MELM-GRBF: A modified version of the extreme learning machine for generalized radial basis function neural networks

In this paper, we propose a methodology for training a new model of artificial neural network called the generalized radial basis function (GRBF) neural network. This model is based on generalized Gaussian distribution, which parametrizes the Gaussian distribution by adding a new parameter τ. The generalized radial basis function allows different radial basis functions to be represented by updating the new parameter τ. For example, when GRBF takes a value of τ=2, it represents the standard Gaussian radial basis function. The model parameters are optimized through a modified version of the extreme learning machine (ELM) algorithm. In the methodology proposed (MELM-GRBF), the centers of each GRBF were taken randomly from the patterns of the training set and the radius and τ values were determined analytically, taking into account that the model must fulfil two constraints: locality and coverage. An thorough experimental study is presented to test its overall performance. Fifteen datasets were considered, including binary and multi-class problems, all of them taken from the UCI repository. The MELM-GRBF was compared to ELM with sigmoidal, hard-limit, triangular basis and radial basis functions in the hidden layer and to the ELM-RBF methodology proposed by Huang et al. (2004) [1]. The MELM-GRBF obtained better results in accuracy than the corresponding sigmoidal, hard-limit, triangular basis and radial basis functions for almost all datasets, producing the highest mean accuracy rank when compared with these other basis functions for all datasets.     

A historical perspective of algorithmic lateral inhibition and accumulative computation in computer vision

Certainly, one of the prominent ideas of Professor José Mira was that it is absolutely mandatory to specify the mechanisms and/or processes underlying each task and inference mentioned in an architecture in order to make operational that architecture. The conjecture of the last fifteen years of joint research has been that any bottom-up organization may be made operational using two biologically inspired methods called “algorithmic lateral inhibition”, a generalization of lateral inhibition anatomical circuits, and “accumulative computation”, a working memory related to the temporal evolution of the membrane potential. This paper is dedicated to the computational formulation of both methods. Finally, all of the works of our group related to this methodological approximation are mentioned and summarized, showing that all of them support the validity of this approximation.     

Designing adaptive learning itineraries using features modelling and swarm intelligence

In this paper, Bayesian network (BN) and ant colony optimization (ACO) techniques are combined in order to find the best path through a graph representing all available itineraries to acquire a professional competence. The combination of these methods allows us to design a dynamic learning path, useful in a rapidly changing world. One of the most important advances in this work, apart from the variable amount of pheromones, is the automatic processing of the learning graph. This processing is carried out by the learning management system and helps towards understanding the learning process as a competence-oriented itinerary instead of a stand-alone course. The amount of pheromones is calculated by taking into account the results acquired in the last completed course in relation to the minimum score required and by feeding this into the learning tree in order to obtain a relative impact on the path taken by the student. A BN is used to predict the probability of success, by taking historical data and student profiles into account. Usually, these profiles are defined beforehand; however, in our approach, some characteristics of these profiles, such as the level of knowledge, are classified automatically through supervised and/or unsupervised learning. By using ACO and BN, a fitness function, responsible for automatically selecting the next course in the learning graph, is defined. This is done by generating a path which maximizes the probability of each user??s success on the course. Therefore, the path can change in order to adapt itself to learners?? preferences and needs, by taking into account the pedagogical weight of each learning unit and the social behaviour of the system.     

Shape-sensitive MLS deformation

This work presents methods for deforming meshes in a shape-sensitive way using Moving Least Squares (MLS) optimization. It extends an approach for deforming space (Cuno et al. in Proceedings of the 27th Computer Graphics International Conference, pp. 115–122, 2007) by showing how custom distance metrics may be used to achieve deformations which preserve the overall mesh shape. Several variant formulations are discussed and demonstrated, including the use of geodesic distances, distances constrained to paths contained in the mesh, the use of skeletons, and a reformulation of the MLS scheme which makes it possible to affect the bending behavior of the deformation. Finally, aspects of the implementation of these techniques in parallel architectures such as GPUs (graphics processing units) are described and compared with CPU-only implementations.     

A retunable PID multi-rate controller for a networked control system

This paper introduces a control strategy based on retuning a multi-rate PID controller in accordance with the variable delays detected in a networked control system, in order to avoid a decreased control performance. The basic idea is minimising the first-order Taylor terms of a performance measure via gain scheduling, i.e., making the controller gains delay dependent. As network delay is time-variant, the stability of this control approach will be proved by means of linear matrix inequalities.     

