Browsing by Author "Lee, SY"

Now showing 1 - 20 of 70

2D C-tree spatial representation for iconic image
(1999-04-01)
Hsu, FJ
;
Lee, SY
;
Lin, BS
;
資訊科學與工程研究所
;
Institute of Computer Science and Engineering
The 2D string approaches provide a natural way of constructing iconic indexing for images. The 2D C-string representation with an efficient cutting mechanism is more characteristic of spatial knowledge and efficient in the representation of iconic images. However, the computation of object ranks in a 2D C-string might make the inference of spatial reasoning somewhat complicated. This shortcoming is overcome by the 2D C-tree representation. The 2D C-tree not only keeps the comprehensive spatial knowledge in the original 2D C-string, but also the ordered labeled tree is more suitable for spatial reasoning and image retrieval. The spatial knowledge can be derived directly from the inference rules embedded in the characteristic structure of the 2D C-tree representation. (C) 1999 Academic Press.
A framework for temporal similarity measures of content-based scene retrieval
(2001-04-01)
Shan, MK
;
Lee, SY
;
資訊科學與工程研究所
;
Institute of Computer Science and Engineering
The distinguished features of video retrieval lie in the similarity measures and content-based retrieval. Most research on content-based video retrieval represents the content of video as a set of frames, leaving out the temporal ordering of frames in the shot. In this paper, the similarity measures of video content are investigated. We propose a series of similarity measures based on the similarity of frame sequence which take temporal ordering into consideration. All the algorithms corresponding to the similarity measures are based on the approach of dynamic programming. (C) 2001 Published by Elsevier Science B.V.
A multi-granularity locking model for concurrency control in object-oriented database systems
(1996-02-01)
Lee, SY
;
Liou, RL
;
資訊科學與工程研究所
;
Institute of Computer Science and Engineering
A locking model adopting a multi-granularity approach is proposed for concurrency control in object-oriented database systems. The model is motivated by a desire to provide high concurrency and low locking overhead in accessing objects. Locking in schemas and locking in instances are developed separately and then are integrated. Schema changes and composite objects are also taken into account. A dual queue scheme for efficient scheduling of lock requests is developed. The model consists of a rich set of lock modes, a compatibility matrix, and a locking protocol. Characteristic query examples on single class, class lattice, and composite objects are used to illustrate the comparison between the ORION model and the proposed model. It is shown that our locking model has indeed made some improvements and is suitable for concurrency control in object-oriented databases.
A new locking-based protocol for mobile transactions
(2001)
Chen, GC
;
Lee, SY
;
資訊工程學系
;
Department of Computer Science
Due to long network delay of wireless links, transactions will be long-lived transactions in mobile computing environments. In such a situation, the occurrence of handoff is inevitable and thus a wireless link held by a mobile unit crossing cell boundaries might be forced to terminate. It is undesirable that an active transaction is forced to terminate. A queueing scheme has been proposed to solve the problem of forced termination of transactions in our previous research. In the queueing scheme, the execution of a transaction which cannot get a channel when handoff occurs, is suspended and is resumed when it gets a channel again. However when two-phase locking (2PL) protocol is employed, suspending an active transaction will elongate the lock, holding time and thus degrade the system performance. In this paper we propose a new locking-based protocol to improve the performance of 2PL in mobile environments.
A novel object detection framework in compressed domain
(2004)
Ahmad, BMA
;
Ahmad, AMA
;
Lee, SY
;
資訊工程學系
;
Department of Computer Science
In this paper we propose a novel approach for robust motion vector based object detection in MPEG-2 video streams. By processing the extracted motion vector fields that are directly extracted from MPEG-2 video streams in the compressed domain, through post processing operation, in order to reduce the noise within the motion vector content, obtain more robust object information, and refine this information through our proposed system which composed of a Spatial filter Component, a Temporal filter Component and a Texture filter component. As a result, the object detection algorithm is more capable of accurately detecting objects with more efficient performance in terms of runtime. We compare the performance of our proposed system with other popular and commonly related work and techniques. Based on the experimental results performed over the MPEG7 testing dataset and measuring performance by using the standard recall and precision metrics, object detection using our proposed system is remarkably superior to the alternative techniques. In addition to these results, we describe a user system interface that we developed, where users can maintain the parameters interactively.
