1,721,180 research outputs found
Feature selection based on geometric distance for high-dimensional data
No full textA novel feature selection method based on geometric distance is proposed. It utilises both the average distance between classes along with the evenness of these distances to evaluate feature subsets. The feature evaluation and selection process used therein is very easy to understand, because it lends itself to a simple geometrical analysis. Moreover, because the method does not calculate the relevance or redundancy between features, it is faster than other filter methods that use information or statistical dependency concepts. The experiments demonstrate its markedly better classification performance as well as fast computation compared with existing methods.1112Nsciescopu
Robust segment-based object tracking using generalized hyperplane approximation
No full textTracking based on gradient descent algorithm using image gradient is one of the popular object tracking method. However, it easily fails to track when illumination changes. Although several illumination invariant features have been proposed, applying the invariant feature to the gradient descent method is not easy because the invariant feature is represented as a non-linear function of image pixel values and its Jacobian cannot be calculated in a closed-form. To make it possible, we introduce the generalized hyperplane approximation technique and apply it to histogram of oriented gradient (HOG) feature, one of the well-known illumination invariant feature. In addition, we achieve partial occlusion invariance using image segments. The hyperplanes are calculated from training segment images obtained by perturbing the motion parameter around the target region. Then, it is used to map the difference in non-linear feature of image onto the increment of alignment parameters. This process is mathematically same to the gradient descent method. The information from each segment is integrated by a simple weighted linear combination with confidence weights of segments. Compared to the previous tracking algorithms, our method shows very fast and stable tracking results in experiments on several practical image sequences. (C) 2012 Elsevier Ltd. All rights reserved.X1167sciescopu
Automatic rule generation of fuzzy logic controllers based on asynchronous coevolution of rule-level subpopulations
No full textThis paper proposes a rule-level coevolutionary approach based on multiple subpopulations to evolve fuzzy logic controllers (FLCs). Each rule is used as an individual and the subpopulations, each comprising a number of candidate rules, are randomly probed for evolution [asynchronous coevolution] via evolution strategy (ES). The rules belonging to the same subpopulation compete while those in different subpopulations cooperate to achieve the goal of finding a better FLC. During this process, the rules within each subpopulation become specialized into a kind of expert in the corresponding problem domain. For this approach, a simple credit assignment scheme for rule evaluation is introduced to reduce the search space effectively. The superiority of the proposed algorithm over traditional FLC-level evolution approaches has been demonstrated by evolving FLCs for two typical nonlinear control problems - the ball-and-beam and the cart-pole systems.X11sciescopu
Real-time neural network based camera localization and its extension to mobile robot control
No full textThe feasibility of using neural networks for camera localization and mobile robot control is investigated here. This approach has the advantages of eliminating the laborious and error-prone process of imaging system modeling and calibration procedures. Basically, two different approaches of using neural networks are introduced of which one is a hybrid approach combining neural networks and the pinhole-based analytic solution while the other is purely neural network based. These techniques have been tested and compared through both simulation and real-time experiments and are shown to yield more precise localization than analytic approaches. Furthermore, this neural localization method is also shown to be directly applicable to the navigation control of an experimental mobile robot along the hallway purely guided by a dark wall strip. It also facilitates multi-sensor fusion through the use of multiple sensors of different types for control due to the network's capability of learning without models.X111sciescopu
Hybrid control for autonomous mobile robot navigation using neural network based behavior modules and environment classification
No full textA hybrid control architecture combining behavior based reactive navigation and model based environment classification has been developed. It is also hybrid in the sense that both competitive coordination and cooperative coordination are used for the BBC (Behavior Based Control) part. The contributions are as follows. First, a Neural Network (NN) in charge of environment classification has been developed based on 16 prototypes of topological maps roughly describing various local navigation environments. This environment classification NN not only enables the navigator to avoid local minimum points but also eliminates the requirement for prior detailed modeling of the environment since it needs to memorize only "rough" information on local environments encountered along the way that might be sufficient for navigation. Next, an NN based reactive behavior controller will be trained to learn human steering commands for each of the 16 prototype local environments. Third, the modified potential field (MPF) method obtained by adding the free space vector as the third component is used to select a particular reactive behavior in conjunction with the classification NN. Finally, a hybrid control architecture integrating all three of these concepts was developed. It avoids local minimum traps as well as solves the problems of poor obstacle clearance or oscillation. It is robust against sensor noise and adaptive to dynamic environments. This hybrid architecture is also amenable to easy addition of new behaviors due to the modularity of the BBC architecture. The effectiveness of the proposed architecture has been verified through both computer simulation and an actual robot called MORIS (MObile Robot as an Intelligent System).X1133sciescopu
Real-time dynamic visual tracking using PSD sensors and extended trapezoidal motion planning
