Current Issue : July - September Volume : 2014 Issue Number : 3 Articles : 5 Articles
This paper presents a rectification algorithm for multi-view images captured by a matrix multi-camera arrangement. In this work,\nan ergodic method is proposed to trace the ideal rotation matrix and the preferred location of the optical centers. Moreover, the\ncamera�s intrinsic parameters are optimized to obtain the ideal projection matrix, from which the image rectifying transform\nmatrix(IRTM) is further derived. Then the multi-view image is rectified with the IRTM. Experimental results show that this\nproposed algorithm has great effect on multi-view image rectification for matrix camera arrangement....
Capacity utilization rate is one of the most important indicators of the efficiency of the manufacturing industry, and therefore\nof the return of the investments made. Estimation of these rates accurately renders it possible to make important economic\ndecisions such as taking sectorial investment decisions, defining the optimal distribution of sectorial credits, determining noncompetitive\nsectors, making development plans and developing unemployment policies. In this study, we estimated the capacity\nutilization rates of 21 sub-sectors of the Turkish manufacturing industry using support vector machines and compared the results\nwith the results obtained from the methods of artificial neural networks and vector auto-regression. This study is the first in the\nliterature in that it was carried out using this method....
Optical measurement devices for eye movements are generally expensive and it is often necessary to restrict user head movements\nwhen various eye-gaze input interfaces are used. Previously, we proposed a novel eye-gesture input interface that utilized\nelectrooculography amplified via an AC coupling that does not require a head mounted display. Instead, combinations of eyegaze\ndisplacement direction were used as the selection criteria. When used, this interface showed a success rate approximately\n97.2%, but it was necessary for the user to declare his or her intention to perform an eye gesture by blinking or pressing an enter\nkey. In this paper, we propose a novel eye-glance input interface that can consistently recognize glance behavior without a prior\ndeclaration, and provide a decision algorithm that we believe is suitable for eye-glance input interfaces such as small smartphone\nscreens. In experiments using our improved eye-glance input interface, we achieved a detection rate of approximately 91% and\na direction determination success rate of approximately 85%. A smartphone screen design for use with the eye-glance input\ninterface is also proposed....
Ordinal decision problems are very common in real-life. As a result, ordinal classification models have drawn much attention\nin recent years. Many ordinal problem domains assume that the output is monotonously related to the input, and some ordinal\ndata mining models ensure this property while classifying. However, no one has ever reported how accurate these models are\nin presence of varying levels of non-monotone noise. In order to do that researchers need an easy-to-use tool for generating\nartificial ordinal datasets which contain both an arbitrary monotone pattern as well as user-specified levels of non-monotone\nnoise. An algorithm that generates such datasets is presented here in detail for the first time. Two versions of the algorithm\nare discussed. The first is more time consuming. It generates purely monotone datasets as the base of the computation. Later,\nnon-monotone noise is incrementally inserted to the dataset. The second version is basically similar, but it is significantly faster.\nIt begins with the generation of almost monotone datasets before introducing the noise. Theoretical and empirical studies of the\ntwo versions are provided, showing that the second, faster, algorithm is sufficient for almost all practical applications. Some\nuseful information about the two algorithms and suggestions for further research are also discussed....
Reinforcement learning requires information about states, actions, and outcomes as the basis for learning. For many applications,\nit can be difficult to construct a representative model of the environment, either due to lack of required information or because\nof that the model�s state space may become too large to allow a solution in a reasonable amount of time, using the experience of\nprior actions. An environment consisting solely of the occurrence or nonoccurrence of specific events attributable to a human\nactor may appear to lack the necessary structure for the positioning of responding agents in time and space using reinforcement\nlearning. Digital pheromones can be used to synthetically augment such an environment with event sequence information to create\namore persistent and measurable imprint on the environment that supports reinforcement learning.We implemented this method\nand combined it with the ability of agents to learn from actions not taken, a concept known as fictive learning. This approach was\ntested against the historical sequence of Somali maritime pirate attacks from2005 to mid-2012, enabling a set of autonomous agents\nrepresenting naval vessels to successfully respond to an average of 333 of the 899 pirate attacks, outperforming the historical record\nof 139 successes...
Loading....