• Journal of Intelligence and Information Systems
• /
• v.24 no.1
• /
• pp.205-225
• /
• 2018

Convolutional Neural Network (ConvNet) is one class of the powerful Deep Neural Network that can analyze and learn hierarchies of visual features. Originally, first neural network (Neocognitron) was introduced in the 80s. At that time, the neural network was not broadly used in both industry and academic field by cause of large-scale dataset shortage and low computational power. However, after a few decades later in 2012, Krizhevsky made a breakthrough on ILSVRC-12 visual recognition competition using Convolutional Neural Network. That breakthrough revived people interest in the neural network. The success of Convolutional Neural Network is achieved with two main factors. First of them is the emergence of advanced hardware (GPUs) for sufficient parallel computation. Second is the availability of large-scale datasets such as ImageNet (ILSVRC) dataset for training. Unfortunately, many new domains are bottlenecked by these factors. For most domains, it is difficult and requires lots of effort to gather large-scale dataset to train a ConvNet. Moreover, even if we have a large-scale dataset, training ConvNet from scratch is required expensive resource and time-consuming. These two obstacles can be solved by using transfer learning. Transfer learning is a method for transferring the knowledge from a source domain to new domain. There are two major Transfer learning cases. First one is ConvNet as fixed feature extractor, and the second one is Fine-tune the ConvNet on a new dataset. In the first case, using pre-trained ConvNet (such as on ImageNet) to compute feed-forward activations of the image into the ConvNet and extract activation features from specific layers. In the second case, replacing and retraining the ConvNet classifier on the new dataset, then fine-tune the weights of the pre-trained network with the backpropagation. In this paper, we focus on using multiple ConvNet layers as a fixed feature extractor only. However, applying features with high dimensional complexity that is directly extracted from multiple ConvNet layers is still a challenging problem. We observe that features extracted from multiple ConvNet layers address the different characteristics of the image which means better representation could be obtained by finding the optimal combination of multiple ConvNet layers. Based on that observation, we propose to employ multiple ConvNet layer representations for transfer learning instead of a single ConvNet layer representation. Overall, our primary pipeline has three steps. Firstly, images from target task are given as input to ConvNet, then that image will be feed-forwarded into pre-trained AlexNet, and the activation features from three fully connected convolutional layers are extracted. Secondly, activation features of three ConvNet layers are concatenated to obtain multiple ConvNet layers representation because it will gain more information about an image. When three fully connected layer features concatenated, the occurring image representation would have 9192 (4096+4096+1000) dimension features. However, features extracted from multiple ConvNet layers are redundant and noisy since they are extracted from the same ConvNet. Thus, a third step, we will use Principal Component Analysis (PCA) to select salient features before the training phase. When salient features are obtained, the classifier can classify image more accurately, and the performance of transfer learning can be improved. To evaluate proposed method, experiments are conducted in three standard datasets (Caltech-256, VOC07, and SUN397) to compare multiple ConvNet layer representations against single ConvNet layer representation by using PCA for feature selection and dimension reduction. Our experiments demonstrated the importance of feature selection for multiple ConvNet layer representation. Moreover, our proposed approach achieved 75.6% accuracy compared to 73.9% accuracy achieved by FC7 layer on the Caltech-256 dataset, 73.1% accuracy compared to 69.2% accuracy achieved by FC8 layer on the VOC07 dataset, 52.2% accuracy compared to 48.7% accuracy achieved by FC7 layer on the SUN397 dataset. We also showed that our proposed approach achieved superior performance, 2.8%, 2.1% and 3.1% accuracy improvement on Caltech-256, VOC07, and SUN397 dataset respectively compare to existing work.

