Parallel Computer Vision Algorithms for Graphics Processing Units

La evolución de los teléfonos móviles inteligentes, dotados de cámaras digitales, está provocando una creciente demanda de aplicaciones cada vez más complejas que necesitan algoritmos de visión artificial en tiempo real; puesto que el tamaño de las señales de vídeo no hace sino aumentar y en cambio el rendimiento de los procesadores de un solo núcleo se ha estancado, los nuevos algoritmos que se diseñen para visión artificial han de ser paralelos para poder ejecutarse en múltiples procesadores y ser computacionalmente escalables. Una de las clases de procesadores más interesantes en la actualidad se encuentra en las tarjetas gráficas (GPU), que son dispositivos que ofrecen un alto grado de paralelismo, un excelente rendimiento numérico y una creciente versatilidad, lo que los hace interesantes para llevar a cabo computación científica. En esta tesis se exploran dos aplicaciones de visión artificial que revisten una gran complejidad computacional y no pueden ser ejecutadas en tiempo real empleando procesadores tradicionales. En cambio, como se demuestra en esta tesis, la paralelización de las distintas subtareas y su implementación sobre una GPU arrojan los resultados deseados de ejecución con tasas de refresco interactivas. Asimismo, se propone una técnica para la evaluación rápida de funciones de complejidad arbitraria especialmente indicada para su uso en una GPU. En primer lugar se estudia la aplicación de técnicas de síntesis de imágenes virtuales a partir de únicamente dos cámaras lejanas y no paralelas—en contraste con la configuración habitual en TV 3D de cámaras cercanas y paralelas—con información de color y profundidad. Empleando filtros de mediana modificados para la elaboración de un mapa de profundidad virtual y proyecciones inversas, se comprueba que estas técnicas son adecuadas para una libre elección del punto de vista. Además, se demuestra que la codificación de la información de profundidad con respecto a un sistema de referencia global es sumamente perjudicial y debería ser evitada. Por otro lado se propone un sistema de detección de objetos móviles basado en técnicas de estimación de densidad con funciones locales. Este tipo de técnicas es muy adecuada para el modelado de escenas complejas con fondos multimodales, pero ha recibido poco uso debido a su gran complejidad computacional. El sistema propuesto, implementado en tiempo real sobre una GPU, incluye propuestas para la estimación dinámica de los anchos de banda de las funciones locales, actualización selectiva del modelo de fondo, actualización de la posición de las muestras de referencia del modelo de primer plano empleando un filtro de partículas multirregión y selección automática de regiones de interés para reducir el coste computacional. Los resultados, evaluados sobre diversas bases de datos y comparados con otros algoritmos del estado del arte, demuestran la gran versatilidad y calidad de la propuesta. Finalmente se propone un método para la aproximación de funciones arbitrarias empleando funciones continuas lineales a tramos, especialmente indicada para su implementación en una GPU mediante el uso de las unidades de filtraje de texturas, normalmente no utilizadas para cómputo numérico. La propuesta incluye un riguroso análisis matemático del error cometido en la aproximación en función del número de muestras empleadas, así como un método para la obtención de una partición cuasióptima del dominio de la función para minimizar el error. ABSTRACT The evolution of smartphones, all equipped with digital cameras, is driving a growing demand for ever more complex applications that need to rely on real-time computer vision algorithms. However, video signals are only increasing in size, whereas the performance of single-core processors has somewhat stagnated in the past few years. Consequently, new computer vision algorithms will need to be parallel to run on multiple processors and be computationally scalable. One of the most promising classes of processors nowadays can be found in graphics processing units (GPU). These are devices offering a high parallelism degree, excellent numerical performance and increasing versatility, which makes them interesting to run scientific computations. In this thesis, we explore two computer vision applications with a high computational complexity that precludes them from running in real time on traditional uniprocessors. However, we show that by parallelizing subtasks and implementing them on a GPU, both applications attain their goals of running at interactive frame rates. In addition, we propose a technique for fast evaluation of arbitrarily complex functions, specially designed for GPU implementation. First, we explore the application of depth-image–based rendering techniques to the unusual configuration of two convergent, wide baseline cameras, in contrast to the usual configuration used in 3D TV, which are narrow baseline, parallel cameras. By using a backward mapping approach with a depth inpainting scheme based on median filters, we show that these techniques are adequate for free viewpoint video applications. In addition, we show that referring depth information to a global reference system is ill-advised and should be avoided. Then, we propose a background subtraction system based on kernel density estimation techniques. These techniques are very adequate for modelling complex scenes featuring multimodal backgrounds, but have not been so popular due to their huge computational and memory complexity. The proposed system, implemented in real time on a GPU, features novel proposals for dynamic kernel bandwidth estimation for the background model, selective update of the background model, update of the position of reference samples of the foreground model using a multi-region particle filter, and automatic selection of regions of interest to reduce computational cost. The results, evaluated on several databases and compared to other state-of-the-art algorithms, demonstrate the high quality and versatility of our proposal. Finally, we propose a general method for the approximation of arbitrarily complex functions using continuous piecewise linear functions, specially formulated for GPU implementation by leveraging their texture filtering units, normally unused for numerical computation. Our proposal features a rigorous mathematical analysis of the approximation error in function of the number of samples, as well as a method to obtain a suboptimal partition of the domain of the function to minimize approximation error.

