• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.19 no.6
• /
• pp.227-233
• /
• 2019

Recently, along with the recent development of deep learning technique, neural networks are achieving success in computer vision filed. Convolutional neural network have shown outstanding performance in not only for a simple image classification task, but also for tasks with high difficulty such as object segmentation and detection. However many such deep learning models are based on supervised-learning, which requires more annotation labels than image-level label. Especially image semantic segmentation model requires pixel-level annotations for training, which is very. To solve these problems, this paper proposes a weakly-supervised semantic segmentation method which requires only image level label to train network. Existing weakly-supervised learning methods have limitations in detecting only specific area of object. In this paper, on the other hand, we use multi-classifier deep learning architecture so that our model recognizes more different parts of objects. The proposed method is evaluated using VOC 2012 validation dataset.

• Korean Institute of Interior Design Journal
• /
• v.19 no.6
• /
• pp.75-82
• /
• 2010

The purposes of this study are to present the criteria for a sensibility rating scale for measuring the general public's perception of museums' spatial environment, particularly lobby space, through an empirical analysis; and to clarify the characteristics of the presented rating scale in terms of each rating element and factor. For this study, a survey was conducted during September 11-17, 2010, and a total of 370 museum visitors participated in the survey. A sensibility rating scale used for the survey consisted of a total of 32 adjectives selected from a literature review of previous studies. To specify the dimensions of semantic space using the semantic adjectives, words with opposite meanings were analyzed with the semantic differential technique developed by Osgood et al. Using SPSS, a reliability analysis, factor analysis, and cluster analysis were conducted on the data obtained from the survey. The results of this study can be summarized as follows: According to the general public's perception of museum lobbies, five factors were found from the 19 semantic ratings of the Gyeonggido Museum of Modern Art and the 20 semantic ratings of the Seoul Museum of Art, respectively. In the case of Gyeonggido Museum of Modern Art, three additional semantic words of 'orderly', 'open', and 'original', which did not appear in the case of Seoul Museum of Art, were discovered. In the case of Seoul Museum of Art, more detailed semantic words such as 'restrained', 'ordinary', 'concrete', and 'intellectual (rational)' were obtained. Five semantic elements, which describe the two museums, were: Feelings of 'pleasantness', 'value, 'usage', 'aesthetics', and 'materials'. According to a comparative analysis of the two lobby spaces in terms of semantic rating elements, Gyeonggido Museum of Modern Art was perceived to be an orderly, original, open, soft, and female-like space, whereas Seoul Museum of Art was perceived to be aesthetic, restrained, concrete, realistic, intellectual and rational. In the coming years, the results of this study will serve as valuable data for constructing a sensibility rating scale for evaluating spatial environments of museums.

• The KIPS Transactions:PartD
• /
• v.13D no.2 s.105
• /
• pp.223-234
• /
• 2006

Branching and merging have been being the outstanding methods for SCM in terms of supporting parallel developments. Since well-known commercial merging tools based on textual merging have not detecting semantics conflicts, they can cause semantic errors in the result of merging. Although a lot of researches for detecting semantic conflict and merging up to recently, these researches have been doing individually. Therefore, it is necessary for a research detecting semantic conflict on textual merging and solving it. In this paper, we propose a new method for merging which preserve semantics on textual merging. The method merging two revisions from a source program is as follows : 1) defining changing operations, which include Update, Delete, and Insert operation, per line on two revisions corresponding to the line in source program, 2) detecting textual conflicts and semantic conflict in terms of executional behaviors, 3) solving these conflicts before merging. So, the proposed method can be regarded as a hybrid method that combines a method of textual merging and a behavioral semantic merging.

• Journal of KIISE:Databases
• /
• v.31 no.2
• /
• pp.195-204
• /
• 2004

The Semantic Web is proposed as the next generation Web technology. In the environment of the Semantic Web, resources are related with each other semantically and computers can process this information easily. The Resource Description Framework (RDF) supports this semantic relationship. RDF is the data model for describing metadata of the Web resources. To establish and develop the Semantic Web, methods for managing RDF data efficiently are the most important. So, in this research, we propose methods for storing and querying RDF data using an XML database system. Using an XML database system, XML data, main data of the Semantic Web, and RDF data, the metadata of XML data, can be managed in the same storage and by the same mechanism efficiently. In addition, we propose an efficient data retrieval method and several techniques to improve the system performance. Our query processing technique performs better than an existing system.

• The Journal of the Korea Contents Association
• /
• v.10 no.6
• /
• pp.18-26
• /
• 2010

Many personalized services that provide users with adaptive information according to users' preferences have been researched and developed. Push services are especially expected to be more economic impact because push services satisfy user's potential needs even if the user does not require anything. In this paper, we propose Semantic Web approach in order to enhance the performance of push services. Our approach provides infrastructure to recommend contents based on semantic association by enabling information of contents and user preferences to be described on service-specific ontologies that reflect features of each service. In addition, our approach can recommend users with adaptive information based on information represented in our description model. Our approach enables information of contents and user preferences to be described with rich expressiveness, and it provides semantic interoperability.