No full textA real-time visual servo tracking system for an industrial robot has been implemented using PSD (Position Sensitive Detector) cameras, neural networks, and an extended trapezoidal motion planning method. PSD and directly transduces the light's projected position on its sensor plane into an analog current and lends itself to fast real-time tracking. A neural network, after proper training, transforms the PSD sensor reading into a 3D position of the target, which is then input to an extended trapezoidal motion planning algorithm. This algorithm implements a continuous motion update strategy in response to an ever-changing sensor information from the moving target, while greatly reducing the tracking delay. This planning method is found to be very useful for sensor-based control such as moving target tracking or weld-seam tracking in which the robot needs to change its motion in real time in response to incoming sensor information. Further, for real-time usage of the neural net, a new architecture called LANN (Locally Activated Neural Network) has been developed based on the concept of CMAC input partitioning and local learning. Experimental evidence shows that an industrial robot can smoothly track a moving target of unknown motion with speeds of up to 1 m/s and with oscillation frequency up to 5 Hz.X1113sciescopu
A new paradigm for real-time parallel storage and recognition of patterns based on a hierarchical organization of associative memories utilizing Walsh function encoding
No full textA new hierarchical Walsh memory which can store and quickly recognize any number of patterns is proposed. A Walsh function based associative memory was found to be capable of storing and recognizing patterns in parallel via purely a software algorithmic technique (namely, without resorting to parallel hardware) while the memory itself only takes a single pattern space of computer memory, due to the Walsh encoding of each pattern. This type of distributed associative memory lends itself to high speed pattern recognition and has been reported earlier in a single memory version. In this paper, the single memory concept has first been extended to a parallel memory module and then to a tree-shaped hierarchy of these parallel modules that are capable of storing and recognizing any number of patterns for practical large scale data applications exemplified by image and speech recognition. The memory hierarchy was built by successively applying k-means clustering to the training data set. In the proposed architecture, the clustered data subsets are stored respectively into a parallel memory module where the module allocation is optimized using the genetic algorithm to realize a minimal implementation of the memory structure. The system can recognize all the training patterns with 100% accuracy and further, can also generalize on similar data. In order to demonstrate its efficacy with large scale real world data, we stored and recognized over 500 faces while at same time, achieving much reduced recognition time and storage space than template matching.X111sciescopu
Hybrid position and image based visual servoing for mobile robots
No full textVisual servoing requires the target object to be in the field of view of the camera all the time. At the same time, we also want to achieve optimal path planning and controllability of the robot pose. This paper presents a new hybrid fuzzy control method for visual servoing of mobile robots to meet these requirements. IBVS (Image Based Visual Servoing) calculates the motion plan directly from the image space using the inverse image Jacobian so that the target object always stays within the field of view. In contrast, PBVS (Position Based Visual Servoing) uses an image-to-work space transform to plan an optimal pose trajectory directly in the cartesian. The proposed fuzzy control then integrates these two types of visual servoing through a warning signal indicating the target may escape the field of view. Also, we use the neural network for the prediction of the target position for a robust timely tracking of the object. Simulation and real experimental work based on MORIS, our mobile robot test bed, verify the efficacy of this approach.X117sciescopu
Automatic extraction of eye and mouth fields from a face image using eigenfeatures and ensemble networks
No full textThis paper presents a novel algorithm for the extraction of the eye and mouth (facial features) fields from 2D gray level images. Eigenfeatures are derived from the eigenvalues and eigenvectors of the binary edge data set constructed from eye and mouth fields. Such eigenfeatures are ideal features for finely locating fields efficiently. The eigenfeatures are extracted from a set of the positive and negative training samples for facial features and are used to train a multilayer perceptron (MLP) whose output indicates the degree to which a particular image window contains the eyes or the mouth within itself. An ensemble network consisting of a multitude of independent MLPs was used to enhance the generalization performance of a single MLP. It was experimentally verified that the proposed algorithm is robust against facial size and even slight variations of the pose.X1117sciescopu
A constructive design method for two-layer perceptrons and its application to the design of modular neural networks
No full textA multilayer perceptron is known to be capable of approximating any smooth function to any desired accuracy if it has a sufficient number of hidden neurons. But its training, based on the gradient method, is usually a time consuming procedure that may converge toward a local minimum, and furthermore its performance is greatly influenced by the number of hidden neurons and their initial weights. Usually these crucial parameters are determined based on the trial and error procedure, requiring much experience on the designer's part. In this paper, a constructive design method (CDM) has been propose for a two-layer perceptron that can approximate a class of smooth functions whose feature vector classes are linearly separable. Based on the analysis of a given data set sampled from the target function, feature vectors that can characterize the function 'well' are extracted and used to determine the number of hidden neurons and the initial weights of the network. But when the classes of the feature vectors are not linearly separable, the network may not be trained easily, mainly due to the interference among the hyperplanes generated by hidden neurons. Next, to compensate for this interference, a refined version of the modular neural network (MNN) has been proposed where each network module is created by CDM. After the input space has been partitioned into many local regions, a two-layer perceptron constructed by CDM is assigned to each local region. By doing this, the feature vector classes are more likely to become linearly separable in each local region and as a result, the function may be approximated with greatly improved accuracy by MNN. An example simulation illustrates the improvements in learning speed using a smaller number of neurons.X112sciescopu
- …