Dimensionality Reduction Through Sub-Space Mapping for Nearest Neighbour Algorithms

Postprint

Exploring automatic feature selection for transition-based dependency parsing

En este artículo se investigan técnicas automáticas para encontrar un modelo óptimo de características en el caso de un analizador de dependencias basado en transiciones. Mostramos un estudio comparativo entre algoritmos de búsqueda, sistemas de validación y reglas de decisión demostrando al mismo tiempo que usando nuestros métodos es posible conseguir modelos complejos que proporcionan mejores resultados que los modelos que siguen configuraciones por defecto.

Improving kNN multi-label classification in Prototype Selection scenarios using class proposals

Prototype Selection (PS) algorithms allow a faster Nearest Neighbor classification by keeping only the most profitable prototypes of the training set. In turn, these schemes typically lower the performance accuracy. In this work a new strategy for multi-label classifications tasks is proposed to solve this accuracy drop without the need of using all the training set. For that, given a new instance, the PS algorithm is used as a fast recommender system which retrieves the most likely classes. Then, the actual classification is performed only considering the prototypes from the initial training set belonging to the suggested classes. Results show that this strategy provides a large set of trade-off solutions which fills the gap between PS-based classification efficiency and conventional kNN accuracy. Furthermore, this scheme is not only able to, at best, reach the performance of conventional kNN with barely a third of distances computed, but it does also outperform the latter in noisy scenarios, proving to be a much more robust approach.

On the suitability of Prototype Selection methods for kNN classification with distributed data

In the current Information Age, data production and processing demands are ever increasing. This has motivated the appearance of large-scale distributed information. This phenomenon also applies to Pattern Recognition so that classic and common algorithms, such as the k-Nearest Neighbour, are unable to be used. To improve the efficiency of this classifier, Prototype Selection (PS) strategies can be used. Nevertheless, current PS algorithms were not designed to deal with distributed data, and their performance is therefore unknown under these conditions. This work is devoted to carrying out an experimental study on a simulated framework in which PS strategies can be compared under classical conditions as well as those expected in distributed scenarios. Our results report a general behaviour that is degraded as conditions approach to more realistic scenarios. However, our experiments also show that some methods are able to achieve a fairly similar performance to that of the non-distributed scenario. Thus, although there is a clear need for developing specific PS methodologies and algorithms for tackling these situations, those that reported a higher robustness against such conditions may be good candidates from which to start.

A study of methods for selection of quotient digits during digital division,

Vita.

The dynamics of a genetic algorithm under stabilizing selection

A formalism recently introduced by Prugel-Bennett and Shapiro uses the methods of statistical mechanics to model the dynamics of genetic algorithms. To be of more general interest than the test cases they consider. In this paper, the technique is applied to the subset sum problem, which is a combinatorial optimization problem with a strongly non-linear energy (fitness) function and many local minima under single spin flip dynamics. It is a problem which exhibits an interesting dynamics, reminiscent of stabilizing selection in population biology. The dynamics are solved under certain simplifying assumptions and are reduced to a set of difference equations for a small number of relevant quantities. The quantities used are the population's cumulants, which describe its shape, and the mean correlation within the population, which measures the microscopic similarity of population members. Including the mean correlation allows a better description of the population than the cumulants alone would provide and represents a new and important extension of the technique. The formalism includes finite population effects and describes problems of realistic size. The theory is shown to agree closely to simulations of a real genetic algorithm and the mean best energy is accurately predicted.