• Journal of the Korea Society for Simulation
• /
• v.17 no.4
• /
• pp.183-189
• /
• 2008

The next-generation war game simulators need a technique that reuses resources disperse on the web, and reorganizes federates on the fly based on the various events in real time. So far, HLA-based federates limit their interoperability to military networks, and in syntax-level. Web services techniques are widely used in enterprise applications and provide many proven practices to extend interoperability between WAN resources in semantic level. Two problems are met in order to utilize web services into war-game simulator : 1) How to achieve semantic-level interoperability between federates disperse on WAN, 2) How to interoperate web-based federates and RTI-based federates. In this paper, we provide solutions to the problems and highlight advantages using web-based federates with an example of ASuW(Anti-Surface Warfare).

• Journal of Internet Computing and Services
• /
• v.20 no.3
• /
• pp.33-41
• /
• 2019

In this paper, we propose a method to learn the redesigned class with background and object for semantic segmentation of indoor scene image. Semantic image segmentation is a technique that divides meaningful parts of an image, such as walls and beds, into pixels. Previous work of semantic image segmentation has proposed methods of learning various object classes of images through neural networks, and it has been pointed out that there is insufficient accuracy compared to long learning time. However, in the problem of separating objects and backgrounds, there is no need to learn various object classes. So we concentrate on separating objects and backgrounds, and propose method to learn after class simplification. The accuracy of the proposed learning method is about 5 ~ 12% higher than the existing methods. In addition, the learning time is reduced by about 14 ~ 60 minutes when the class is configured differently In the same environment, and it shows that it is possible to efficiently learn about the problem of separating the object and the background.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.6
• /
• pp.2049-2068
• /
• 2021

AI combined with NLP techniques has promoted the use of Virtual Assistants and have made people rely on them for many diverse uses. Conversational Agents are the most promising technique that assists computer users through their operation. An important challenge in developing Conversational Agents globally is transferring the groundbreaking expertise obtained in English to other languages. AI is making it possible to transfer this learning. There is a dire need to develop systems that understand secular languages. One such difficult language is Hindi, which is the fourth most spoken language in the world. Semantic similarity is an important part of Natural Language Processing, which involves applications such as ontology learning and information extraction, for developing conversational agents. Most of the research is concentrated on English and other European languages. This paper presents a Corpus-based word semantic similarity measure for Hindi. An experiment involving the translation of the English benchmark dataset to Hindi is performed, investigating the incorporation of the corpus, with human and machine similarity ratings. A significant correlation to the human intuition and the algorithm ratings has been calculated for analyzing the accuracy of the proposed similarity measures. The method can be adapted in various applications of word semantic similarity or module for any other language.

• The Journal of the Korea Contents Association
• /
• v.21 no.3
• /
• pp.23-32
• /
• 2021

Semantic image segmentation technology in the field of computer vision is a technology that classifies an image by dividing it into pixels. This technique is also rapidly improving performance using a machine learning method, and a high possibility of utilizing information in units of pixels is drawing attention. However, this technology has been raised from the early days until recently for 'lack of detailed segmentation' problem. Since this problem was caused by increasing the size of the label map, it was expected that the label map could be improved by using the edge map of the original image with detailed edge information. Therefore, in this paper, we propose a post-processing algorithm that maintains semantic image segmentation based on learning, but modifies the resulting label map based on the edge map of the original image. After applying the algorithm to the existing method, when comparing similar applications before and after, approximately 1.74% pixels and 1.35% IoU (Intersection of Union) were applied, and when analyzing the results, the precise targeting fine segmentation function was improved.

• KIPS Transactions on Computer and Communication Systems
• /
• v.6 no.6
• /
• pp.281-288
• /
• 2017

In Smart-cities, various types of integrated services must be linked to provide services to applications. Therefore, flexibility must be ensured between services so that various services can be efficiently provided. In order to secure the flexibility among services, it is very important to have a function to dynamically discover and invoke a desired service by searching for a semantic service by reflecting a recognized context through real-time context-aware in smart-cities. To date, quite a number of semantic service discovery techniques have been developed. However, they have not been verified as suitable for use in the smart-city domain. In this study, we tried to verify the existing ones to use a suitable one. We tested most of existing semantic service discovery techniques, but we found that none of them is suitable to our research. Therefore, we developed our own semantic service discovery technique. This paper introduces our work and presents the performance evaluation results that demonstrate that our developed works well and show good performance. For the performance evaluation, the experimental system was actually constructed and the real performance was measured. In the experiment, we implemented the semantic service discovery scenario that dynamically searches and calls the services needed to provide fire accident management services in smart cities.