• Journal of The Korean Association For Science Education
• /
• v.32 no.1
• /
• pp.15-29
• /
• 2012

The purpose of this study was to investigate the educational factors on career choice of science-gifted students to science and technology bound universities and the difference of perception in regards to group factors. In addition, this study aimed to examine the effects of science-gifted education and critical events in relation to career choice to science and technology bound universities. For the study, 104 university freshmen, 75 males and 29 females, were sampled from UNIST (Ulsan National Institute of Science and Technology), that many science high school graduates entered this year. The survey was conducted with questionnaires to do with the perceptions concerning career choice and educational factors that cause them to choose such career directions. The educational factors on career choice to science and technology bound universities were classified as 3 main categories such as educational environment factor (teaching-learning factor), human factor, attitude towards science factor and the subcategories within each category. The research findings are as follows: First, the factors were closely connected with each other and 'the project centered classes' were highly interrelated with other educational environment factors such as 'the experiment activity and environment for the activity' and 'influence of teachers (professors).' Second, the female students and graduates of the science high school were more positively influenced by the educational environment and human factors on their decision for career than male students and graduates of the general high school. Third, this research found that historical scientific knowledge, perception of scientists' social status and job applications in the science field gave less influence rather than other factors on their decision for career. As a result of examining critical events for science-gifted education in relation to career choice to science and technology bound universities, numerous students mentioned that the extracurricular science activities such as science camps and field trips gave significant effects on students' career choices to science and engineering fields.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.23 no.4
• /
• pp.16-28
• /
• 2009

Stage art basically builds upon the essence of "seeing," and at the same time, possesses relativity in showing and seeing. Stage lighting uses artificial light to solve the essence of "seeing", which is the foundation of stage art, and coming into the modern age, its role has been enhanced to an important medium for visual expression in stage art, due to the lighting tools that developed at a rapid pace along with the discovery of electricity, as well as the development of optics. Therefore, not only does lighting use a medium known as light in a field of stage art that gives mental and emotional inspiration to the audience, and aesthetically expresses time and space. In other words, stage lighting is a complex function of light engineering (technology and science) and aesthetic sense (feeling and art). This study aims to do research on methods for the visualization of stage space through lighting, mainly focused on dancing. I have studied the basics of stage lighting, its relations with other fields of stage art, and the functions and characteristics of lighting. Results show that lighting could be used to maximize the visualization of dancing and emphasizing the artistic growth of lighting and its ability to aesthetically express and I came to the following conclusions. First, lighting uses the forms and directions of light that various tools are able to produce in order to visualize the space on stage, and can maximally express the image that the work seeks. Second, it is possible to use lighting, through the movement of light, as a visual representation of the configuration of space in dancing works. Third, through the expression of visual and spatial aspects created by light, the work's dramatic catharsis can bring out mental and emotional feelings form the audience. Fourth, lighting can be seen not as a supporting role, but as an original visual design. To conclude, in order for lighting to be freed form the simple function of "lighting up the stage," which a majority of people think is common knowledge, and grow as one area in art, lighting designers must understand the intentions of the choreographer and the work with creativity and artistry they must consider light and color as an aesthetic language in order to heighten the effects of the work and allow it to partake as one element of work creation, so that lighting will be treated as a form of art.

• Journal of the Korean Institute of Landscape Architecture
• /
• v.37 no.2
• /
• pp.14-25
• /
• 2009

We live in an age which is exponentially growing as the knowledge paradigm is changing. New sites are subject to contemporary landscape architecture function as "fields" in which this hybrid aspect is both actively practiced and becoming a catalyst for change in the area of landscape architecture. With this as its background, this study attempts to deal with how the aspect of integration in environmental design is manifested. For this purpose, the tendencies for the discussion of integration in various fields of practice were examined: planning theories, urban theories, architecture, public environment, engineering, and landscape architecture. As yet, the discussions of interdisciplinary integration, which occur in practice in these respective fields, mainly tend to be oriented toward the effective implementation of the merits of other related fields. Seen from these examples of practice, integrated design approaches can be found in the following three aspects: design objects, respective professional areas, and methodologies of approaches and design. In terms of design objects, the positions of individual design subjects present themselves as most obvious, and integration or combination of the physical targets that come to exist through design can be easily seen. Most examples of integration turn out to be this, in almost every case of which the theme and the target of expression are integrated via a small number of certain methods. In terms of professional areas, what can be mainly evidenced is how the individual subject acts when the subject designs. The strong points of professionals from each field seem to create synergy, achieving through integration optimum results. In terms of methodologies of approaches and design, there are attempts to create integrated approaches as ways of effective decision-making, in which case the integration of all of the interest parties is of primary concern. As yet, few instances have been found in which integrated design has had enough strength to be seen as a concrete design methodology based on practical examples. However, it is encouraging that theoretical approaches and the necessity for integrated design have been identified from multiple perspectives, and that a practical movement such as landscape urbanism has come into active being. The authors of this study find this point in time to be ripe for discussions on integrated practices in terms of environmental design, on the basis of the synthetic approaches mentioned above.

