M. A. van Wyk

A RKHS interpolator-based graph matching algorithm

We present an algorithm for performing attributed graph matching. This algorithm is derived from a generalized framework for describing functionally expanded interpolators which is based on the theory of reproducing kernel Hilbert spaces (RKHS). The algorithm incorporates a general approach to a wide class of graph matching problems based on attributed graphs, allowing the structure of the graphs to be based on multiple sets of attributes. No assumption is made about the adjacency structure of the graphs to be matched.

Kronecker product graph matching

Abstract In this paper the Interpolator-based Kronecker product graph matching (IBKPGM) algorithm for performing attributed graph matching is presented. The IBKPGM algorithm is based on the Kronecker product graph matching (KPGM) formulation. This new formulation incorporates a general approach to a wide class of graph matching problems based on attributed graphs, allowing the structure of the graphs to be based on multiple sets of attributes. Salient features of the IBKPGM algorithm are that no assumption is made about the adjacency structure of the graphs to be matched, and that the explicit calculation of compatibility values between all vertices of the reference and input graphs as well as between all edges of the reference and input graphs are avoided.

A framework for multiscale and hybrid RKHS-based approximators

A generalized framework for deriving multiscale and hybrid functionally expanded approximators that are linear in the adjustable weights is presented. The basic idea here is to define one or more appropriate function spaces and then to impose a geometric structure on these to obtain reproducing kernel Hilbert spaces (RKHSs). The weight identification problem is formulated as a minimum norm optimization problem that produces an approximation network structure that comprises a linear weighted sum of displaced reproducing kernels fed by the input signals. Examples of the application of this framework are discussed. Results of numerical experiments are presented.

Viterbi decoding and the power spectral densities of some rate one half binary DC free modulation codes

The Hamming distance properties are investigated, and some experimental results obtained with the following R = 1/2 binary dc free modulation codes are presented: the (b, l, C) = (1, 5, 3) Miller2code and codes with (b, l, C), respectively, (1, 4, 3), (0, 3, 3), and (0, 1, 2). A R = 3/6, (b, l, C) = (0, 3, 2) combined error-correcting/ modulation code is also investigated. State systems, power spectral densities, and the bit error rates after computer simulated decoding of these codes on both the binary symmetric channel and a burst erasure channel are presented.

PHASE VOCODER TECHNOLOGY FOR THE SIMULATION OF ENGINE SOUND

For a driving simulator which should be an exact replica of a certain vehicle, an accurate sound model is of extreme importance. The most games select between three or more prerecorded engine sounds, depending on the engine speed. Other methods use linear interpolation between engine sounds for a more accurate approximation, but this is still not ideal. By using vocoders, a technique used for the manipulation of voice, a much higher level of accuracy and realism can be obtained. This article proposes the use of vocoders for the modeling of engine sound for driving simulation and computer driving games.

MATHEMATICAL MODELING OF AN AUTOMATIC DRIVER

The design of driving simulators is common practice within the simulation industry. Normally, the focus is on the modeling of realistic vehicle dynamics models. However, the design of a realistic simulation environment is of equal importance. A human driver usually steers one vehicle, but the rest of the vehicles used in the simulation should be managed by a computer program. In this article, an automatic driver model to be used within the simulation environment, is described. The automatic driver uses the same vehicle dynamics model as the human driver would use. It also uses the vehicle characteristics in such a way to obtain the optimal performance of the vehicle.

A LEARNING-BASED FRAMEWORK FOR GRAPH MATCHING

This paper presents a unifying review of a learning-based framework for kernel-based attributed graph matching. The framework, which includes as special cases the RKHS Interplator-Based Graph Matching (RIGM) and Interpolator-Based Kronecker Product Graph Matching (IBKPGM) algorithms, incorporates a general approach where no assumption is made about the adjacency structure of the graphs to be matched. Corresponding pairs of attributed adjacency matrices and attribute vectors of an input and reference graph are used as the input–output training set of a constrained multi-input multi-output multi-variable mapping to be learned. It is shown that a Reproducing Kernel Hilbert Space (RKHS) based interpolator can be used to infer this mapping. Partially constraining the inferred mapping by the generation of additional consistency input–output training pairs and the use of polynomial feature augmentation lead to improved performance. The proposed learning-based framework avoids the explicit calculation of compatibility values.

Chaotic Maps, Control Parameter, and Liapunov Exponent

We describe a one-dimensional chaotic map wherethe Liapunov exponent is a smooth function of a controlparameter.

An evaluation of several data reduction methods for an absolute capacitance measurement experiment

Certain types of electromagnetic interference mitigation techniques require knowledge of the absolute capacitance of a circuit node. A technique has been developed in order to measure this capacitance under operating conditions in the circuit. The procedure is however very sensitive to measurement errors, and does not directly yield good results. In order to combat this, a large number of similar measurements are made with certain parameter variations. This leads to a large amount of data, which must be reduced in order to arrive at the best possible estimate of the absolute capacitance. In this paper several data reduction methods are applied to this type of problem in order to find the most suitable one. Monte Carlo methods are applied to the data reduction methods in order to evaluate their sensitivity to measurement uncertainty

A projection-based joint probabilistic data association algorithm [multi-target tracking applications]

A novel projection-based joint probabilistic data association (PJPDA) algorithm is presented. The JPDA problem is formulated as a constrained optimization problem and solved using projection onto convex sets (POCS) methods. The PJPDA algorithm is extremely simple to implement, has no strenuous memory requirements and completely avoids the generation of feasible event matrices. Algorithm complexity is discussed and simulation results presented