Hyunsook Chung

Hybrid genetic algorithm approach for precedence-constrained sequencing problem

The objective of precedence-constrained sequencing problem (PCSP) is to locate the optimal sequence with the shortest traveling time among all feasible sequences. Various methods for effectively solving the PCSP have been suggested. This paper proposes a new concept of hybrid genetic algorithm (HGA) with adaptive local search scheme in order that the PCSP should be effectively solved. By the use of the adaptive local search scheme, the local search is automatically adapted into the loop of genetic algorithm. Two types of the PCSP are presented and analyzed to compare the efficiency among the proposed HGA approach and other competing conventional approaches. Finally, it is proved that the proposed HGA approach outperforms the other competing conventional approaches.

Conditional random field-based gesture recognition with depth information

Abstract. Gesture recognition is useful for human-computer interaction. The difficulty of gesture recognition is that instances of gestures vary both in motion and shape in three-dimensional (3-D) space. We use depth information generated using Microsoft’s Kinect in order to detect 3-D human body components and apply a threshold model with a conditional random field in order to recognize meaningful gestures using continuous motion information. Body gesture recognition is achieved through a framework consisting of two steps. First, a human subject is described by a set of features, encoding the angular relationship between body components in 3-D space. Second, a feature vector is recognized using a threshold model with a conditional random field. In order to show the performance of the proposed method, we use a public data set, the Microsoft Research Cambridge-12 Kinect gesture database. The experimental results demonstrate that the proposed method can efficiently and effectively recognize body gestures automatically.

Ontology construction using online ontologies based on selection, mapping and merging

The reuse of ontologies existing on the Web should be required to speed up ontology construction process. However, current ontology mapping approaches assume that two input ontologies are given not discovered from available semantic data of the web. In this paper, we present a new approach, which integrates ontology selection, mapping, and merging processes in order to minimise human mediation. Our ontology selection mechanism accepts a hand-crafted ontology as query ontology to search already constructed relevant ontologies. In addition, we develop a ranking method based on syntactic and semantic structure of classes to provide the best search result to the user.

Shape from Focus through Laplacian Using 3D Window

One of the fundamental objectives of computer vision is to reconstruct a three-dimensional (3D) structure of objects from two-dimensional (2D) images. The basic idea of image focus is that objects at different distances from a lens are focused at different distances. Shape from Focus (SFF) is the problem of reconstructing the depth of the scene changing actively the optics of the camera until the point of interest is in focus. The point in focus gives information about its depth through the thin lens Gaussian law. An effective focus measure operator should be a high-pass filter. Usually, the variation of frequency components are not enough that focus measure could be computed pixel-wise, therefore, sum of pixels in small 2D windows are used for detecting the high frequency components. In this paper, we propose to use 3D windows instead of 2D windows for detecting the high frequency components in the images. The proposed algorithm using 3D window gives better depth map than the previous algorithms using 2D windows.

Test-retest reliability of the driving habits questionnaire in older self-driving adults.

[Purpose] The purpose of this study was to investigate the test-retest reliability of the Driving Habits Questionnaire in community-dwelling older self-drivers. [Subjects and Methods] Seventy-four participants were recruited by convenience sampling from local rehabilitation centers. This was a cross-sectional study design that used two clinical measures: the Driving Habits Questionnaire and Mini-mental State Examination. To examine the test-retest reliability of the Driving Habits Questionnaire, the clinical tool was measured twice, five days apart. [Results] The Driving Habits Questionnaire showed good reliability for older community-dwelling self-drivers. The Cronbach’s alpha coefficients for the four domains of dependence (0.572), difficulty (0.871), crashes and citations (0.689), and driving space (0.961) of the Driving Habits Questionnaire indicated good or high internal consistency. Driving difficulty correlated significantly with self-reported crashes and citations and driving space. [Conclusion] The results of this study suggest that the Driving Habits Questionnaire is a reliable measure of self-reported interview-based driving behavior in the community-dwelling elderly.