• Journal of Intelligence and Information Systems
• /
• v.17 no.4
• /
• pp.227-240
• /
• 2011

New concepts and ideas often result from extensive recombination of existing concepts or ideas. Both researchers and developers build on existing concepts and ideas in published papers or registered patents to develop new theories and technologies that in turn serve as a basis for further development. As the importance of patent increases, so does that of patent analysis. Patent analysis is largely divided into network-based and keyword-based analyses. The former lacks its ability to analyze information technology in details while the letter is unable to identify the relationship between such technologies. In order to overcome the limitations of network-based and keyword-based analyses, this study, which blends those two methods, suggests the keyword network based analysis methodology. In this study, we collected significant technology information in each patent that is related to Light Emitting Diode (LED) through text mining, built a keyword network, and then executed a community network analysis on the collected data. The results of analysis are as the following. First, the patent keyword network indicated very low density and exceptionally high clustering coefficient. Technically, density is obtained by dividing the number of ties in a network by the number of all possible ties. The value ranges between 0 and 1, with higher values indicating denser networks and lower values indicating sparser networks. In real-world networks, the density varies depending on the size of a network; increasing the size of a network generally leads to a decrease in the density. The clustering coefficient is a network-level measure that illustrates the tendency of nodes to cluster in densely interconnected modules. This measure is to show the small-world property in which a network can be highly clustered even though it has a small average distance between nodes in spite of the large number of nodes. Therefore, high density in patent keyword network means that nodes in the patent keyword network are connected sporadically, and high clustering coefficient shows that nodes in the network are closely connected one another. Second, the cumulative degree distribution of the patent keyword network, as any other knowledge network like citation network or collaboration network, followed a clear power-law distribution. A well-known mechanism of this pattern is the preferential attachment mechanism, whereby a node with more links is likely to attain further new links in the evolution of the corresponding network. Unlike general normal distributions, the power-law distribution does not have a representative scale. This means that one cannot pick a representative or an average because there is always a considerable probability of finding much larger values. Networks with power-law distributions are therefore often referred to as scale-free networks. The presence of heavy-tailed scale-free distribution represents the fundamental signature of an emergent collective behavior of the actors who contribute to forming the network. In our context, the more frequently a patent keyword is used, the more often it is selected by researchers and is associated with other keywords or concepts to constitute and convey new patents or technologies. The evidence of power-law distribution implies that the preferential attachment mechanism suggests the origin of heavy-tailed distributions in a wide range of growing patent keyword network. Third, we found that among keywords that flew into a particular field, the vast majority of keywords with new links join existing keywords in the associated community in forming the concept of a new patent. This finding resulted in the same outcomes for both the short-term period (4-year) and long-term period (10-year) analyses. Furthermore, using the keyword combination information that was derived from the methodology suggested by our study enables one to forecast which concepts combine to form a new patent dimension and refer to those concepts when developing a new patent.

