Go to the main menu
Skip to content
Go to bottom
REFERENCE LINKING PLATFORM OF KOREA S&T JOURNALS
> Journal Vol & Issue
The KIPS Transactions:PartB
Journal Basic Information
Journal DOI :
Korea Information Processing Society
Editor in Chief :
Volume & Issues
Volume 10B, Issue 7 - Dec 2003
Volume 10B, Issue 6 - Oct 2003
Volume 10B, Issue 5 - Aug 2003
Volume 10B, Issue 4 - Aug 2003
Volume 10B, Issue 3 - Jun 2003
Volume 10B, Issue 2 - Apr 2003
Volume 10B, Issue 1 - Feb 2003
Selecting the target year
A Study of using Emotional Features for Information Retrieval Systems
Kim, Myung-Gwan ; Park, Young-Tack ;
The KIPS Transactions:PartB, volume 10B, issue 6, 2003, Pages 579~586
DOI : 10.3745/KIPSTB.2003.10B.6.579
In this paper, we propose a novel approach to employ emotional features to document retrieval systems. Fine emotional features, such as HAPPY, SAD, ANGRY, FEAR, and DISGUST, have been used to represent Korean document. Users are allowed to use these features for retrieving their documents. Next, retrieved documents are learned by classification methods like cohesion factor, naive Bayesian, and, k-nearest neighbor approaches. In order to combine various approaches, voting method has been used. In addition, k-means clustering has been used for our experimentation. The performance of our approach proved to be better in accuracy than other methods, and be better in short texts rather than large documents.
Function Approximation for Reinforcement Learning using Fuzzy Clustering
Lee, Young-Ah ; Jung, Kyoung-Sook ; Chung, Tae-Choong ;
The KIPS Transactions:PartB, volume 10B, issue 6, 2003, Pages 587~592
DOI : 10.3745/KIPSTB.2003.10B.6.587
Many real world control problems have continuous states and actions. When the state space is continuous, the reinforcement learning problems involve very large state space and suffer from memory and time for learning all individual state-action values. These problems need function approximators that reason action about new state from previously experienced states. We introduce Fuzzy Q-Map that is a function approximators for 1 - step Q-learning and is based on fuzzy clustering. Fuzzy Q-Map groups similar states and chooses an action and refers Q value according to membership degree. The centroid and Q value of winner cluster is updated using membership degree and TD(Temporal Difference) error. We applied Fuzzy Q-Map to the mountain car problem and acquired accelerated learning speed.
Improvement of Accuracy of Decision Tree By Reprocessing
Lee, Gye-Sung ;
The KIPS Transactions:PartB, volume 10B, issue 6, 2003, Pages 593~598
DOI : 10.3745/KIPSTB.2003.10B.6.593
Machine learning organizes knowledge for efficient and accurate reuse. This paper is concerned with methods of concept learning from examples, which glean knowledge from a training set of preclassified ‘objects’. Ideally, training facilitates classification of novel, previously unseen objects. However, every learning system relies on processing and representation assumptions that may be detrimental under certain circumstances. We explore the biases of a well-known learning system, ID3, review improvements, and introduce some improvements of our own, each designed to yield accurate and pedagogically sound classification.
Active Documents: Programs by Form Designers
Nam, Chul-Ki ; Bae, Jae-Hak ; Yoo, Hae-Young ;
The KIPS Transactions:PartB, volume 10B, issue 6, 2003, Pages 599~610
DOI : 10.3745/KIPSTB.2003.10B.6.599
The Web plays an important role as information source and most Web applications are document-centric. A document implies an intention of its own designer, which can be utilized actively in automation of business processes. Through an understanding of an intrinsic nature of a document function, we can see a document as an executable computer program in a special case. For this approach, we propose an active document model that is composed of form, knowledge base, rules, and queries. For reusability and interoperability of a document, each component of the proposed model is uniformly represented in XML. The proposed active document not only plays a passive role in providing user interfaces, but also is a document that a machine can infer and process with reading a procedure of document processing and business rules intended by document designers. Through this approach, document can interact with machines and can cooperate with other applications. For applicability of our active document, we show a case study for the processing of purchase orders in a B2B e-Commerce system. This paper is expected to provide the framework of accelerating the development of intelligent applications through our approach regards form document as a computer program. In short, the proposed active document contains knowledge representation and processing method, consequently our document will play an important role in providing a concept of document of pursuing in Semantic Web.
