http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Oh, I.,Jo, J.,Park, J.,Lee, J.,Shin, H.,Yoo, J.W. Elsevier 2016 CURRENT APPLIED PHYSICS Vol.16 No.11
<P>Glassy carbon can be used in a variety of harsh environments because it exhibits low electrical resistance, high thermal stability, chemically impermeability and hardness. In this work, we studied the potential application of the pyrolized glassy carbon as a thermoelectric material by electron beam irradiation. Irradiation with electron beam energy of 1.4 MeV and 0.2 MeV triggered local crystallization of glassy carbon, which was determined by a Raman spectroscopy. As the irradiation time was increased, the number of sp(2) hybridized carbons was gradually increased while the number of disordered carbons was decreased. This electron beam induced local crystallization caused a synergetic effect on a thermoelectric power factor of the studied glassy carbon films. Although the electrical conductivity was decreased by the e-beam irradiation, the Seebeck coefficient was significantly enhanced resulting in an improved power factor. The optimal power factor we obtained for the irradiated glassy carbon was similar to 50% higher than that of the non-irradiated sample. (C) 2016 Elsevier B.V. All rights reserved.</P>
오광식,박외정 대구효성가톨릭대학교 1994 연구논문집 Vol.49 No.1
In regression diagnostics, a number of outlier measures based on various statistical tools have been discussed. These measures usually pay·attention to dectecting individual outlier. However, there exists a situation in which observations are multiple but not individual outliers or the other way about, The situation is revered to as a masking effect which means that the outlier of one observation is masked by the presence of another observation. The multiple outlier is often overlooked because it is much more difficult to detect than the individual outliers. We propose a new expression of John and Draper's outlier sum of square. The advantages from the theoretical point of view are that the expression is based not only on the multiple case but also on the single case, and that we can detect multiple outliers with the masking effect by investigating the difference between the multiple outliers and sum of the individual outliers. Furthermore we suggest a search method for multiple outliers after investigating some properties of the new expression. .
An Analysis of Shin keikō haiku and Dentō ha haiku from the Fundamental Informatics Perspective
Ohi, Nami World Association for Triple Helix and Future Stra 2013 Journal of Contemporary Eastern Asia Vol.12 No.2
This study aims to understand haiku phenomena from the viewpoint of communication, which emerges from reciprocal relationships among haiku poets, haiku societies, media on haiku, and other establishments relating to haiku. Such an analytic point of view, which tries to consider both the operation of each autonomous agent and that of the whole system consisted by the agents, is included in the realm of second-order cybernetics. The operation of a system cannot be reduced completely to a system's individual components, which is why such a viewpoint is required. Fundamental informatics, which is employed as a theoretical framework, and two haiku movements, which include shin keiko haiku ("new-trend" haiku) and dento ha haiku (Hototogisu-school haiku), forming an important part of modern haiku history, are the focus of this study. As a result, modern haiku history is considered to be an evolution of a haiku system, whose main incentive is an awareness of the production mechanism of haiku communication occurring through second-order observations in the system. This study also illuminates how haiku poets, haiku societies, and media covereage of haiku play roles in the evolution of the haiku system.
HIROSHI OHI,TOMOYA YOKOYAMA 한국펄프·종이공학회 2006 한국펄프종이학회 기타 간행물 Vol.- No.-
The pulp yield was improved by about 4.5-5% when polysulfide (PS) and anthraquinone (AQ) were added to the kraft cooking liquor (white liquor). The exchange of the black liquor with fresh white liquor further increased the yield. The highest pulp yield was obtained when the PS cooking liquor containing 70% of total active alkali (AA) and 100% of AQ was used from the beginning of the reaction and the black liquor was exchanged with fresh white liquor containing the residual 30% of AA just after temperature reached 135℃. There was a good correlation between kraft pulp yields of a hardwood species and the ratios of the amount of xylose to glucose (X/G ratio), liberated by an acid hydrolysis of the pulps. However, the correlation was dependent on raw material wood species. Therefore, it is required in advance to establish a correlation between the yields and X/G ratios for raw material wood species of a target pulp in order to estimate pulp yield using X/G ratio. The X/G ratios of relatively high yield pulps showed higher values than those expected from the correlation. In a mill trial, the superiority of the PS-AQ isothermal cooking (ITC) process over the kraft ITC process was confirmed by examining X/G ratio of pulps obtained. The pulp yield in the PS-AQ ITC process was estimated at about 57.0%. This yield is very high, which indicates that reaction conditions of the PS-AQ ITC process are optimal.
A Lightweight Speaker Recognition System Using Timbre Properties
Abu Quwsar Ohi,M. F. Mridha,Md. Abdul Hamid,Muhammad Mostafa Monowar,Dongsu Lee,Jinsul Kim 한국디지털콘텐츠학회 2020 The Journal of Contents Computing Vol.2 No.1
Speaker recognition is an active research area that contains notable usage in biometric security and authentication system. Currently, there exist many well-performing models in the speaker recognition domain. However, most of the advanced models implement deep learning that requires GPU support for real-time speech recognition, and it is not suitable for low-end devices. In this paper, we propose a lightweight text-independent speaker recognition model based on random forest classifier. It also introduces new features that are used for both speaker verification and identification tasks. The proposed model uses human speech based timbral properties as features that are classified using random forest. Timbre refers to the very basic properties of sound that allow listeners to discriminate among them. The prototype uses seven most actively searched timbre properties, boominess, brightness, depth, hardness, roughness, sharpness, and warmth as features of our speaker recognition model. The experiment is carried out on speaker verification and speaker identification tasks and shows the achievements and drawbacks of the proposed model. In the speaker identification phase, it achieves a maximum accuracy of 78%. On the contrary, in the speaker verification phase, the model maintains an accuracy of 80% having an equal error rate (ERR) of 0.24.