A video information system for sport motion analysis
(1997-06-01)
Chang, CW
;
Lee, SY
;
資訊科學與工程研究所
;
Institute of Computer Science and Engineering
In this paper, we present several important methods in a video information system for sport motion analysis. We first describe an efficient method for computing the position of one or more moving objects in a video sequence. We utilize computer vision routines for both target object tracking and feature value extraction. Then, we propose a multiple subsequence matching mechanism for comparing the difference in object motion between two video sequences. A mapping function, penalty table and the overall distance between two video sequences have been defined. We use the prominent feature points as our segmentation positions and separate the sequence into several subsequence groups. After an alignment process, we can find the mapping of the best matching and get an evaluation report between the two video sequences for motion analysis. The video information system described in this paper has successfully analysed several examples of sport motion, and is representative of techniques commonly used in practice. (C) 1997 Academic Press Limited.
A video information system for sport motion analysis
(1997-06-01)
Chang, CW
;
Lee, SY
;
交大名義發表
;
資訊科學與工程研究所
;
National Chiao Tung University
In this paper, we present several important methods in a video information system for sport motion analysis. We first describe an efficient method for computing the position of one or more moving objects in a video sequence. We utilize computer vision routines for both target object tracking and feature value extraction. Then, we propose a multiple subsequence matching mechanism for comparing the difference in object motion between two video sequences. A mapping function, penalty table and the overall distance between two video sequences have been defined. We use the prominent feature points as our segmentation positions and separate the sequence into several subsequence groups. After an alignment process, we can find the mapping of the best matching and get an evaluation report between the two video sequences for motion analysis. The video information system described in this paper has successfully analysed several examples of sport motion, and is representative of techniques commonly used in practice. (C) 1997 Academic Press Limited.
Allocation of signature file on parallel device for WWW index servers
(1999-03-01)
Shan, MK
;
Lee, SY
;
資訊科學與工程研究所
;
Institute of Computer Science and Engineering
Signature files are widely used in information retrieval and database. They act as search filters for content-based retrieval. In a large database server, a parallel device is utilized to achieve concurrency access. Efficient allocation of signature files on parallel devices minimizes the query response time and is important in the design of access methods for large scale index servers. We have developed an algorithm to organize the storage of signatures in parallel secondary storage to reduce the query response time. First, signature file is clustered into signature pages. Then, the clustered signature pages are distributed among the disks using the parity check matrix of error correcting code in coding theory. Through the construction of error correcting code, the least frequently simultaneously accessed pages are allocated on the same disk. Performance analysis shows that this algorithm improves the efficiency of access.
An analytic model for performance analysis of concurrency control strategies in mobile environments
(1999-01-01)
Chen, GC
;
Lee, SY
;
資訊工程學系
;
Department of Computer Science
Many performance studies have shown that locking based protocols outperform other concurrency control protocols in most database configurations and workloads. However, in mobile environments there are many special characteristics, such as long network delay and the expensive wireless communication. Besides, transactions may be aborted:due to forced termination when handoff occurs for the consideration of concurrency control. The performances of concurrency control strategies should be re-evaluated in such environments, In this paper, we develop an analytic model to evaluate the performances of concurrency control strategies. The accuracy of this model is verified by simulation. According to the experimental results we show that the behavior of mobile users and the degree of data contention have a significant impact on the relative performances of the concurrency control strategies. In particular, we point out that the optimistic concurrency control strategy outperforms other strategies over a wide range of system configurations. We also explain why the optimistic algorithm is well suited in mobile environments.
An analytic model for performance analysis of concurrency control strategies in mobile environments
(1999-01-01)
Chen, GC
;
Lee, SY
;
資訊工程學系
;
Department of Computer Science
Many performance studies have shown that locking based protocols outperform other concurrency control protocols in most database configurations and workloads. However, in mobile environments there are many special characteristics, such as long network delay and the expensive wireless communication. Besides, transactions may be aborted:due to forced termination when handoff occurs for the consideration of concurrency control. The performances of concurrency control strategies should be re-evaluated in such environments, In this paper, we develop an analytic model to evaluate the performances of concurrency control strategies. The accuracy of this model is verified by simulation. According to the experimental results we show that the behavior of mobile users and the degree of data contention have a significant impact on the relative performances of the concurrency control strategies. In particular, we point out that the optimistic concurrency control strategy outperforms other strategies over a wide range of system configurations. We also explain why the optimistic algorithm is well suited in mobile environments.