Shot Boundary Detection of Video Data Based on Fuzzy Inference
Jang, Seok-Woo ;
The KIPS Transactions:PartB, volume 10B, issue 6, 2003, Pages 611~618
DOI : 10.3745/KIPSTB.2003.10B.6.611
In this paper, we describe a fuzzy inference approach for detecting and classifying shot transitions in video sequences. Our approach basically extends FAM (Fuzzy Associative Memory) to detect and classify shot transitions, including cuts, fades and dissolves. We consider a set of feature values that characterize differences between two consecutive frames as input fuzzy sets, and the types of shot transitions as output fuzzy sets. The inference system proposed in this paper is mainly composed of a learning phase and an inferring phase. In the learning phase, the system initializes its basic structure by determining fuzzy membership functions and constructs fuzzy rules. In the inferring phase, the system conducts actual inference using the constructed fuzzy rules. In order to verify the performance of the proposed shot transition detection method experiments have been carried out with a video database that includes news, movies, advertisements, documentaries and music videos.
Effective Watermark Detection Using Asymmetric Thresholds
Shin, Chang-Doon ; Oh, Hae-Seok ;
The KIPS Transactions:PartB, volume 10B, issue 6, 2003, Pages 619~628
DOI : 10.3745/KIPSTB.2003.10B.6.619
In this paper, an effective watermark detection technique in the wavelet transform domain is proposed. In this proposed method, the image is 2-level wavelet transformed, and then the watermark with a binary logo is embedded into middle band except baseband and high band to consider Invisibility and robustness. In this paper, we use an asymmetric thresholds watermarking (ATW) in which detection threshold is higher than inserting threshold in order to enhance watermark detection ratio in attacked images. In watermark detection phase, the detection value is not changed when the difference of the selected wavelet neighboring coefficient pairs Is smaller than specific value. The experimental results show that the proposed method has good quality and is robust to various attacks such as the JPEG lossy compression, noise addition, cropping, blurring, etc.
A adaptively robust method of DCT-based watermarking
Jun, Young-Min ; Kim, Gye-Young ; Choi, Hyung-Il ;
The KIPS Transactions:PartB, volume 10B, issue 6, 2003, Pages 629~638
DOI : 10.3745/KIPSTB.2003.10B.6.629
In this paper, we propose an improved method of watermarking to increase the invisibility of a watermark and robustness against attacks for the purpose of removing the embedded watermark. The proposed method adaptively selects DCT blocks and determines position and intensity for watermarking based on the characteristics of human visual system. The used features are texture, luminance and contrast. We show the experimental results against image processing attacks such as cropping, image enhancement, low pass filtering, and JPEG compression, and then discuss the performance of the proposed method.
A Visual Model for the Perception of the Optical illusions from Discrete Dot Stimuli
Jung, Eun-Hwa ; Hong, Keong-Ho ;
The KIPS Transactions:PartB, volume 10B, issue 6, 2003, Pages 639~646
DOI : 10.3745/KIPSTB.2003.10B.6.639
This paper proposes a neural network model for extracting optical illusions produced by a sequence of discontinuous dot stimuli. The proposed model is based on visual cell`s characters founded by visual information processing path. This study approaches on the basis of physiological observation of the perceptual phenomena that some simple ways of discrete dots are perceived as a continuous virtual contour rather than as separate dots. This paper presents the implementation of the optical illusions from discrete dot stimuli that are composed of virtual polygons from 6 to 10 dots. This experimental data are similar to those of Smith & Vos`s physiological experiments. The proposed model shows that it can extract continuous illusion contours from discrete dot stimuli successfully.
A Study on Vector-based Automatic Caricature Generation
Park, Yeon-Chool ; Oh, Hae-Seok ;
The KIPS Transactions:PartB, volume 10B, issue 6, 2003, Pages 647~656
DOI : 10.3745/KIPSTB.2003.10B.6.647
This paper proposes the system to generate caricature (character`s face) resembling human face using extracted facial features automatically. Since this system is vector-based, the generated character`s face has no size limit and constraint. So it is available to transform the shape freely and to apply various facial expressions to 2D face. Moreover, owing to the vector file`s advantage, it can be used in mobile environment as small file site.