Noisy fitness evaluation in genetic algorithms and the dynamics of learning

A theoretical model is presented which describes selection in a genetic algorithm (GA) under a stochastic fitness measure and correctly accounts for finite population effects. Although this model describes a number of selection schemes, we only consider Boltzmann selection in detail here as results for this form of selection are particularly transparent when fitness is corrupted by additive Gaussian noise. Finite population effects are shown to be of fundamental importance in this case, as the noise has no effect in the infinite population limit. In the limit of weak selection we show how the effects of any Gaussian noise can be removed by increasing the population size appropriately. The theory is tested on two closely related problems: the one-max problem corrupted by Gaussian noise and generalization in a perceptron with binary weights. The averaged dynamics can be accurately modelled for both problems using a formalism which describes the dynamics of the GA using methods from statistical mechanics. The second problem is a simple example of a learning problem and by considering this problem we show how the accurate characterization of noise in the fitness evaluation may be relevant in machine learning. The training error (negative fitness) is the number of misclassified training examples in a batch and can be considered as a noisy version of the generalization error if an independent batch is used for each evaluation. The noise is due to the finite batch size and in the limit of large problem size and weak selection we show how the effect of this noise can be removed by increasing the population size. This allows the optimal batch size to be determined, which minimizes computation time as well as the total number of training examples required.

Gaussian processes for classification: mean field algorithms:mean-field algorithms

We derive a mean field algorithm for binary classification with Gaussian processes which is based on the TAP approach originally proposed in Statistical Physics of disordered systems. The theory also yields an approximate leave-one-out estimator for the generalization error which is computed with no extra computational cost. We show that from the TAP approach, it is possible to derive both a simpler 'naive' mean field theory and support vector machines (SVM) as limiting cases. For both mean field algorithms and support vectors machines, simulation results for three small benchmark data sets are presented. They show 1. that one may get state of the art performance by using the leave-one-out estimator for model selection and 2. the built-in leave-one-out estimators are extremely precise when compared to the exact leave-one-out estimate. The latter result is a taken as a strong support for the internal consistency of the mean field approach.

Data visualization with simultaneous feature selection

Data visualization algorithms and feature selection techniques are both widely used in bioinformatics but as distinct analytical approaches. Until now there has been no method of measuring feature saliency while training a data visualization model. We derive a generative topographic mapping (GTM) based data visualization approach which estimates feature saliency simultaneously with the training of the visualization model. The approach not only provides a better projection by modeling irrelevant features with a separate noise model but also gives feature saliency values which help the user to assess the significance of each feature. We compare the quality of projection obtained using the new approach with the projections from traditional GTM and self-organizing maps (SOM) algorithms. The results obtained on a synthetic and a real-life chemoinformatics dataset demonstrate that the proposed approach successfully identifies feature significance and provides coherent (compact) projections. © 2006 IEEE.

Robust automatic mapping algorithms in a network monitoring scenario

Automatically generating maps of a measured variable of interest can be problematic. In this work we focus on the monitoring network context where observations are collected and reported by a network of sensors, and are then transformed into interpolated maps for use in decision making. Using traditional geostatistical methods, estimating the covariance structure of data collected in an emergency situation can be difficult. Variogram determination, whether by method-of-moment estimators or by maximum likelihood, is very sensitive to extreme values. Even when a monitoring network is in a routine mode of operation, sensors can sporadically malfunction and report extreme values. If this extreme data destabilises the model, causing the covariance structure of the observed data to be incorrectly estimated, the generated maps will be of little value, and the uncertainty estimates in particular will be misleading. Marchant and Lark [2007] propose a REML estimator for the covariance, which is shown to work on small data sets with a manual selection of the damping parameter in the robust likelihood. We show how this can be extended to allow treatment of large data sets together with an automated approach to all parameter estimation. The projected process kriging framework of Ingram et al. [2007] is extended to allow the use of robust likelihood functions, including the two component Gaussian and the Huber function. We show how our algorithm is further refined to reduce the computational complexity while at the same time minimising any loss of information. To show the benefits of this method, we use data collected from radiation monitoring networks across Europe. We compare our results to those obtained from traditional kriging methodologies and include comparisons with Box-Cox transformations of the data. We discuss the issue of whether to treat or ignore extreme values, making the distinction between the robust methods which ignore outliers and transformation methods which treat them as part of the (transformed) process. Using a case study, based on an extreme radiological events over a large area, we show how radiation data collected from monitoring networks can be analysed automatically and then used to generate reliable maps to inform decision making. We show the limitations of the methods and discuss potential extensions to remedy these.

