Search

In this paper, we present an ontology definition which is helpful in symbolising the ontology and ontology management. By investigating interactions between agents in business scenarios, we adopt the process algebra to model interactions and corresponding effects. For the purpose of solving the conflicts or inconsistency in understanding a particular issue, we investigate the agent negotiation mechanisms on the basis of process algebra. Moreover, private strategies in reaching agreements are discussed. Our approach reveals that reflections from other agents in negotiations can be captured and in turn provide sufficient information about how to manage the ontology along with proposed activities in ontology evolution. [ABSTRACT FROM AUTHOR]

In this paper we present a novel approach that customize the BDI model to define a so-called "RBDIA: Rapport-Belief-Desire-Intention-Adaptation" as a generic method to support progress from individual autonomous agent concept towards a collaborative multiple agents. Rapport here refers to the component that connects an agent to its environment, whereas Adaptation module incorporates mechanisms of learning. The contribution of this paper is twofold: first, we develop a novel modeling approach that enable us to combine the internal and social structures of collaborative multigent, and second the proposed methodology is applied to a real-world application for assistance in product development process. We believe that the five proposed tiers for multiagent systems (MAS) development serves for mastering the complexity and the difficulty of setting up effective autonomous collaborative MAS. [ABSTRACT FROM AUTHOR]

The paper presents some properties of the communication network supported by the intelligent agents. The intelligent agents were placed into network nodes and they were immobile. They were used to regulate the transfer of mobile agents from the network input, through the routing nodes and finally, towards the service processing nodes. These nodes were the programs running on the computers. A new model of distributed and collaborating intelligent agents was designed and presented. Continuous adaptation and agent's collaboration was achieved by reinforcement Q-learning. For such a model the results show rapid tendency to reduce state time of mobile agents in the service. The obstructive effects expansion to the other agents when a change of processing capabilities in the region occurs was detected and described. [ABSTRACT FROM AUTHOR]

This paper presents a model of a learning mechanism for situated agents. The learning is described explicitly in terms of plans and conducted as intentional actions within the BDI (Beliefs, Desires, Intentions) agent model. Actions of learning direct the task-level performance towards improvements or some learning goals. The agent is capable of modifying its own plans through a set of actions on the run. The use of domain independent patterns of actions is introduced as a strategy for constraining the search for the appropriate structure of plans. The model is demonstrated to represent Q-learning algorithm, however different variation of pattern can enhance the learning. [ABSTRACT FROM AUTHOR]

The Data-Intensive Web Sites provide access to a large number of Web pages whose content is dynamically extracted from structured databases. They may be used for shopping or paying in e-commerce, for reading news in a newspaper site or to consult digital library. In this context, users often need rich content and fresh data. Several techniques have been developed to meet the demand for faster and more efficient access to the DIWS. Among them a main role is acquired by the replication, the caching, the materialization, and the refreshing of data. Software agents have proved to be a good tool that may give a high performance results on the Web. In this paper, we address their cases of use in the DIWS. We discuss cases in which agents can be used to improve the data management performance. The aim is to specify tasks that may profit from the increase development in agent technologies. [ABSTRACT FROM AUTHOR]

The use of probabilistic reasoning is a key capability in information fusion systems for a variety of domains such as military situation assessment. In this paper, we discuss two key approaches to probabilistic reasoning in military situation assessment: Probabilistic Relational Models and Object Oriented Probabilistic Relational Models. We compare the modeling and inferencing capabilities of these two languages and compare these capabilities against the requirements of the military situation assessment domain. [ABSTRACT FROM AUTHOR]

Currently, the major focus on the network security is securing individual components as well as preventing unauthorized access to network service. In this paper, we propose a network blocking algorithm and architecture, which provides a comprehensive, host-by-host perspective on IP over Ethernet network security. The proposed system is also an effective tool for managing network resources containing IP address, MAC address and hostname, etc. under diverse and complicated network environment. Therefore, we will be able to enhance local network security on the wired and wireless environment with the network resource and security management system based on our proposed framework. [ABSTRACT FROM AUTHOR]