Human Motion Tracking by Combining View-based and Model-based Methods for Monocular Video Sequences
Park, Ji-Hun ; Park, Sang-Ho ; Aggarwal, J.K. ;
The KIPS Transactions:PartB, volume 10B, issue 6, 2003, Pages 657~664
DOI : 10.3745/KIPSTB.2003.10B.6.657
Reliable tracking of moving humans is essential to motion estimation, video surveillance and human-computer interface. This paper presents a new approach to human motion tracking that combines appearance-based and model-based techniques. Monocular color video is processed at both pixel level and object level. At the pixel level, a Gaussian mixture model is used to train and classily individual pixel colors. At the object level, a 3D human body model projected on a 2D image plane is used to fit the image data. Our method does not use inverse kinematics due to the singularity problem. While many others use stochastic sampling for model-based motion tracking, our method is purely dependent on nonlinear programming. We convert the human motion tracking problem into a nonlinear programming problem. A cost function for parameter optimization is used to estimate the degree of the overlapping between the foreground input image silhouette and a projected 3D model body silhouette. The overlapping is computed using computational geometry by converting a set of pixels from the image domain to a polygon in the real projection plane domain. Our method is used to recognize various human motions. Motion tracking results from video sequences are very encouraging.
Speech Synthesis using Diphone Clustering and Improved Spectral Smoothing
Jang, Hyo-Jong ; Kim, Kwan-Jung ; Kim, Gye-Young ; Choi, Hyung-Il ;
The KIPS Transactions:PartB, volume 10B, issue 6, 2003, Pages 665~672
DOI : 10.3745/KIPSTB.2003.10B.6.665
This paper describes a speech synthesis technique by concatenating unit phoneme. At that time, a major problem is that discontinuity is happened from connection part between unit phonemes, especially from connection part between unit phonemes recorded by different persons. To solve the problem, this paper uses clustered diphone, and proposes a spectral smoothing technique, not only using formant trajectory and distribution characteristic of spectrum but also reflecting human`s acoustic characteristic. That is, the proposed technique performs unit phoneme clustering using distribution characteristic of spectrum at connection part between unit phonemes and decides a quantity and a scope for the smoothing by considering human`s acoustic characteristic at the connection part of unit phonemes, and then performs the spectral smoothing using weights calculated along a time axes at the border of two diphones. The proposed technique removes the discontinuity and minimizes the distortion which can be occurred by spectrum smoothing. For the purpose of the performance evaluation, we test on five hundred diphones which are extracted from twenty sentences recorded by five persons, and show the experimental results.
A Classified Space VQ Design for Text-Independent Speaker Recognition
Lim, Dong-Chul ; Lee, Hanig-Sei ;
The KIPS Transactions:PartB, volume 10B, issue 6, 2003, Pages 673~680
DOI : 10.3745/KIPSTB.2003.10B.6.673
In this paper, we study the enhancement of VQ (Vector Quantization) design for text independent speaker recognition. In a concrete way, we present a non-iterative method which makes a vector quantization codebook and this method performs non-iterative learning so that the computational complexity is epochally reduced The proposed Classified Space VQ (CSVQ) design method for text Independent speaker recognition is generalized from Semi-noniterative VQ design method for text dependent speaker recognition. CSVQ contrasts with the existing desiEn method which uses the iterative learninE algorithm for every traininE speaker. The characteristics of a CSVQ design is as follows. First, the proposed method performs the non-iterative learning by using a Classified Space Codebook. Second, a quantization region of each speaker is equivalent for the quantization region of a Classified Space Codebook. And the quantization point of each speaker is the optimal point for the statistical distribution of each speaker in a quantization region of a Classified Space Codebook. Third, Classified Space Codebook (CSC) is constructed through Sample Vector Formation Method (CSVQ1, 2) and Hyper-Lattice Formation Method (CSVQ 3). In the numerical experiment, we use the 12th met-cepstrum feature vectors of 10 speakers and compare it with the existing method, changing the codebook size from 16 to 128 for each Classified Space Codebook. The recognition rate of the proposed method is 100% for CSVQ1, 2. It is equal to the recognition rate of the existing method. Therefore the proposed CSVQ design method is, reducing computational complexity and maintaining the recognition rate, new alternative proposal and CSVQ with CSC can be applied to a general purpose recognition.
Implementation of Content-based Image Retrieval System using Color Spatial and Shape Information
Ban, Hong-Oh ; Kang, Mun-Ju ; Choi, Heyung-Jin ;
The KIPS Transactions:PartB, volume 10B, issue 6, 2003, Pages 681~686
DOI : 10.3745/KIPSTB.2003.10B.6.681
In recent years automatic image indexing and retrieval have been increasingly studied. However, content-based retrieval techniques for general images are still inadequate for many purposes. The novelty and originality of this thesis are the definition and use of a spatial information model as a contribution to the accuracy and efficiency of image search. In addition, the model is applied to represent color and shape image contents as a vector using the method of image features extraction, which was inspired by the previous work on the study of human visual perception. The indexing scheme using the color, shape and spatial model shows the potential of being applied with the well-developed algorithms of features extraction and image search, like ranking operations. To conclude, user can retrieved more similar images with high precision and fast speed using the proposed system.
