent(Scotchbond Multi-Purpose Plus, 3M) and an All-In-One system(Prompt L-pop, 3M ESPE) were applied on the dentin varying the application time and the frequency in primary teeth. Shear bond strength was measured and the patterns of hybrid layers were observed by SEM. The following results were obtained ; 1. The shear bonding strength of single bonding agent was significantly higher than that of All-In-One system(P<0.05). 2. The shear bonding strength of All-In-One system applied twice or 3 times were higher than that of applied once (P<0.05). And thickness of the hybrid layer was increased when applied twice or 3 times compared to once. 3. The shear bonding strength of All-In-One system when applied for 15 second and 30 second were higher than that of 7 second (P<0.05). And the hybrid layer thickness of 15 second and 30 second's application time were higher than that of 7 second. 4. Thickness of hybrid layer applied with single bonding agent was $2-4{\mu}m$ and that of All-In-One system was $1-2{\mu}m$.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.17 no.11
• /
• pp.1185-1198
• /
• 1992

In this paper, the differential weighted algorithm proposed in order to improve th noise smoothing characteristics of conventional Median filter and FIR-Median Hybrid filter. Performance of some image restoration filter(median filter, FIR-Median Hybird filter, FIR-Median Hybrid filter to proposed differential weighted algorithm) are compared and evaluated on the noise smoothing characteristics and sharp edge conservation characteristics. Test and Real images used in this paper are Lenna and Urological images corrupted by impulse, gaussian, exponential and laplacian noise. Experimental results show that the FIR-Median Hybrid filter applied to the differential weighted algorithm are comparatively superior to others. But the filter orders have increased, the more time consumed to image processing. Hence if the adequate filtering by the type of image is selected. now after a great support will be take consideration into the various parts of application by computer science and of medical image processing.