• Journal of Korea Multimedia Society
• /
• v.18 no.3
• /
• pp.359-367
• /
• 2015

In this paper, we propose intelligent fire learning and detection system using hybrid visual attention mechanism of human. Proposed fire learning system generates leaned data by learning process of fire and smoke images. The features used as learning feature are selected among many features which are extracted based on bottom-up visual attention mechanism of human, and these features are modified as learned data by calculating average and standard variation of them. Proposed fire detection system uses learned data which is generated in fire learning system and features of input image to detect fire.

• Proceedings of the KSR Conference
• /
• 2008.06a
• /
• pp.975-979
• /
• 2008

Since tram has the advantages to reduce construction cost of infrastructure, to improve accessibility of passengers, and to offer visual pleasures, nowadays, it is one of light rails attracting public attention. Tram can be classified into two groups, one is a conventional steel-wheeled type, and the other is a rubber-tired type (bi-modal tram). The bi-modal tram propelled by the serial CNG hybrid propulsion unit has been developing since 2003 in Korea, which can realize both scheduled operation of railway and route flexibility of bus. Because the bimodal will be operated on both railway mode and bus mode, however, specific criteria and regulations for its design, certification, construction, operation and maintenance have not been determined definitely yet. In consideration of mobility enhancement for the old and the handicapped, motor vehicle safety standard and urban transit (railway vehicle) safety standard, several design specifications were proposed for car body and interiors of the bimodal tram vehicle. The design specifications proposed in this paper can be expected to promote passengers' comfort and safety, operation efficiency of the bimodal tram.

• Journal of the Korean Society of Costume
• /
• v.58 no.6
• /
• pp.110-123
• /
• 2008

The objectives of this study are to define the fundamental notions of collaboration in design field. The results of this study are summarized as follows: Firstly, collaboration means a cooperative work which are cooperatively done by collabarator and collaboratee on the equal status within a certain period of time, based on core ability of each of them. This cooperative work is presented with a visual image based on the identity of participants, and it creates an added value. Collaboration is different from convergence and hybrid in the methodological aspects and from brand alliance and strategic alliance in the aspects of the subject of project. Secondly, collaboration has been developing through step of using the specific ability, step of sharing each core abilities, and step of fusing their identities, in a view of the sphere of cooperative work and the intimate relations in reciprocity. Thirdly, according to strategic purpose, collaboration can be classified into value ascending collaboration, image changing collaboration, business-field extension collaboration, and event collaboration. Fourthly, production methods of collaboration consist of limited edition method, premium line method, common line method, producing a collaborative product, and so on. Fifthly, differentiation effect, image upgrade effect, production of higher value-added products, sale increase, diversification of business, sharing the targets, and promotion effect are achievable through collaboration. Sixthly, to make progress in collaboration successfully, the compatibility of combination between collaborator and collaboratee, the superiority of collaboratee, definitude of collaboration, the fit method of collaboration, the novelty of collatoration, possibility of issuing of collaboration, and perfection of collabo-product must be the points which deserve our attention.

• Fisheries and Aquatic Sciences
• /
• v.19 no.5
• /
• pp.26.1-26.8
• /
• 2016

Background: The nematode species belonging to genus Anisakis occur at their third larval stage in numerous marine teleost fish species worldwide and known to cause accidental human infection through the ingestion of raw or undercooked fish or squids. They may also draw the attention of consumers because of the visual impact of both alive and dead worms. Therefore, the information on their geographical distribution and clear species identification is important for epidemiological survey and further prevention of human infection. Results: For identification of anisakid nematodes species isolated from largehead hairtail (Trichiurus japonicus), polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) analysis of internal transcribed spacers of ribosomal DNA were conducted. Mitochondrial cytochrome c oxidase subunit 2 gene was also sequenced, and phylogenetic analysis was conducted. From the largehead hairtail (n = 9), 1259 nematodes were isolated in total. Most of the nematodes were found encapsulated throughout the viscera (56.2 %, 708/1259) or moving freely in the body cavity (41.5 %, 523/1259), and only 0.3 % (4/1259) was found in the muscles. By PCR-RFLP, three different nematode species were identified. Anisakis pegreffii was the most dominantly found (98.7 %, 1243/1259) from the largehead hairtail, occupying 98.7 % (699/708) of the nematodes in the mesenteries and 98.1 % (513/523) in the body cavity. Hybrid genotype (Anisakis simplex ${\times}$ A. pegreffii) occupied 0.5 %, and Hysterothylacium sp. occupied 0.2 % of the nematodes isolated in this study. Conclusions: The largehead hairtail may not significantly contribute accidental human infection of anisakid nematode third stage larvae because most of the nematodes were found from the viscera or body cavity, which are not consumed raw. But, a high prevalence of anisakid nematode larvae in the largehead hairtail is still in concern because they may raise food safety problems to consumers. Immediate evisceration or freezing of fish after catch will be necessary before consumption.