On the use of small training sets for neural network-based characterization of mixed pixels in remotely sensed hyperspectral images

In this work, neural network-based models involved in hyperspectral image spectra separation are considered. Focus is on how to select the most highly informative samples for effectively training the neural architecture. This issue is addressed here by several new algorithms for intelligent selection of training samples: (1) a border-training algorithm (BTA) which selects training samples located in the vicinity of the hyperplanes that can optimally separate the classes; (2) a mixed-signature algorithm (MSA) which selects the most spectrally mixed pixels in the hyperspectral data as training samples; and (3) a morphological-erosion algorithm (MEA) which incorporates spatial information (via mathematical morphology concepts) to select spectrally mixed training samples located in spatially homogeneous regions. These algorithms, along with other standard techniques based on orthogonal projections and a simple Maximin-distance algorithm, are used to train a multi-layer perceptron (MLP), selected in this work as a representative neural architecture for spectral mixture analysis. Experimental results are provided using both a database of nonlinear mixed spectra with absolute ground truth and a set of real hyperspectral images, collected at different altitudes by the digital airborne imaging spectrometer (DAIS 7915) and reflective optics system imaging spectrometer (ROSIS) operating simultaneously at multiple spatial resolutions.     

Epitomic Location Recognition

This paper presents a novel method for location recognition, which exploits an epitomic representation to achieve both high efficiency and good generalization. A generative model based on epitomic image analysis captures the appearance and geometric structure of an environment while allowing for variations due to motion, occlusions, and non-Lambertian effects. The ability to model translation and scale invariance together with the fusion of diverse visual features yields enhanced generalization with economical training. Experiments on both existing and new labeled image databases result in recognition accuracy superior to state of the art with real-time computational performance.     

Invariant-driven specifications in Maude

This work presents a general mechanism for executing specifications that comply with given invariants, which may be expressed in different formalisms and logics. We exploit Maude’s reflective capabilities and its properties as a general semantic framework to provide a generic strategy that allows us to execute Maude specifications taking into account user-defined invariants. The strategy is parameterized by the invariants and by the logic in which such invariants are expressed. We experiment with different logics, providing examples for propositional logic, (finite future time) linear temporal logic and metric temporal logic.     

Learning a color distance metric for region-based image segmentation

In this paper we describe an experiment where we studied empirically the application of a learned distance metric to be used as discrimination function for an established color image segmentation algorithm. For this purpose we chose the Mumford–Shah energy functional and the Mahalanobis distance metric. The objective was to test our approach in an objective and quantifiable way on this specific algorithm employing this particular distance model, without making generalization claims. The empirical validation of the results was performed in two experiments: one applying the resulting segmentation method on a subset of the Berkeley Image Database, an exemplar image set possessing ground-truths and validating the results against the ground-truths using two well-known inter-cluster validation methods, namely, the Rand and BGM indexes, and another experiment using images of the same context divided into training and testing set, where the distance metric is learned from the training set and then applied to segment all the images. The obtained results suggest that the use of the specified learned distance metric provides better and more robust segmentations, even if no other modification of the segmentation algorithm is performed.     

TextonBoost for Image Understanding: Multi-Class Object Recognition and Segmentation by Jointly Modeling Texture,Layout, and Context

This paper details a new approach for learning a discriminative model of object classes, incorporating texture, layout, and context information efficiently. The learned model is used for automatic visual understanding and semantic segmentation of photographs. Our discriminative model exploits texture-layout filters, novel features based on textons, which jointly model patterns of texture and their spatial layout. Unary classification and feature selection is achieved using shared boosting to give an efficient classifier which can be applied to a large number of classes. Accurate image segmentation is achieved by incorporating the unary classifier in a conditional random field, which (i) captures the spatial interactions between class labels of neighboring pixels, and (ii) improves the segmentation of specific object instances. Efficient training of the model on large datasets is achieved by exploiting both random feature selection and piecewise training methods. High classification and segmentation accuracy is demonstrated on four varied databases: (i) the MSRC 21-class database containing photographs of real objects viewed under general lighting conditions, poses and viewpoints, (ii) the 7-class Corel subset and (iii) the 7-class Sowerby database used in He et al. (Proceeding of IEEE Conference on Computer Vision and Pattern Recognition, vol. 2, pp. 695–702, June 2004), and (iv) a set of video sequences of television shows. The proposed algorithm gives competitive and visually pleasing results for objects that are highly textured (grass, trees, etc.), highly structured (cars, faces, bicycles, airplanes, etc.), and even articulated (body, cow, etc.). J. Shotton is now working at Toshiba Corporate Research & Development Center, Kawasaki, Japan.