This paper presents an effective strategy to extract the buildings and to reconstruct 3-D buildings using high-resolution multispectral stereo satellite images. Proposed scheme containes three major steps: building enhancement and segmentation using both Background Discriminant Transformation (BDT) and ISODATA algorithm, conjugate building identification using the object matching with Hausdorff distance and color indexing, and 3-D building reconstruction using photogrammetric techniques. IKONOS multispectral stereo images were used to evaluate the scheme. As a result, the BDT technique was verified as an effective tool for enhancing building areas since BDT suppressed the dominance of background to enhance the building as a non-background. In building recognition, color information itself was not enough to identify the conjugate building pairs since most buildings are composed of similar materials such as concrete. When both Hausdorff distance for edge information and color indexing for color information were combined, all segmented buildings in the stereo images were correctly identified. Finally, 3-D building models were successfully generated using the space intersection by the forward Rational Function Model (RFM). [ABSTRACT FROM AUTHOR]

Theoretic methods for travel speed estimation are common in the traffic literature, but treatments of how to account delay caused by traffic signal are less common. This paper is an attempt to improve conventional technique for evaluating travel speed in signalized intersection networks. To do so, this paper employs new concept which is associated with the vehicles' trajectory in the time-space diagram. The most useful advantage of the algorithm proposed in this paper is that, the traffic analysts can reduce much tedious works for screening and selecting a bunch of speed data. [ABSTRACT FROM AUTHOR]

The Variable Message Signs (VMS) are useful way to reduce the socio-economic costs due to the traffic congestions and delays by providing the information on traffic condition to drivers. This paper provided a methodology to determine the locations of VMS's in terms of the minimization of the delay by applying the genetic algorithm. The optimal number of VMS's was also derived by the economic analysis based on the cost and the benefit. The simulation considered the variation of traffic volume, the frequency and duration of the incident, and the traffic conversion in order to reflect the real situation. [ABSTRACT FROM AUTHOR]

Urban traffic congestion is one of the most serious problems in many large cities of the world. The traffic demand management is the most practicable countermeasure, and the role of urban transit is increasing. The electronic fare card system can enhance the service quality of transit in terms of users' convenience and the speed of operation. In addition, the importance of the inter-modal transit service is increasing, and the fare system between two different modes of transportation is a crucial issue. The city of Seoul in South Korea has recently introduced the integrated transit-fare card system, corresponding to the reform of the urban bus system. This paper focuses on the basic architecture of the fare card system. [ABSTRACT FROM AUTHOR]

This paper presents a real application of Web-content mining using an incremental FP-Growth approach. We firstly restructure the semi-structured data retrieved from the web pages of Chinese car market to fit into the local database, and then employ an incremental algorithm to discover the association rules for the identification of car preference. To find more general regularities, a method of attribute-oriented induction is also utilized to find customer's consumption preferences. Experimental results show some interesting consumption preference patterns that may be beneficial for the government in making policy to encourage and guide car consumption. [ABSTRACT FROM AUTHOR]

Building a forecasting model for time-series data is a tough but very valuable research topic in recent years. High variation of time-series features must be considered appropriately for an accurate prediction. For weather forecasting, which is continuous, dynamic and chaotic, it's difficult to extract the most important information present in the knowledge base and determine the importance of each feature. In this paper, taking tropical cyclone (TC) as an example, we present an integrated similarity retrieval model to forecast the intensity of a tropical cyclone using neural network, which is adopted to generate a set of appropriate weights for various associated features of a tropical cyclone. A time adjustment function is used for time-series consideration. The experimental results show that this integrated approach can achieve a better performance. [ABSTRACT FROM AUTHOR]

In this paper, a fuzzy case based reasoning approach to Web access path prediction is developed and tested. It is based on the assumption that new user's access patterns can be predicted by referencing to the behaviors of similar Web users in the past. This method has three phases. Firstly, a Web case base is constructed from the Web log data. This includes the pre-processing and cleaning of the Web log data so that a suitable format is developed. Secondly, contextual information is extracted from the Web pages, and this information is used to develop a similarity measurement between Web pages. This information is added to the Web case base. Finally, fuzzy association rule mining is used to discover the relationship between the browsing behavior (user navigations) and the Web contents using the Web case base. A set of predictive cases from the Web case base is then selected for the access path prediction. From the experimental evaluation, our approach has demonstrated better prediction accuracy than the existing approaches. [ABSTRACT FROM AUTHOR]