• Phonetics and Speech Sciences
• /
• v.11 no.1
• /
• pp.41-49
• /
• 2019

An end-to-end speech recognition model consisting of a single integrated neural network model was recently proposed. The end-to-end model does not need several training steps, and its structure is easy to understand. However, it is difficult to understand how the model recognizes speech internally. In this paper, we visualized and analyzed the attention-based end-to-end model to elucidate its internal mechanisms. We compared the acoustic model of the BLSTM-HMM hybrid model with the encoder of the end-to-end model, and visualized them using t-SNE to examine the difference between neural network layers. As a result, we were able to delineate the difference between the acoustic model and the end-to-end model encoder. Additionally, we analyzed the decoder of the end-to-end model from a language model perspective. Finally, we found that improving end-to-end model decoder is necessary to yield higher performance.

• Journal of Intelligence and Information Systems
• /
• v.23 no.2
• /
• pp.19-38
• /
• 2017

Collaborative filtering (CF) algorithm has been popularly used for implementing recommender systems. Until now, there have been many prior studies to improve the accuracy of CF. Among them, some recent studies adopt 'hybrid recommendation approach', which enhances the performance of conventional CF by using additional information. In this research, we propose a new hybrid recommender system which fuses CF and the results from the social network analysis on trust and distrust relationship networks among users to enhance prediction accuracy. The proposed algorithm of our study is based on memory-based CF. But, when calculating the similarity between users in CF, our proposed algorithm considers not only the correlation of the users' numeric rating patterns, but also the users' in-degree centrality values derived from trust and distrust relationship networks. In specific, it is designed to amplify the similarity between a target user and his or her neighbor when the neighbor has higher in-degree centrality in the trust relationship network. Also, it attenuates the similarity between a target user and his or her neighbor when the neighbor has higher in-degree centrality in the distrust relationship network. Our proposed algorithm considers four (4) types of user relationships - direct trust, indirect trust, direct distrust, and indirect distrust - in total. And, it uses four adjusting coefficients, which adjusts the level of amplification / attenuation for in-degree centrality values derived from direct / indirect trust and distrust relationship networks. To determine optimal adjusting coefficients, genetic algorithms (GA) has been adopted. Under this background, we named our proposed algorithm as SNACF-GA (Social Network Analysis - based CF using GA). To validate the performance of the SNACF-GA, we used a real-world data set which is called 'Extended Epinions dataset' provided by 'trustlet.org'. It is the data set contains user responses (rating scores and reviews) after purchasing specific items (e.g. car, movie, music, book) as well as trust / distrust relationship information indicating whom to trust or distrust between users. The experimental system was basically developed using Microsoft Visual Basic for Applications (VBA), but we also used UCINET 6 for calculating the in-degree centrality of trust / distrust relationship networks. In addition, we used Palisade Software's Evolver, which is a commercial software implements genetic algorithm. To examine the effectiveness of our proposed system more precisely, we adopted two comparison models. The first comparison model is conventional CF. It only uses users' explicit numeric ratings when calculating the similarities between users. That is, it does not consider trust / distrust relationship between users at all. The second comparison model is SNACF (Social Network Analysis - based CF). SNACF differs from the proposed algorithm SNACF-GA in that it considers only direct trust / distrust relationships. It also does not use GA optimization. The performances of the proposed algorithm and comparison models were evaluated by using average MAE (mean absolute error). Experimental result showed that the optimal adjusting coefficients for direct trust, indirect trust, direct distrust, indirect distrust were 0, 1.4287, 1.5, 0.4615 each. This implies that distrust relationships between users are more important than trust ones in recommender systems. From the perspective of recommendation accuracy, SNACF-GA (Avg. MAE = 0.111943), the proposed algorithm which reflects both direct and indirect trust / distrust relationships information, was found to greatly outperform a conventional CF (Avg. MAE = 0.112638). Also, the algorithm showed better recommendation accuracy than the SNACF (Avg. MAE = 0.112209). To confirm whether these differences are statistically significant or not, we applied paired samples t-test. The results from the paired samples t-test presented that the difference between SNACF-GA and conventional CF was statistical significant at the 1% significance level, and the difference between SNACF-GA and SNACF was statistical significant at the 5%. Our study found that the trust/distrust relationship can be important information for improving performance of recommendation algorithms. Especially, distrust relationship information was found to have a greater impact on the performance improvement of CF. This implies that we need to have more attention on distrust (negative) relationships rather than trust (positive) ones when tracking and managing social relationships between users.