Decision-making algorithms in geometric modelling

Many of the applications of geometric modelling are concerned with the computation of well-defined properties of the model. The applications which have received less attention are those which address questions to which there is no unique answer. This thesis describes such an application: the automatic production of a dimensioned engineering drawing. One distinctive feature of this operation is the requirement for sophisticated decision-making algorithms at each stage in the processing of the geometric model. Hence, the thesis is focussed upon the design, development and implementation of such algorithms. Various techniques for geometric modelling are briefly examined and then details are given of the modelling package that was developed for this project, The principles of orthographic projection and dimensioning are treated and some published work on the theory of dimensioning is examined. A new theoretical approach to dimensioning is presented and discussed. The existing body of knowledge on decision-making is sampled and the author then shows how methods which were originally developed for management decisions may be adapted to serve the purposes of this project. The remainder of the thesis is devoted to reports on the development of decision-making algorithms for orthographic view selection, sectioning and crosshatching, the preparation of orthographic views with essential hidden detail, and two approaches to the actual insertion of dimension lines and text. The thesis concludes that the theories of decision-making can be applied to work of this kind. It may be possible to generate computer solutions that are closer to the optimum than some man-made dimensioning schemes. Further work on important details is required before a commercially acceptable package could be produced.

Fuel supplier selection for large scale UK bioenergy schemes

This article presents a potential method to assist developers of future bioenergy schemes when selecting from available suppliers of biomass materials. The method aims to allow tacit requirements made on biomass suppliers to be considered at the design stage of new developments. The method used is a combination of the Analytical Hierarchy Process and the Quality Function Deployment methods (AHP-QFD). The output of the method is a ranking and relative weighting of the available suppliers which could be used to improve optimization algorithms such as linear and goal programming. The paper is at a conceptual stage and no results have been obtained. The aim is to use the AHP-QFD method to bridge the gap between treatment of explicit and tacit requirements of bioenergy schemes; allowing decision makers to identify the most successful supply strategy available.

Combining algorithms to predict bacterial protein sub-cellular location:parallel versus concurrent implementations

We describe a novel and potentially important tool for candidate subunit vaccine selection through in silico reverse-vaccinology. A set of Bayesian networks able to make individual predictions for specific subcellular locations is implemented in three pipelines with different architectures: a parallel implementation with a confidence level-based decision engine and two serial implementations with a hierarchical decision structure, one initially rooted by prediction between membrane types and another rooted by soluble versus membrane prediction. The parallel pipeline outperformed the serial pipeline, but took twice as long to execute. The soluble-rooted serial pipeline outperformed the membrane-rooted predictor. Assessment using genomic test sets was more equivocal, as many more predictions are made by the parallel pipeline, yet the serial pipeline identifies 22 more of the 74 proteins of known location.

Performance evaluation of adaptive context aware rate selection algorithm (ACARS) for road safety applications in vehicular network

One of the major drawbacks for mobile nodes in wireless networks is power management. Our goal is to evaluate the performance power control scheme to be used to reduce network congestion, improve quality of service and collision avoidance in vehicular network and road safety application. Some of the importance of power control (PC) are improving spatial reuse, and increasing network capacity in mobile wireless communications. In this simulation we have evaluated the performance of existing rate algorithms compared with context Aware Rate selection algorithm (ACARS) and also seen the performance of ACARS and how it can be applied to road safety, improve network control and power management. Result shows that ACARS is able to minimize the total transmit power in the presence of propagation processes and mobility of vehicles, by adapting to the fast varying channels conditions with the Path loss exponent values that was used for that environment which is shown in the network simulation parameter. Our results have shown that ACARS is a very robust algorithm which performs very well with the effect of propagation processes that is prone to every transmitted signal in mobile networks. © 2013 IEEE.