The information security of computer networks has become a serious global issue and also a hot topic in computer networking researches. Many approaches have been proposed to tackle these problems, especially the denial of service (DoS) attacks. The Multiple Classifier System (MCS) is one of the approaches that have been adopted in the detection of DoS attacks recently. For a MCS to perform better than a single classifier, it is crucial for the base classifiers which embedded in the MCS to be diverse. Both resampling, e.g. bagging, and feature grouping could promote diversity of base classifiers. In this paper, we propose an approach to select the reduced feature group for each of the base classifiers in the MCS based on the mutual information between the features and class labels. The base classifiers being combined using the weighted sum is examined in this paper. Different feature grouping methods are evaluated theoretically and experimentally. [ABSTRACT FROM AUTHOR]

This paper proposes a multimedia ontology that will support semantic integration of multimedia databases. The ontology is defined by integrating MPEG-7 MDS and TV-AnyTime metadata. A procedure to integrate these two standards is described and relationship of their elements is also presented. We also designed and implemented a framework for integrating multimedia databases and showed that any multimedia metadata can be mapped to the proposed ontology. [ABSTRACT FROM AUTHOR]

WWW is currently undergoing a remarkable change from a collection of pages to a collection of services that interoperate through the Internet. The representative types of these services are XML Web Services and ebXML, which are emerging as significant e-Business application frameworks. Both ebXML and Web Services, however, use different registries for publishing and discovery of services, and this has given rise to some problems when adopting frameworks. Furthermore, these registries support only keyword-based search, which does not make use of semantic information, and which does not address the problem of matching service capabilities and allowing service location based on the functionalities sought. This paper shows that service descriptions in OWL-S can be mapped into UDDI and ebXML registry entries, and can be migrated from registries; therefore, it provides not only a way to integrate registry entries but also a way to build the foundation of semantic service discovery. Also, migrated OWL-S can be updated periodically while registry entries are being modified. [ABSTRACT FROM AUTHOR]

In this paper we describe the design of a hybrid search that combines simple metadata search with a traditional keyword search over unstructured context data. This paradigm provides the inquirer additional options to narrow the search with some semantic aspects through the XML metadata query. We demonstrate a paper search for a case study of the hybrid search, and describe a data integration hub to integrate those data dispersed on the Net. [ABSTRACT FROM AUTHOR]

In this paper, we propose a concept mapping method that can be used to find relations among concepts of different ontology. In order to find the relevant relation, this method uses three kinds of knowledge types such as lexical knowledge, domain knowledge and structure information. When the relations are retrieved, each relation is evaluated by comparing instances of each concept which is involved in that relation. With this method, the concept of one ontology can be mapped to other concept of the other ontology, which enables information sharing. [ABSTRACT FROM AUTHOR]

The enterprise information systems (EIS) offer the cornerstone for managing enterprise business, applying strategic and economical decisions, and holding communication with partners. Bringing systems to work together is becoming increasingly essential for leveraging the enterprise information systems and reaching common goals. Meanwhile, semantic sharing represents the daunting barrier for making these systems work together through a more convenient global architecture and providing interoperability and reusability. This paper concentrates on studying the application of tightening together context and ontologies which can serve as formal background for reaching a suitable global enterprise environment. It brings along architecture and implementation, with special interest for the reusability of components between these systems. [ABSTRACT FROM AUTHOR]

Fuzzy ontology mapping is important for handling uncertain knowledge on the semantic web. However, current ontology mapping technologies are not sufficient for fuzzy ontologies. This paper proposes a framework of mapping fuzzy concepts between fuzzy ontologies. It applies the approximate concept mapping approach, extends atom fuzzy concept sets and defines the least upper bounds to reduce the searching space. It resolves the mapping problem of fuzzy concepts into finding the simplified least upper bounds for atom fuzzy concepts, and gives an algorithm for searching the simplified least upper bounds, which is fast and proved correct. The framework is efficient for mapping fuzzy concepts between fuzzy ontologies. [ABSTRACT FROM AUTHOR]

Shape comparison between 3D models is essential for shape recognition, retrieval, classification, etc. In this paper, we propose a method for comparing 3D shapes, which is invariant under translation, rotation and scaling of models and is robust to non-uniformly distributed and incomplete data sets. A modal model is constructed from input data using vibration modes and then shape similarity is evaluated with modal strain energy. The proposed method provides global-to-local ordering of shape deformation using vibration modes ordered by frequency. Thus, we evaluated similarity in terms of global properties of shape without being affected localized shape features using ordered shape representation and modal strain energy. [ABSTRACT FROM AUTHOR]