Conflict Detection and Resolution in Merging of Topic Maps

Merging between topic maps describes the process of integrating two topic maps into a new topic map. In this paper, we propose a specific ontology merging process for integrating two source topic maps into a new topic map based on given semantic correspondences. Also, we define a taxonomy of merging conflicts which occur during the process of topic maps merging and a method for detecting and resolving them. Our experiments show that the automatic merging module compared with manual merging by an expert has advantages in terms of time and effort.

Design of user-centric semantic rights model for validation of user-generated content

In these days, the creative user really enjoys creating digital items and sharing their works with other people on the Web. Most users, who create digital contents, want to make secured packages of their works and distribute them with the attachment of valid licenses. Current digital rights management (DRM) systems, however, do not provide the functionality that supports the requirement of the creative users who are considered as just consumers by the current available DRM systems. To make the user-centric DRM functionality possible, we found that license management should be more intelligent to enable users make appropriate licenses for the secured distribution of their created works. In this paper, we define the semantic-based rights expression and management model for the user generated content. Each user created content can have one or more licenses with different types, reproduction, distribution, and usage. Based on our semantic license model and big data analytics, we can support a new business model in which users can sell and buy their created digital items in a secure environment.

An intuitive user interaction method using multi-sensors for pencil drawing filter of NPR rendering in mobile devices

The user interface is one of the most important factors to use various functions of mobile device easily. Recently, the importance of the user interface has been increased after Apple emphasized emotional UX(User eXperience) by applying multi-touches to iPhone. Mobile device Manufacturers have tried to apply embedded sensors to apps as user interface. Microphone was applied to execute functions combined with voice recognition and GPS sensor was used to find a location in navigation apps or game apps. Like these services, several sensors have roles of interface between user and device. In this paper, we propose an intuitive user interaction method using multi-sensors in NPR(Non-Photorealistic) rendering. The proposed method renders pencil drawing filter for a given photo image with the direction and strength of line strokes which are applied in vector fields. At this point, the method changes direction of stroke with x-coordinate values and y-coordinate values and control the strength of the stroke with z-coordinate values after catching three coordinate values in real-time from the orientation sensor on a mobile device. The method also adjusts brightness of image by transforming intensity of light from the light sensor. The light sensor updates intensity of light source continuously. If the user brings the mobile device closer to the external light source, the light sensor value will increase and the intensity of brightness becomes stronger. In addition, the user can make an attenuation effect with orientation sensor combined with light sensor for specific scope in a photo image and change the attenuation scope by moving action of his holding mobile device. This is a user interaction method for emotional image processing using sensors in mobile devices. We developed an app which generated pencil drawing filter effect for a photo image. We experimented sensor based emotional user interaction and compared various results of image processing effects using two combinatorial sensors. Finally, this paper proved efficiency through the experimental results and evaluated the effectiveness of the proposed method.

Semantic Model of Syllabus and Learning Ontology for Intelligent Learning System

The syllabus is a blueprint of course for teaching and learning because it contains the important meaning of promise between instructor and students in higher education and university. However, the current most of all syllabus management systems provide simple functionalities including creation, modification, and retrieval of the unstructured syllabus. In this paper, our approach consists of a definition of the ontological structure of the syllabus and semantic relationships of syllabuses, classification and integration of the syllabus based on ACM/IEEE computing curriculum, and formalization of learning goals, learning activity, and learning evaluation in syllabus using Bloom’s taxonomy for improving the usability of the syllabus. Also, we propose an effective method for enhancing the learning effect of students through the construction of subject ontology, which is used in discussion, visual presentation, and knowledge sharing between instructor and students. We prove the retrieval and classification correctness of our proposed methods according to experiments and performance evaluations.

Ontology-Based TV Program Contents Retrieval and Recommendation

In this paper we propose a searching and recommendation method of TV program contents in order to reduce the information overload problem. Most of previous recommendation approaches are dependent on simple ratings of users to determine preference similarity among the users. But our approach is closely related to the combination of ontology-based TV program searching and content-based filtering based on TV ontology and usage history. We search TV programs by computing the similarity between contents ontologies, filter the candidates with preferences of users, and return the ranked list of TV programs. A subjective experiments show that our proposed method is effective in semantic-based searching and recommendation.