• Journal of Mushroom
• /
• v.12 no.2
• /
• pp.132-137
• /
• 2014

Green mold disease caused by Trichoderma species has recently caused considerable damage to oyster mushroom industries in Korea. This disease Trichoderma, Penicillium, Aspergillus, such as in (genus) to be included in a disease caused by a species that collectively the largest incidence and damage is caused by the pathogen Trichoderma genus. T. longibrachiatum, Trichoderma koningii, Trichoderma virens, T. hazianum, T. atroviride, and T. pseudokoningii were detected on oyster mushroom beds and, of them, T. virens, T. hazianum, T. longibrachiatum was the most frequently detected. The knowledge concerning physiological and ecological properties of Trichoderma spp. was essential for their effective control. T. longibrachiatum hyphal growth is very fast, spore formation, and, particularly well-chlamydospore formation characteristics, and reviews are dark green discoloration. T. koningii, fast mycelial growth, aerial hyphae and spores in aerial hyphae formation is concentrated. T. virens, especially if the color change caused by spore-forming, slow, late in infection, the more severe the damage is discovered. T. hazianum fast mycelial growth, white aerial hyphae and late turns dark green. After spore formation hyphae glob of white pustules or tufts on the top of the formation. T. atroviride. aerial hyphae usually the mycelial growth and spore formation in the unlikely event of the formation and smells similar to the smell of coconut is that. Fast T. pseudokoningii mycelial growth, spore formation is formed around the inoculation site, discoloration of the medium color and well formed chlamydospores.

• Proceedings of the Korean Society of Soil and Groundwater Environment Conference
• /
• 1997.11a
• /
• pp.3-31
• /
• 1997

Water has always played a significant role in the lives of people. In urbanised Rome, with its million people. sophisticated supply systems developed and then fled with the empire. only to be rediscovered later But it was the industrial Revolution commencing in the eighteenth century that ushered in major paradigm shifts In use and altitudes towards water. Rapid and concentrated urbanisation brought problems of expanded demands for drinking supplies, waste management and disease. The strategy of using water from local streams, springs and village wells collapsed under the onslaughts of rising urban demands and pollution due to poor waste disposal practices. Expanding travel (railways. and steamships) aided the spread of disease. In England. public health crises peaks, related to water-borne typhoid and the three major cholera outbreaks occurred in the late eighteenth and early nineteenth century respectively. Technological, engineering and institutional responses were successful in solving the public health problem. it is generally accepted that the putting of water into pipe networks both for a clean drinking supply, as well as using it as a transport medium for removal of human and other wastes, played a significant role in towering death rates due to waterborne diseases such as cholera and typhoid towards the end of the nineteenth century. Today, similar principles apply. A recent World Bank report Indicates that there can be upto 76％ reduction in illness when major water and sanitation improvements occur in developing countries. Water management, technology and thinking in Australia were relatively stable in the twentieth century up to the mid to late 1970s. Groundwater sources were investigated and developed for towns and agriculture. Dams were built, and pipe networks extended both for supply and waste water management. The management paradigms in Australia were essentially extensions of European strategies with the minor adaptions due to climate and hydrogeology. During the 1970s and 1980s in Australia, it was realised increasingly that a knowledge of groundwater and hydrogeological processes were critical to pollution prevention, the development of sound waste management and the problems of salinity. Many millions of dollars have been both saved and generated as a consequence. This is especially in relation to domestic waste management and the disposal of aluminium refinery waste in New South Wales. Major institutional changes in public sector water management are occurring in Australia. Upheveals and change have now reached ail states in Australia with various approaches being followed. Market thinking, corporatisation, privatisation, internationalisation, downsizing and environmental pressures are all playing their role in this paradigm shift. One casualty of this turmoil is the progressive erosion of the public sector skillbase and this may become a serious issue should a public health crisis occur such as a water borne disease. Such crises have arisen over recent times. A complete rethink of the urban water cycle is going on right now in Australia both at the State and Federal level. We are on the threshold of significant change in how we use and manage water, both as a supply and a waste transporter in Urban environments especially. Substantial replacement of the pipe system will be needed in 25 to 30 years time and this will cost billions of dollars. The competition for water between imgation needs and environmental requirements in Australia and overseas will continue to be an issue in rural areas. This will be especially heightened by the rising demand for irrigation produced food as the world's population grows. Rapid urbanisation and industrialisation in the emerging S.E Asian countries are currently producing considerable demands for water management skills and Infrastructure development. This trend e expected to grow. There are also severe water shortages in the Middle East to such an extent that wars may be fought over water issues. Environmental public health crises and shortages will help drive the trends.