In this paper, we propose a new camera calibration method for the 3D-based image synthesis and 3D reconstruction. We improve the problem as changing the principle point for obtaining the linear equation. According to the error rate, we adapt the non-linear method that minimizes the intrinsic parameters. Namely, it minimizes the intrinsic parameters error with maintaining the computational conciseness. As a result, we can find optimized camera intrinsic parameters and adapt to image synthesis and reconstruction. Experimental results show the performance of the proposed method is the better than the previous. We also demonstrate examples of the 3D-based image synthesis and 3D reconstruction from uncalibrated images. [ABSTRACT FROM AUTHOR]

Although airbags in the car play an important role for the safety of occupants, in fact, many peoples have been injured or killed by the deployment of airbags themselves. Such conventional airbags are deployed by the shock sensors. As an alternative approach, a vision-based smart airbag system could be promising. This paper describes a new method by which 3-D pose of the car occupant can be recognized. We combine 2-D head tracking information with a disparity map of the occupant for 3-D pose tracking. Result shows that the system can locate the head position around the passenger's seat with a real-time basis. [ABSTRACT FROM AUTHOR]

In this paper, an approach that uses a combination of neural network classifiers (CNNC) is applied to human face recognition. We present a divide-and-conquer approach for system composed of several separate networks. Decomposing the complex problem into sub-problems for solving them by a binary base classifier is presented. Each of that learns to recognize a subject of the complete set of training database. Combining the results of sub-problems with max rule accomplished to achieve better performance. The recognition rate of 100% for ORL and Yale database was obtained using the mentioned devised algorithm. Keywords: Face Recognition, Neural Networks, Classifier Combination, Multilayer perceptron, Decomposition [ABSTRACT FROM AUTHOR]

It is known that deformation of the hippocampus shape is involved with several neurological diseases. In this paper, we propose a hybrid shape representation scheme, which consists of multiresolution skeletons, voxels and meshes for the shape analysis of the hippocampus. Initially, a hippocampal surface model is reconstructed from MRI and then it is placed into a canonical coordinate system, where the position, orientation and scaling are normalized. From the voxel representation of the hippocampus, multiresolution skeletons are extracted and Iterative Closest Point normalization is carried out. Then the shape similarity of two hippocampal models is computed with a hierarchical fashion. In addition, we have implemented a neural network based classifier to discriminate whether a hippocampal model is normal or not. Results indicate that the proposed hybrid representation and the skeleton-based normalization using ICP are very effective in 3D shape analysis of the hippocampus. [ABSTRACT FROM AUTHOR]

In this paper, we propose a semantic supervised clustering approach to classify lands in geo-images. We use the Maximum Likelihood Method to generate the clustering. In addition, we complement the analysis applying spatial semantics to improve the classification. The approach considers the a priori knowledge of the multispectral image to define the training sites (classes) related to the geographic environment. In this case the spatial semantics is defined by the spatial properties, functions and relations that involve the geo-image. By using these characteristics, it is possible to determine the training data sites with a priori knowledge. This method attempts to improve the supervised clustering, adding the intrinsic semantics of the geo-images to determine the training sites that involve the analysis with more precision. [ABSTRACT FROM AUTHOR]

In this paper, we present the design of an intelligent web service for ontology-based query-answering dialogues. This web service is realized through OWL-QL message exchanges between the querying agent and the answering agent. The OWL-QL is a formal language and protocol for query-answering dialogues among Semantic Web agents using knowledge represented in the standard Ontology Web Language(OWL). In OWL-QL settings, the answering agent uses automated reasoning methods to derive answers to queries. We explain the organizational details of each agent and then discuss the usefulness of the graphical OWL-QL query composer included in the querying agent. [ABSTRACT FROM AUTHOR]

Intermediary agents extend the functionality and performance without violating the principles employed in the design of the Web. Such agents intervene in the client-server interactions shaping the response from the web server before it reaches the client. In this paper, we study and characterize the problem of choosing the intermediary agent nearest to the web client. Agent-client proximity is considered in the context of low latency and high bandwidth. Each intermediary agent can support every client but the nearest agent is the best choice. We performed extensive experiments on the Internet. We show that the geographic distance for selecting the nearest intermediary agent is accurate and effective. [ABSTRACT FROM AUTHOR]

Data mining methodology and tools are employed in different application areas. This paper proposes a novel application field for data mining research, namely analysis and long-term forecasting of Internet performance, especially for the needs of Computational Grids. Using data mining the performance problems studied for Internet can be considered from new points of view, and sometimes with better understanding than through applying conventional data analysis methods. This knowledge has been mined by means of professional data mining package in order to build the decision model for advising in further exploitation and usage scheduling of Grid links for a particular time and date. The results show that the data mining can be efficiently used in this research area. [ABSTRACT FROM AUTHOR]