An efficient scheme of general weighted fair allocation for enhanced best-effort service in high speed networks
(2001-03-01)
Chen, YJ
;
Lee, SY
;
資訊工程學系
;
Department of Computer Science
This research is focused on enhanced best-effort service, which provides the minimum bandwidth guarantee and distributes the available bandwidth fairly to packet flows in a high speed network. These goals can be reached when the bandwidth allocation is general weighted (GW) fair among the flows. GW fairness ensures the minimum rate for every flow and, furthermore, allocates the excess bandwidth (the available bandwidth minus the sum of the minimum rates) to the flows in a way that is proportional to their weights. In the proposed scheme, switches coordinate to efficiently achieve GW fair allocation. In the process of achieving fairness, a switch not only takes O(log N) time for bandwidth allocation among N active flows when receiving feedback from its neighbor node, but also keeps the packet queue length below a target threshold. As a result, better performance in terms of end-to-end delay, packet loss ratio, and throughput can be achieved. A proof to show that the scheme can achieve GW fair allocation is presented. Finally, simulation results illustrate and suggest a good and justifiable system design.
Automatic cel painting in computer-assisted cartoon production using similarity recognition
(1997-07-01)
Chang, CW
;
Lee, SY
;
資訊科學與工程研究所
;
Institute of Computer Science and Engineering
The task of painting each eel of an animated sequence has always been a tedious work using traditional methods. In order to simplify and speed up the process, the development of an automatic eel painting system is desired. This paper presents a method for assisting the production of cartoon animation using computers in order to reduce the production expense and to improve the overall quality. It signals a great improvement in the working environment where 2-D animation Is produced. The basic idea of our automatic eel painting mechanism is to extract both statistical and topological features of regions from input animation drawings, and to search for similar corresponding regions in the previous drawing, If the similarity measures are within the accepted thresholds, the corresponding colour will be inherited. A new integrated labelling algorithm constructing statistical region and topological graph features is proposed. Furthermore, a similarity measure of regions is defined and an algorithm of partial matching of attribute graphs is presented. The experimental result shows the feasibility and practicability of the proposed approach. (C) 1997 by John Wiley & Sons, Ltd.
Automatic cel painting in computer-assisted cartoon production using similarity recognition
(1997-07-01)
Chang, CW
;
Lee, SY
;
資訊科學與工程研究所
;
Institute of Computer Science and Engineering
The task of painting each eel of an animated sequence has always been a tedious work using traditional methods. In order to simplify and speed up the process, the development of an automatic eel painting system is desired. This paper presents a method for assisting the production of cartoon animation using computers in order to reduce the production expense and to improve the overall quality. It signals a great improvement in the working environment where 2-D animation Is produced. The basic idea of our automatic eel painting mechanism is to extract both statistical and topological features of regions from input animation drawings, and to search for similar corresponding regions in the previous drawing, If the similarity measures are within the accepted thresholds, the corresponding colour will be inherited. A new integrated labelling algorithm constructing statistical region and topological graph features is proposed. Furthermore, a similarity measure of regions is defined and an algorithm of partial matching of attribute graphs is presented. The experimental result shows the feasibility and practicability of the proposed approach. (C) 1997 by John Wiley & Sons, Ltd.
Automatic video summary and description
(2000)
Lee, SY
;
Lee, ST
;
Chen, DY
;
資訊工程學系
;
Department of Computer Science
Multimedia technology has been applied to many kinds of applications and the amount of multimedia data is growing dramatically. Especially the usage of digital video data is very popular today. Thus, content-based indexing and retrieval in video is getting more important in the future. In this paper, we investigate an efficient and reasonable solution to describe video contents, and propose an automatic generation of the summary of video contents with as least manual interaction as possible. Based on the summary of video contents, we extract the key features of the summary to represent its content and produce the description of the video content in MPEG-7 format using XML. We illustrate our approach by developing a system that can support content-based browsing and querying on the WWW.
Bounded tag fair queueing for broadband packet switching networks
(2000-01-01)
Chen, YJ
;
Lee, SY
;
資訊工程學系
;
Department of Computer Science
Fair Queueing (FQ) is an attractive packet scheduling mechanism, which can establish firewall among packet flows. Thus, the quality of service (QoS) of each flow can be guaranteed. To implement the mechanism, three important performance issues are concerned: fairness, bounded delay, and efficiency. We propose a scheme, named BTFQ, which satisfies the three criteria. The scheme BTFQ + is also proposed to improve the fairness property of BTFQ and is designed to combine with a low-cost, high-speed hardware for reducing computation time. For comparison, we classify some well-known packet scheduling schemes into six classes according to their fairness properties. BTFQ and BTFQ + are excellent in bounded delay and efficiency. Their fairness properties, although a little bit weak, are but still very good if the traffic load is under 95%, according to the simulation results. Because of the tradeoff among the performance issues, BTFQ and BTFQ + provide another good choices. (C) 2000 Elsevier Science B.V. All rights reserved.