• Journal of Korean Society for Atmospheric Environment
• /
• v.10 no.E
• /
• pp.311-324
• /
• 1994

The quantitative knowledge of N $O_{\gamma}$ (＝N $O_{x}$ ＋HN $O_3$＋/PAN＋N $O_3$ ＋$N_2$ $O_{5}$ ＋HN $O_2$＋N $O_3$$^{-10}$ ＋organic nitrates＋......)distribution is essential in tropospheric chemistry, especially, especially that related to understanding the processes leading to ozone production. Ambient concentrations of NO, N $O_2$, HN $O_3$ and PAN as well as total N $O_{\gamma}$ were measured during June and early July 1992 at a rural site(Candor, NC), in the central Piedmont region of NC. The measurements of N $O_{\gamma}$ species were made in an effort to provide a comprehensive understanding of nitrogen chemistry and to investigate the total nitrogen budget at the site. N $O_{\gamma}$, N $O_2$, and NO showed diurnal variations with maxima in the morning. The maximum N $O_{\gamma}$ concentration reached was 14.5 ppbv, and the maximum concentrations of NO and N $O_2$ were 5.4 and 7.8 PPbv, respectively. The mean N $O_{\gamma}$ concentration as found to be 2.88$\pm$1.58 ppbv(n=743). The mean concentrations of NO and N $O_2$, were found to be 0.15 $\pm$ 0.29 ppbv(n＝785) and 1.31 $\pm$ 0.99 ppbv(n＝769). Products of photochemical oxidants, (N <$O_{\gamma}$-N $O_{x}$ ), such as HN $O_3$ and PAN, as well as ozone showed diurnal variation with maxima in the afternoon and minima at night The fractions of individual reactive nitrogen species to total N $O_{\gamma}$ were investigated and contrasted to the results from remote marine site and rural continental sites. N $O_{x}$ was the major species to total N $O_{\gamma}$(45%). NO concentrations appeared to be nearly constant whether the Prevailing winds were from continental areas or from oceanic areas. Linear regression of $O_3$ with (N $O_{\gamma}$- N $O_{x}$ )/N $O_{\gamma}$ (i.e. percent N $O_{x}$ converted to the photochemical products of N $O_{\gamma}$) yielded ( $O_3$) ＝25.8 〔 N $O_{\gamma}$-N $O_{x}$ 〕/(N $O_{\gamma}$) ＋27, ( $r^{2}$＝0.58). The regression intercept is interpreted as the ozone back ground (intercept=27ppbv) and the slope suggests that 8.6 molecules of ozone are formed per molecule of N $O_{x}$ oxidized products (when the average N $O_{\gamma}$ concentration, about 3 ppbv at the site, is used). The N $O_{x}$ N $O_{\gamma}$ ratio was used as an indicator of the chemical age of airmasses and the ratio showed strong positive correlations with HN $O_3$( $r^{2}$=0.58), PAN ( $r^{2}$=0.46) and $O_3$( $r^{2}$=0.62). Larger N $O_{\gamma}$ and N $O_{x}$ N $O_{\gamma}$ ratio were found when winds came from continental sides. It may suggest that synoptic meteorological conditions and transport of N $O_{x}$ are important in the distribution of N $O_{\gamma}$ and its relationship with photochemical oxidants at the site.