Through integration of Web services into agents, we can obtain enhanced functionality and credibility. If agents could access and use Web services, the agents might offer them to other agents as extended services. Additionally, if Web service clients and servers could access and use agent services, agent developers would be able to offer the benefits of agent services to a Web service environment. In this paper, we propose Agent Service to Web Service Gateway as a means for increasing availability and usefulness of legacy biological information resources. Specifically, we illustrates as an example Protein Data Base(PDB), which is one of the most important resources for structural proteomics research. We explain the agentification of PDB using a wrapper and then describe how to provide the corresponding Web service through the service gateway in detail. [ABSTRACT FROM AUTHOR]

In this paper, we present a computational approach to understanding and augmenting the conversational knowledge process. We introduce the concept of the conversation quantization, a technique of approximating a continuous flow of conversation by a series of conversation quanta that represent points of the discourse. To investigate what the nature of conversation quanta is, we attempt to extract conversation quanta from two types of the meeting videos by hand. As a result, we have obtained some profitable suggestions about conversation quanta. [ABSTRACT FROM AUTHOR]

With the help of real source identity in packets, network security system can intelligently protect and counteract the attacks. Packet marking is an important method of source identification, and there are some issues on it. For large amount of packets, analysis time and complicated computation are necessary while detect marking information. This paper focuses on this direction, and proposes a simple and efficient method to mark all packets belonging to upstream traffic with a deterministic, plain form identity. With this approach, we just need low processing power on some specific edge routers as well as a little extra network traffic to settle it. Furthermore, distilling mark from packets is easy since the mark is in plain text format. [ABSTRACT FROM AUTHOR]

This paper presents a set of multilevel-secure optimistic concurrency control (MLS/OCC) scheme that has several desirable properties: If lower-level transactions were somehow allowed to continue with its execution in spite of the conflict of high-level transactions, covert timing channel-freeness would be satisfied. This sort of optimistic approach for conflict insensitiveness and the properties of non-blocking and deadlock freedom make the optimistic concurrency control scheme especially attractive to multilevel-secure transaction processing. Unlike pessimistic approaches, the MLS/OCC scheme never delays or rejects an operation submitted by a lower-level transaction which is passed the mandatory access control. Instead, the read and write operations are processed freely without updating the actual database. Therefore, it is reasonable to assert that MLS/OCC scheme is allowed to avoid the abort of lower-level transactions in order to close covert timing channel, nevertheless guaranteeing conflict-preserving serializability. The basic refinement philosophy for the solution on starvation problem is an incorporation of multiple versions of low-level datainto MLS/OCC. This kind of intelligent channel-free concurrency control scheme satisfies the B3 or higher level of the US TCSEC requirements. [ABSTRACT FROM AUTHOR]

Keystroke dynamics is an intelligent data processing technique of analyzing the user's habitual typing patterns to identify him. Keystroke dynamics combined with password authentication has been widely used as a means to enhance user authentication system. However, the user authentication system's security does not rely solely on the keystroke dynamics. To guarantee a high level of security, more secure password authentication is needed. The design and development of a secure password authentication protocol for keystroke dynamics is discussed in this paper. We propose a new efficient password authentication protocol that is secure against all types of attacks considered in the paper. We also show that our two-party protocol is extended to a three-party protocol, where each user only shares a password with a trusted server. As a result, our protocols with keystroke dynamics can provide a secure and intelligent means of authentication and access control of computer users. [ABSTRACT FROM AUTHOR]

In applications of multimedia communication, the remote user authentication is an important issue to protect the multimedia resources. Recently, Lin-Lai proposed an improvement on the Lee-Ryu-Yoo fingerprint-based remote user authentication scheme using smart cards to not only keep the original advantages but also enhance their security. Lin-Lai claimed that their scheme can withstand masquerade attacks and secure access control in multimedia communication. However, we have found that the scheme is still susceptible to impersonation attacks. Accordingly, the current paper demonstrates the vulnerability of the Lin-Lai scheme to impersonation attacks and presents an improved scheme with better security strength and efficiency. The computational costs of the proposed scheme are less than those of the Lin-Lai scheme. Keyword: Security techniques, Multimedia, User authentication, Fingerprint verification, Biometrics, Smart card [ABSTRACT FROM AUTHOR]

