• Proceedings of the Korea Information Processing Society Conference
• /
• 2019.10a
• /
• pp.895-897
• /
• 2019

모바일의 기술 발전과 소셜미디어 사용의 증가로 수없이 많은 멀티미디어 콘텐츠들이 생성되고 있다. 이러한 많은 양의 콘텐츠 중에서 사용자가 원하는 이미지를 효율적으로 찾기 위해 의미 기반 이미지 검색을 이용한다. 이 검색 기법은 이미지에 의미 있는 정보들을 이용하여 사용자가 찾고 자하는 이미지를 정확하게 찾을 수 있다. 본 연구에서는 모바일 환경에서 이미지가 가질 수 있는 의미적 정보를 어노테이션 하고 이와 더불어 모바일에 있는 이미지에 풍성한 어노테이션을 위해 딥러닝 기술을 이용하여 다양한 태그들을 자동 생성하도록 구현하였다. 이렇게 생성된 어노테이션 정보들은 의미적 기반 태그를 통해 RDF 트리플로 확장된다. SPARQL 질의어를 이용하여 의미 기반 이미지 검색을 할 수 있다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2009.04a
• /
• pp.504-507
• /
• 2009

사용자 참여, 개방, 공유를 기반으로 하는 웹2.0 환경 하에서 보다 효율적이고 합법적으로 UCC(User Created Contents)가 제작, 공유, 유통되기 위한 저작권 시스템으로서, CCL(Creative Commons License)이 빠르게 확산되고 있다. 그러나, UCC에 CCL메타데이터를 직접 첨부하지 않는 기존의 CCL어노테이션 방식은 의도적으로 또는 비의도적으로 UCC와 CCL메타데이터가 분리될 수 있는 문제점을 가지고 있다. 그 결과, UCC로부터 분리되어진 저작권관련정보는 변경 또는 분실되어 추후에 저작권분쟁의 원인이 될 수 있다. 본 연구에서는 RDF와 XMP를 기반으로 CCL메타데이터를 이미지 컨텐츠에 직접 부착하는 방법을 제안하고 이를 지원하기 위한 어노테이션 지원도구(CCL-MASTer)를 개발하였다. 본 연구결과는, 기존의 어노테이션 기법의 문제점을 해결하고 다양하고 풍부한 CCL메타데이터의 효율적인 어노테이션을 가능케 하여 UCC 저작권분쟁 해결방안 및 UCC와 UCC관련 CCL메타데이터의 생성과 검색 등의 토대가 된다.

• Journal of KIISE:Software and Applications
• /
• v.37 no.7
• /
• pp.548-557
• /
• 2010

In this paper, we propose a semi-automatic modeling approach of ontology to annotate VOD to realize the IPTV's intelligent searching. The ontology is made by combining partial tree that extracts hypernym, hyponym, and synonym of keywords related to a service domain from WordNet. Further, we add to the partial tree new keywords that are undefined in WordNet, such as foreign words and words written in Chinese characters. The ontology consists of two parts: generic hierarchy and specific hierarchy. The former is the semantic model of vocabularies such as keywords and contents of keywords. They are defined as classes including property restrictions in the ontology. The latter is generated using the reasoning technique by inferring contents of keywords based on the generic hierarchy. An annotation generates metadata (i.e., contents and genre) of VOD based on the specific hierarchy. The generic hierarchy can be applied to other domains, and the specific hierarchy helps modeling the ontology to fit the service domain. This approach is proved as good to generate metadata independent of any specific domain. As a result, the proposed method produced around 82% precision with 2,400 VOD annotation test data.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2013.05a
• /
• pp.583-585
• /
• 2013

The Universal Plug and Play (UPnP) specification enables sharing of diverse multimedia contents among different devices. Contents discovered via UPnP are organized using directory hierarchy. If there are multiple sources of content providers and if the directory becomes large, managing and searching of the shared contents can be difficult. In this paper, an enhanced contents searching method is proposed by using semantic annotations. UPnP metadata is expanded to include semantic annotations to enable natural contents searching.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2016.01a
• /
• pp.33-35
• /
• 2016