• The Journal of the Society of Korean Medicine Diagnostics
• /
• v.12 no.1
• /
• pp.42-62
• /
• 2008

Objectives: It is well known that some parameters of the photoplethysmogram (PPG) acquired by time domain contour analysis can be used as markers of vascular aging. But the previous studies that have been performed for frequency domain analysis of the PPG to date have provided only restrictive and fragmentary information. The aim of the present investigation was to determine whether the harmonics extracted from the PPG using a fast Fourier transformation could be used as an index of vascular aging. Methods: The PPG was measured in 600 recruited subjects for 30 second durations, To grasp the gross age-related change of the PPG waveform, we grouped subjects according to gender and age and averaged the PPG signal of one pulse cycle. To calculate the conventional indices of vascular aging, we selected the 5-6 cycles of pulse that the baseline was relatively stable and then acquired the coordinates of the inflection points. For the frequency domain analysis we performed a power spectral analysis on the PPG signals for 30 seconds using a fast Fourier transformation and dissociated the harmonic components from the PPG signals. Results: A final number of 390 subjects (174 males and 216 females) were included in the statistical analysis. The normalized power of the harmonics decreased with age and on a logarithmic scale reduction of the normalized power in the third (r=-0.492, P<0.0001), fourth (r=-0.621, P<0.0001) and fifth harmonic (r=-0.487, P<0.0001) was prominent. From a multiple linear regression analysis, Stiffness index, reflection index and corrected up-stroke time influenced the normalized power of the harmonics on a logarithmic scale. Conclusions: The normalized harmonic power decreased with age in healthy subjects and may be less error prone due to the essential attributes of frequency domain analysis. Therefore, we expect that the normalized harmonic power density can be useful as a vascular aging marker.

• Journal of Korean Society of Forest Science
• /
• v.83 no.2
• /
• pp.191-204
• /
• 1994

Due to its topographic complexities and various climatical condition, Korea exhibits diverse forest types. Dominant tree species in this zone are Quercus spp., Betula spp., Zelkova spp., Fraxinus spp., Pinus densiflora, Pinus koraiensis, and Pinus thunbergii ete. Genetic conservation in forest species in Korea there are three ways ; one is in situ, other is ex situ and third is in-facility conservation. In situ conservation include that are the present status of conservation of rare and endangered flora and ecosystem, the reserved forest, the national and provincial park, and the gene pool of natural forests. Ex situ conservation means to be established the new forest from in situ forest stands, progeny and provenance test populations, seed orchard and clone banks, and gene conservation in-facility. As a tool for low temperature storage, several aspects on in vitro system were studied ; (1) establishment of in vitro cultures from juvenile and/or rejuvenated tissues, (2) induction of multiple shoots from the individual micropropagules, (3) elongation of the proliferated shoots. Studies on cold storage for short-and long-term maintenance of in vitro cultures under $4^{\circ}C$ in the refrigerator were conducted. For the cryopreservation at $-196^{\circ}C$, various factors affecting survivability of the plant materials are being examined. The necessity of gene conservation of forest trees is enlarged not only to increase the adaptability for various environments but also to gain the breeding materials in the future. For effective gene conservation of forest trees, I would like to suggest followings ; 1. Forest stands reserved for other than the gene conservation purposes such as national parks should be investigated by botanical and gene-ecological studies for selecting bio-diversity and gene conservation stands. 2. Reserved forest for gene pool should be extented both economically important tree spp. and non-economical species. 3. Reserved forest for progeny test and clone bank should be systematically investigated for the use of Ex situ forest gene conservation. 4. We have to find out a new methodology of genetic analysis determining the proper and effective size of subpopulation for in situ gene conservation. 5. We should develop a new tree breeding systems for successful gene conservation and utilization of the genetic resources. 6. New method of in-facility gene conservation using advanced genetic engineering should be developed to save time and economic resources. 7. For the conservation of species with short-life span of seed or shortage of knowledge of seed physiology, tissue culture techniques will be played a great role for gene conservation of those species. 8. It is are very useful conservation not only of genes but of genotypes which were selected already by breeding program. 9. Institutional and administrative arrangements including legistlation must be necessarily taken for gene conservation of forest trees. 10. It is national problems for conservation of forest resources which have been rapidly destroyed because of degenerating environmental condition and of inexperienced management system of bio-diversity and gene conservation. 11. In order to international cooperation for exchanging data of bio-diversity and gene conservation, we should connect to international net works as soon as possible.