In this paper, an efficient moving object extraction algorithm for surveillance application is proposed which employ change detection strategy to obtain motion information of moving-object instead of complex operator. In addition, background subtraction is introduced to solve the problems of still object and uncovered background which is generally ill-inherency existed in conventional method. After that, the internal part region of moving-object may be confused with real-static region due to frame difference used. Hence, we use the concept of region adjacent graphic to overcome it. Finally, a post-processing step is used to remove noise regions and refine the shape of objects segmented. Moreover shadow effects can be suppressed in the pre-processing step. Experimental results demonstrate various results of segmented video sequence for both indoor and outdoor scenes and show that the proposed algorithm is superior to others in terms of obviating static region of internal part of moving-object and edge defects. [ABSTRACT FROM AUTHOR]

The mean and standard deviation of the magnitudes of coefficients of the transform domain are common used in texture image retrieval. This paper performs an experimental comparison on Gabor wavelet and wavelet frame based features for this application. The detail comparison results indicate the performance of the former is not superior over that of the last in the query-by-example retrieval. [ABSTRACT FROM AUTHOR]

In this paper, we propose motion pattern-based descriptor which exploits both spatial and temporal features to characterize video sequences in a semantics-based manner. The Discrete Cosine Transform (DCT) is applied to convert the high-level features from the time domain to the frequency domain. The energy concentration property of DCT allows us to use only a few DCT coefficients to precisely capture the variations of moving blobs. In comparison with the frequently used motion activity descriptors, the RLD and SAH of MPEG-7, the proposed descriptor yields 38% and 19% average gains over RLD and SAH, respectively. [ABSTRACT FROM AUTHOR]

Recently, fingerprint has been received considerable attentions as a user authentication technique using biometrics. Fingerprint retrieval, which retrieves a candidate list of fingerprints having similar features with the given fingerprint from a fingerprint database, is one of interesting real-time applications. However, using the straightforward approach, it takes a long response time to scan the entire database and to compare the query against each reference on a large-scale database. Even when implemented on a hardware, it is hard to satisfy the real-time requirement. In this paper, to reduce the response time, we employ the geometric hashing technique which has been widely used for model-based object recognition. Using this algorithm, the number of fingerprint minutiae can be restricted. It limits the hardware size used for implementation to fit into FPGA-chips. Furthermore, we develop a scalable implementation for parallel geometric hashing on an FPGA-based PCI platform. Keywords: Fingerprint Retrieval, Real-Time Processing, Special-Purpose Hardware [ABSTRACT FROM AUTHOR]

It is a very important issue to quickly retrieve semantic information from a vast multimedia database. In this paper, we propose a statistic-based algorithm to retrieve the videos that contain the requested object motion from video database. In order to speed up our algorithm, we only utilize the local motion embedded in the region-of-interest as the query to retrieve data from MPEG bitstreams. Experimental results demonstrate that our fast video retrieval algorithm is powerful in terms of accuracy and efficiency. [ABSTRACT FROM AUTHOR]

A favorable difference metric is crucial to the shot boundary detection (SBD) performance. In this paper, we propose a new set of metrics, information theoretic metrics, to quantitatively measure the changes between frames. It includes image entropy difference, joint entropy, conditional entropy, mutual information and divergence. They all can be used to cut detection. Specially, the image entropy and joint entropy are good clues to fade detection, while mutual information, joint entropy and conditional entropy are less sensitive to illumination variations. The theoretic analysis and experimental results show that they are useful in SBD. [ABSTRACT FROM AUTHOR]

Increasing amount of illicit image data transmitted via the internet has triggered the need to develop effective image mining systems for digital forensics purposes. This paper discusses the requirements of digital image forensics which underpin the design of our forensic image mining system. This system can be trained by a hierarchical Support Vector Machine (SVM) to detect objects and scenes which are made up of components under spatial or non-spatial constraints. Forensic investigators can communicate with the system via a grammar which allows object description for training, searching, querying and relevance feedback. In addition, we propose to use a Bayesian networks approach to deal with information uncertainties which are inherent in forensic work. These inference networks will be constructed to model probability interactions between beliefs, adapt to different users' retrieval patterns, and mimic human judgement of semantic content of image patches. An analysis of the performance of the first prototype of the system is also provided. [ABSTRACT FROM AUTHOR]