Design of Indexing Agent for Semantic-based Video Retrieval
Lee, Jong-Hee ; Oh, Hae-Seok ;
The KIPS Transactions:PartB, volume 10B, issue 6, 2003, Pages 687~694
DOI : 10.3745/KIPSTB.2003.10B.6.687
According to the rapid increase of multimedia data quantity recently, various means of video data search has been desired. In order to process video data effectively, it is required that the content information of video data is loaded in database and semantic-based retrieval method can be available for various query of users. Currently existent contents-based video retrieval systems search by single method such as annotation-based or feature-based retrieval, and show low search efficiency and requires many efforts of system administrator or annotator form less perfect automatic processing. In this paper, we propose semantic-based video retrieval system which support semantic retrieval of various users by feature-based retrieval and annotation-based retrieval of massive video data. By user`s fundamental query and selection of image for key frame that extracted from query, the agent gives the detail shape for annotation of extracted key frame. Also, key frame selected by user become query image and searches the most similar key frame through feature based retrieval method that propose. Therefore, we design the system that can heighten retrieval efficiency of video data through semantic-based retrieval.
A transcode scheduling technique to reduce early-stage delay time in playing multimedia in mobile terminals
Hong, Maria ; Yoon, Joon-Sung ; Lim, Young-Hwan ;
The KIPS Transactions:PartB, volume 10B, issue 6, 2003, Pages 695~704
DOI : 10.3745/KIPSTB.2003.10B.6.695
This paper proposes a new scheduling technique to play multimedia data streams in mobile terminals. The paper explores the characteristics of multimedia data streams , firstly. On basis of these characteristics, selection of specific data stream can be possible as well as transcoding protest. Our approach aims at reducing the early-stage delay time more effectively since it makes possible to select and transcodes some specific streams by employing a selection policy rather than transcoding all streams in the playing process Thus, this paper suggests a stream selection policy for the transcoding based on EPOB (End Point of Over Bandwidth). It aims to lower the required bandwidth of multimedia streams than the network bandwidth level and also to minimize early-stage delay time for multimedia streams, which is to be played in mobile terminals.
Car Frame Extraction using Background Frame in Video
Nam, Seok-Woo ; Oh, Hea-Seok ;
The KIPS Transactions:PartB, volume 10B, issue 6, 2003, Pages 705~710
DOI : 10.3745/KIPSTB.2003.10B.6.705
Recent years, as a rapid development of multimedia technology, video database system to retrieve video data efficiently seems to core technology in the oriented society. This thesis describes an efficient automatic frame detection and location method for content based retrieval of video. Frame extraction part is consist of incoming / outgoing car frame extraction and car number frame extraction stage. We gain star/end time of car video also car number frames. Frames are selected at fixed time interval from video and key frames are selected by color scale histogram and edge operation method. Car frame recognized can be searched by content based retrieval method.
Development of Face Tracking System Using Skin Color and Facial Shape
Lee, Hyung-Soo ;
The KIPS Transactions:PartB, volume 10B, issue 6, 2003, Pages 711~718
DOI : 10.3745/KIPSTB.2003.10B.6.711
In this paper, we propose a robust face tracking algorithm. It is based on Condensation algorithm  and uses skin color and facial shape as the observation measure. It is hard to integrate color weight and shape weight. So we propose the method that has two separate trackers which uses skin color and facial shape as the observation measure respectively. One tracker tracks skin colored region and the other tracks facial shape. We used importance sampling technique to limit sampling region of two trackers. For skin-colored region tracker, we propose an adaptive color model to avoid the effect of illumination change. The proposed face tracker performs robustly in clutter background and in the illumination changes.
An Improvement Of Efficiency For kNN By Using A Heuristic
Lee, Jae-Moon ;
The KIPS Transactions:PartB, volume 10B, issue 6, 2003, Pages 719~724
DOI : 10.3745/KIPSTB.2003.10B.6.719
This paper proposed a heuristic to enhance the speed of kNN without loss of its accuracy. The proposed heuristic minimizes the computation of the similarity between two documents which is the dominant factor in kNN. To do this, the paper proposes a method to calculate the upper limit of the similarity and to sort the training documents. The proposed heuristic was implemented on the existing framework of the text categorization, so called, AI :: Categorizer and it was compared with the conventional kNN with the well-known data, Router-21578. The comparisons show that the proposed heuristic outperforms kNN about 30∼40% with respect to the execution time.