• Cartoon and Animation Studies
• /
• s.18
• /
• pp.37-56
• /
• 2010

This paper focalizes on the 'evolution' of monster animation and . Actually, Many characters' 'Evolution' is a very powerful convention of Japanese genre animation. Lots of young people feel a huge pleasure about character's evolution. It is a fantastic spectacle to boys and girls. The first characteristic point of characters' evolution is growing-up body to an adult. For example, magical girls genre shows a woman's sexy body. This scene gives a sexual fantasy to young girl spectators, a voyeuristic visual pleasure to young boy spectators. The second point is that a transforming-body gains some super power or magical power. The third point is that a great many characters is in one text. In , characters' numbers go on increasing through 6 ways-hybrid type, mutant-reiteration type, metamorphosis type, simple-grownup type, narrative type, multiple evolution type. The evolution of has to pay attention to only one transforming way. Most Digimon characters are becoming to a human-weapons cyborg. In my opinion, this is very dangerous phenomenon with a political ideology.

• Journal of Sericultural and Entomological Science
• /
• v.13 no.2
• /
• pp.123-133
• /
• 1971

The author is the first man who tried to use an anesthetic on insect specially for silkworm in orde to evaluate the silkworm health and silk yielding ability and the obtained results are as followings. 1. The necessary ether vapor induction for narcosis on silkworms is varied by the glowth of the silkworm which the larger worm is, the longer induction is required. For instance, it was 2∼3 minutes for the worms of third day fifth instar silkworm in case use of ether anesthetic. 2. The longer anesthetic induction for silkworms, the longer recovery needs from anesthesia. In case five minutes ether vapor induction, silkworms recovered narcosis during in 5∼130 minutes which had varied very much by the health variation. 3. The ether induction caused silkworm to vomit digestive juice from a few per cent of the worms, but the chloroform induction showed majority of the worms to vomit the digestive juice out of mouth. So, the ether was found as better anesthetic for silkworms. 4. When ether induction last more than 20∼30 minutes, the recovered silkworms can eat mulberry, but when it gets more than three hours they can not eat mulberry. And when it last more than ten minutes, the silkworm may eat mulberry leaf, but they can not spinn cocoon properly. 5. In case five minutes ether induction for silkworms on third day fifth instar, the stronger variety is, the rallier recovered from narcosis. 6. The recovering duration from narcosis varies regarding with each worm health which shows Poisson′s distribution even in a same variety silkworm. 7. The female worms recover from narcosis earlier than male worms which means the female worm is stronger than male one. 8. The later recovered silkworm from narcosis spinned more rich cocoon silk and ended with smaller pupae weight. Such a tendency showed until at some recovery duration, then the silk yield droped down on the worms recovered in more longer duration. The author (Choe) had named such a relation curve as "Silk Yield Curve against Silkworm Health." 9. The silk yield or cocoon layer ratio had varied from 13 to 27% even in a same worm varity cocoon which showed serious variation and call attention carefulness for the duplication work of a variety silkworm eggs. 10. Not always the rich silk yielding worm is the best worm during the silkworm selection and it should be considered with the silkworm health evaluation. 11. At present situation, only specific breeding expert is allowed to join in the selection service because of need many years experience by use of visual observation, but the ether anesthesia method may help such an evaluation with more accuracy and easy way even for the people in fresh on the field. 12. The effect of the narcosis on the silkworm for the next generation or hybrid worm will be reported in next publication.