Vector quantization (VQ) is widely used in low bit rate image compression. In this paper, two predictive vector quantization (PVQ) algorithms that combine the concept of side-match are proposed. By controlling the quantization distortion after encoding or searching the reconstructed vector with minimum side distortion before encoding, the two proposed algorithms can decrease the quantization distortion and computational complexity respectively. The bit rates are also reduced in both algorithms by using side-math technology. The performances of the proposed algorithms are compared with several previous VQ algorithms. Simulation results have shown the efficiency of the proposed algorithms. [ABSTRACT FROM AUTHOR]

A new scheme for vector quantization (VQ) is proposed in this paper. We employ side-match and classified criteria for designing VQ codebooks to combat blocking effects induced from high compression rates, and then we use the proposed algorithm to improve the reconstructed image quality. Simulation results demonstrate the better image quality to compare with that produced from conventional schemes objectively and subjectively, at the cost of reasonable encoding complexity. [ABSTRACT FROM AUTHOR]

The paper proposed a fast motion estimation algorithm based on the spatial homogeneity and temporal standstill of video sequences. It classified a frame into different regions such as standstill, stationary or dynamic region. For MBs lying in the standstill region, motion estimation is skipped; for MBs lying in the stationary region, motion estimation is applied once to an MB in the region, and the resulting motion vector will be used by other MBs in the same region. Normal motion estimation is only carried out in the dynamic region. The new algorithm can be used either by itself alone, or by combining with other existing fast motion estimation methods. Experimental results show that the new algorithm can significantly improve the time efficiency of the H.264 encoder, and is able to achieve an average reduction of 27% in encoding, with an average PSNR loss of only 0.09 dB and 0.42% bit rate increase, compared with the original H.264 reference software. [ABSTRACT FROM AUTHOR]

One of the new features in the H.264/AVC encoder is the use of Lagrangian Rate-Distortion Optimised (RDO) mode decision at the macroblock (MB) level. This brute-force algorithm searches through the 10 different MB coding modes to obtain the best one for encoding that MB, and is hence computationally expensive. This paper proposes a novel algorithm where the RDO can be reorganized in a better order such that the most likely MB modes will be tried first, and an early termination of the RDO process will be used once the calculated rate-distortion cost (RDcost) is below a preset threshold. The preset threshold is dependent on the RDcost of previous and neighbouring MBs that have been coded. This is based on the observation that the RDcost of an MB in the current frame is highly correlated to a co-located MB in the previous frame. Experimental results have shown that the new algorithm has dramatically reduced the encoding times of up to 61.8%, with negligible increases in bit-rates or PSNR. [ABSTRACT FROM AUTHOR]

JPEG is a well-known method for digital image compression. The drawback of JPEG is distortion to image quality by truncation errors from quantizing the DCT coefficients. Another problem is the damage on the extracted watermarks from JPEG compression for the spatial-domain watermarks. In this paper, we proposed an improved method to enhance the quality of the decompressed image. The purpose of our proposed method is to adjust the reconstructed DCT coefficients closed as their original values. The coefficient adjustment blocks are used to guide the decompression operation during the coefficient reconstruction. Experimental results show that the proposed method has indeed made the image quality improvement as comparing to the images in the traditional method. [ABSTRACT FROM AUTHOR]

This paper proposes novel low-power MPEG-4 motion estimator with deep submicron technologies below 0.13m. The proposed motion estimator reduces both dynamic and static power consumption so that it is suitable for large leakage process technologies. It exploits breaking-off search to reduce dynamic power consumption. To reduce static power consumption, block-wise shutdown method is employed. From the simulation results, power consumption was reduced to about 60%. The proposed motion estimator was designed in Verilog HDL and the estimated gate counts are about 45,000 gates. [ABSTRACT FROM AUTHOR]

This paper presents an artistic rendering system for generating 3D images of Chinese paintings using graphics hardware. The user can adjust suitable parameters flexibly to generate different brush styles as his/her hobby, and see rendering results in real time. In this system, we propose a hardware-accelerated method to draw Chinese painting strokes efficiently along visible silhouettes. Three-dimensional texture and multi-texture from normal graphics hardware is used to speed up generating various brushes with Chinese painting stylized strokes. The features of the traditional Chinese painting such as ink diffusion and moisture effects are simulated. Several examples of aesthetically pleasing Chinese-paintings rendered from 3D models are demonstrated using the proposed method. [ABSTRACT FROM AUTHOR]