One of the interferences between inheritance and concurrency is inheritance anomaly. From the view of cognitive comp.tational neurosci.nce, a direct information representation method is presented based on neural syste...
详细信息
One of the interferences between inheritance and concurrency is inheritance anomaly. From the view of cognitive comp.tational neurosci.nce, a direct information representation method is presented based on neural system dynamics and graphic theory. A group of neurons and their connections representing perceptual information directly and the dynamical behaviors of neurons are defined firstly, and then a two-layer neural network is designed to record characteristics of stimulus and connect a specialized neural circuit that responding to the perception of that stimulus respectively. This could be achieved by the structure learning algorithm. The circuit constituted by neurons in two layers is also served as an associative memory of stimulus whose credibility is decided by the degree of connection of the circuit. The direct representation method is of very significance to the research of semantic representation and inference driven by semantics in artificial intelligence.
An image-based rendering method that can successfully support the representation of 3-D surface details and view motion parallax is presented. Expensive task of hole-filling is finished beforehand in the preprocessing...
详细信息
An image-based rendering method that can successfully support the representation of 3-D surface details and view motion parallax is presented. Expensive task of hole-filling is finished beforehand in the preprocessing stage, only texture-mapping hardware is employed in the rendering stage for acceleration. For correct representation, each pixel in a parallel projection image with depth is decomp.sed into several consecutive layers so that a bar texture is generated for that pixel. Since all the bar textures as a whole constitute a continuous 3-D approximation to the scene represented in the original depth image, novel views can be obtained by mapping these bar textures in a from-back-to-front order. Experimental results show the new approach can produce high visual quality and runs faster than many existing image-based rendering techniques with only moderate storage overhead.
A novel unilevel formulation for reliability based design optimization is developed. In the formulation, the lower level optimization is replaced by its corresponding first order Karush-Kuhn-Tucker (KKT) necessary opt...
详细信息
A novel unilevel formulation for reliability based design optimization is developed. In the formulation, the lower level optimization is replaced by its corresponding first order Karush-Kuhn-Tucker (KKT) necessary optimality conditions at the upper level optimization. The replacement is equivalent to solving the original nested optimization if the constraint qualification conditions are satisfied. The proposed formulation is found to be numerically robust and comp.tationally efficient as comp.red to the existing approaches for reliability based design optimization.
The knowledge organization of ALICE and its deduction mechanism are analyzed. Some shortcomings can arise when ALICE is used in processing Chinese. Relevant solutions on extended semantic grammar, pattern-matching alg...
详细信息
The knowledge organization of ALICE and its deduction mechanism are analyzed. Some shortcomings can arise when ALICE is used in processing Chinese. Relevant solutions on extended semantic grammar, pattern-matching algorithm and dynamic answer extraction are proposed. A domain independent Chinese natural language interface CNLIS is implemented based on ALICE. Experiments prove that CNLIS can be easily combined with domain knowledge and achieve 91.45% precision and 91.70% recall.
The aim of this paper is to make SVMs (Support Vector Machines) more applicable to on-line text categorization applications. As SVMs are of good generation ability even with small training sets and text feature vector...
详细信息
The aim of this paper is to make SVMs (Support Vector Machines) more applicable to on-line text categorization applications. As SVMs are of good generation ability even with small training sets and text feature vectors are clustery in the feature space, an algorithm for text categorization, namely, semantic Support Vector Machine (Semantic SVM), is proposed by substituting the original training text set with the semantic center set. This semantic center set is used as the training text and support vector candidates. The steps to generate the semantic center set and the framework of the on-line learning algorithm of semantic SVM are then presented, as well as the implementation of the on-line learning algorithm based on Sequential Minimal Optimization. Experimental results show that, comp.red with the standard SVMs, the proposed semantic SVM and its algorithm can improve the on-line learning speed and the classifying speed by orders with a high classifying veracity.
An omni-directional camera combining hyperbolic mirror with regular lens is implemented. After an analysis of reflection mirror design, image formation and camera architecture, a design method of the omni-directional ...
详细信息
An omni-directional camera combining hyperbolic mirror with regular lens is implemented. After an analysis of reflection mirror design, image formation and camera architecture, a design method of the omni-directional camera having a single view point is proposed. A method of omni-directional image warping making use of the re-projection of points in the coordinate system of real image plane to a virtual image plane is proposed. Experimental results confirmed the validity of the proposed methods. The image captured by the omni-directional camera can be used in video surveillance, object tracking, and stereo vision.
An important problem in biology informatics, that is, protein structure prediction, is studied in this paper. Enlightened by the law of reciprocity among things in the physics world, a continuous model in 2D Euclidean...
详细信息
An important problem in biology informatics, that is, protein structure prediction, is studied in this paper. Enlightened by the law of reciprocity among things in the physics world, a continuous model in 2D Euclidean space is given which gain an advantage over the discretion model in some aspects. The continuous model's strong point is probably that it utilizes an external 'inartificial leading' which is the gravitation among hydrophobe amino acid. Accordingly, in the presupposition of the equivalent result, the continuous model redounds to improving the comp.ting rapidity. Then, a related quasiphysical algorithm is formulated according to the 2D continuous model. Finally some experiment results are given, which also prove the advantages of the continuous model and its quasiphysical algorithm.
Intrusion Tolerance System has become another essential technique to protect Internet servers against.intrusion succeeding to the firewall and IDS. The article proposes a parallel model of intrusion tolerance system (...
详细信息
Intrusion Tolerance System has become another essential technique to protect Internet servers against.intrusion succeeding to the firewall and IDS. The article proposes a parallel model of intrusion tolerance system (ITS): RC model. This model classifies the effects of intrusions on system resources and controls when attacks happen. It divides the resources into 3 levels: measurable resource, information resource and attribute resource, and divides the control status into 2 levels: code level and active level. The 5 levels can characterize the system security state accurately from different views, though they are abstract. The article also describes how to analyze an ITS using RC model, and provides some recovery methods and inst.nces of RC Model.
IPDV is an important performance metric of IP network which reflects network path's workload characteristics. Based on a path queueing model, it is found that the reflection IPDV has on path workload is influenced...
详细信息
IPDV is an important performance metric of IP network which reflects network path's workload characteristics. Based on a path queueing model, it is found that the reflection IPDV has on path workload is influenced by the interval of probes' sending time;the IPDV measured by the existing method just reflects path workload superficially for inappropriately selected reference delay and the limitations of the sampling method. The packet-pair sampling measurement method proposed can improve the reflection capability by selecting the proper reference delay and avoiding the limitations of the sampling method. The minimum interval between the two packets in a packet pair is also determined. Experiments on an Internet path validate the improved reflection capability of the packet-pair sampling method and the interval time's influence on it.
This paper explores new challenges and possible approaches for developing mobile information systems, with an emphasis on model-based approaches on the conceptual and logical levels. Over the last few years, we have e...
详细信息
This paper explores new challenges and possible approaches for developing mobile information systems, with an emphasis on model-based approaches on the conceptual and logical levels. Over the last few years, we have experienced these new challenges through our involvement in several research and industrial projects on mobile solutions, usability and model-based approaches. We summarise the main challenges of how model-based approaches can support the development of mobile information systems that are to be used together with other types of systems in a primarily professional setting and indicate upcoming research issues in this very dynamic area.
暂无评论