Buffer-sharing techniques in service-guaranteed video servers
(1999-09-01)
Tsai, WJ
;
Lee, SY
;
資訊工程學系
;
Department of Computer Science
Advances in networking and storage technology have made it possible to deliver on-demand services over networks such as the emerging video-on- demand (VOD) applications. Although a variety of studies have been focused on designing video servers suitable for VOD applications, the number of concurrent on-demand services supported by such servers is often limited by the I/O bandwidth of the storage systems. Recently, several researchers have focused on providing guaranteed services instead of on-demand services in a VOD system. A service is said to be guaranteed if every viewer can be served within a specified waiting time after the video has been subscribed. This paper describes a model called a buffer-sharing autonomy, which combines batching and bridging techniques to provide guaranteed services to VOD systems. We also describe how buffer-sharing techniques can be used in this model to provide the same services with fewer buffer resource through the use of playback-rate alteration. Additionally, in order to make buffer sharing efficient, it is important to group appropriate video streams together to share buffer resource. Four grouping policies are proposed for this purpose and the benefits yielded are also analyzed and compared. Finally, the results of our experiments show that the proposed buffer-sharing techniques promise the provision of guaranteed video services at low cost.
Clustering databases for user location management in PCS network
(1999)
Chen, GC
;
Lee, SY
;
資訊工程學系
;
Department of Computer Science
Personal Communication Service (PCS) network is the integration of cellular Network and conventional wired network. To support user mobility, the location information of the mobile stations (MSs) should be constantly tracked using a database. The widespread deployment of PCS will lead to a tremendous increase in the number of updates and queries to the location database. The centralized stand-alone Home Location Register (HLR) in GSM can be a potential bottleneck. Besides, in case of the location database failure, incoming calls may be lost Replicating a HLR cart increase the capacity and reliability of the HLR. However replicating the HLR will cause extra cost because the location update must be executed in each replica. In this paper; we propose a clustering technique to reduce the cost of location updating in replicated HLR. The performance evaluation model is also presented.
Content-based video retrieval based on similarity of frame sequence
(1998)
Shan, MK
;
Lee, SY
;
資訊科學與工程研究所
;
Institute of Computer Science and Engineering
Video retrieval is one of the important design issues of multimedia database management systems. The distinguished features of video retrieval lie in the similarity measures and content-based retrieval. Most research on content-based video retrieval represents the content of video as a set of frames leaving out the temporal ordering of frames in the shot. In this paper, the similarity measures of video content are investigated. We propose a series of similarity measures based on the similarity of frame sequence which take temporal ordering into consideration. The corresponding algorithms are also presented. The effectiveness of the developed similarity measures measured by precision and recall is also described.
Content-based video retrieval via motion trajectories
(1998)
Shan, MK
;
Lee, SY
;
資訊科學與工程研究所
;
Institute of Computer Science and Engineering
Motion is one of the most prominent features of video. For content-based video retrieval, motion trajectory is the intuitive specification of motion features. In this paper, approaches for video retrieval via single motion trajectory and multiple motion trajectories are addressed. For the retrieval via single motion trajectory, the trajectory is modeled as a sequence of segments and each segment is represented as the slope. Two quantitative similarity measures and corresponding algorithms based on the sequence similarity are presented. For the retrieval via multiple motion trajectories, the trajectories of the video are modeled as a sequence of symbolic pictures. Four quantitative similarity measures and algorithms, which are also based on the sequence similarity, are proposed. All the proposed algorithms are developed based on the dynamic programming approach.
Defeating the blocking effect in the filter steered robust compressed domain object detection in MPEG videos
(2004)
Ahmad, AMA
;
Lee, SY
;
資訊工程學系
;
Department of Computer Science
In this paper we propose a novel and an elegant approach for defeating Blocking Effect in MPEG-2 video streams. We first take an MPEG2 video stream, extract the motion vectors from P-frames Meanwhile, we will extract the DCT coefficients from I frames, those coefficient include the DC coefficient, and the AC components as well. Then, we will pass the DCT coefficients into a module to calculate the texture of each frame. Later, we will propagate this texture information into P frame using Inverse Motion Compensation Technique. After that, we will change each motion vector based on its texture value as what will be described later. We compare the performance of our system Based on experimental results over MPEG7 testing dataset, based on the standard recall and precision as metrics, our system performance is remarkable and strongly effective. We develop user system interface, where users can maintain the system parameters interactively.