본 논문에서는 모바일 디바이스를 이용하여 촬영한 이미지 또는 이미 모바일 디바이스에 저장된 이미지를 사용자가 검색을 위한 질의어로 사용할 수 있는 방법에 대하여 제안한다. 기존의 모바일 검색엔진을 그대로 활용하기 위해 이미지 어노테이션에 기반한 태깅 키워드를 검색 이미지와 매칭하여 질의하는 방식으로 구현하며, 이 과정에서 이미지의 분석과 분류를 위한 SVM(Support Vector Machine)과 SIFT(Scale Invariant Feature Transform) 알고리즘을 사용하였으며, 이미지 어노테이션 태깅에 대한 키워드 매칭을 위해 빅데이터에서의 MapReduce를 응용하였다.

• Journal of Korea Multimedia Society
• /
• v.10 no.3
• /
• pp.381-390
• /
• 2007

3D object contents are used for various applications in the Web virtual space, where the main concerns are to navigate the 3D virtual space and visualize 3D objects. The techniques to manipulate 3D objects like disassembling and assembling and to record the manipulation process are the very first step. Until now, we can record only the result of 3D object manipulation. Thus, we have tried to study the representation technique to record meaningfully and replay the manipulation process of 3D structured objects. We analyzed the structures and their relations between components to construct 3D objects that are described in XML or VRML. Compared to the previous method, we studied a XML based annotation technique to record and store selectively by user. This technique makes 3D structured objects be used in the various applications by the selective recording and also selective replaying.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.10a
• /
• pp.50-53
• /
• 2022

In order to detect an object in yolv5, a process of annotating location information on an existing image is required when learning an image. The most representative method is to draw a bounding box on an image to store location information as meta information. However, if the boundary of the object is ambiguous, it will be difficult to make a bounding box. A representative example would be to classify parts that are not fire and parts that are fire. Therefore, in this paper, images of 100 samples judged to have caught fire were learned by varying the number of boxes. The results showed better fire detection performance in the model where the bounding box was trained by annotating it with three boxes by segmenting it slightly more than annotating it with one box by holding the edge as large as possible during annotating it with one box.

• Journal of KIISE:Software and Applications
• /
• v.36 no.8
• /
• pp.673-682
• /
• 2009

Recently, photo sharing and publishing based Social Network Sites(SNSs) are increasingly attracting the attention of academic and industry researches. Unlike the face recognition environment addressed by existing works, face annotation problem under SNSs is differentiated in terms of daily updated images database, a limited number of training set and millions of users. Thus, conventional approach may not deal with these problems. In this paper, we proposed a face annotation method for sharing and publishing photographs that contain faces under a social network service using random projection, non-linear regression and representational state transfer. Our experiments on several databases show that the proposed method records an almost constant execution time with comparable accuracy of the PCA-SVM classifier.

• Journal of Korea Multimedia Society
• /
• v.12 no.10
• /
• pp.1486-1498
• /
• 2009

Modem efforts on digitizing electronic books focus on preserving authentic image representation of the original sources. Unlike the text-based format, it is difficult to recognize the information in the image, so the new format requires new tools to help users to access, process, and make sense of digital information. This paper presents an approach which assists users of these image sources by giving them a combination of annotation and social navigation support. Especially in the education domain, the proposed technique improves the usability of online education system. This approach is currently fully implemented and under evaluation in a classroom study.

• Journal of Korea Multimedia Society
• /
• v.8 no.2
• /
• pp.137-153
• /
• 2005

This paper presents a transcoding method that dynamically adapts Web pages to various devices. The proposed method is based on a CC/PP profile that is a standard description of a device's context information. Additionally, to support a sophisticated transcoding, we define an annotation schema for representing additional information about original contents. Since a mobile device has a screen of limited size, A Web page is splitted into many small ones. Our method constructs a navigation-map that represents the hierarchical relations among the splitted pages. Experimental results with various Web contents show that the proposed method is superior in terms of user's convenience of navigation and the transcoding Quality.