Desarrollo de un sistema de reconocimiento facial

El objetivo principal alrededor del cual se desenvuelve este proyecto es el desarrollo de un sistema de reconocimiento facial. Entre sus objetivos específicos se encuentran: realizar una primera aproximación sobre las técnicas de reconocimiento facial existentes en la actualidad, elegir una aplicación donde pueda ser útil el reconocimiento facial, diseñar y desarrollar un programa en MATLAB que lleve a cabo la función de reconocimiento facial, y evaluar el funcionamiento del sistema desarrollado. Este documento se encuentra dividido en cuatro partes: INTRODUCCIÓN, MARCO TEÓRICO, IMPLEMENTACIÓN, y RESULTADOS, CONCLUSIONES Y LÍNEAS FUTURAS. En la primera parte, se hace una introducción relativa a la actualidad del reconocimiento facial y se comenta brevemente sobre las técnicas existentes para desarrollar un sistema biométrico de este tipo. En ella se justifican también aquellas técnicas que acabaron formando parte de la implementación. En la segunda parte, el marco teórico, se explica la estructura general que tiene un sistema de reconocimiento biométrico, así como sus modos de funcionamiento, y las tasas de error utilizadas para evaluar y comparar su rendimiento. Así mismo, se lleva a cabo una descripción más profunda sobre los conceptos y métodos utilizados para efectuar la detección y reconocimiento facial en la tercera parte del proyecto. La tercera parte abarca una descripción detallada de la solución propuesta. En ella se explica el diseño, características y aplicación de la implementación; que trata de un programa elaborado en MATLAB con interfaz gráfica, y que utiliza cuatro sistemas de reconocimiento facial, basados cada uno en diferentes técnicas: Análisis por componentes principales, análisis lineal discriminante, wavelets de Gabor, y emparejamiento de grafos elásticos. El programa ofrece además la capacidad de crear y editar una propia base de datos con etiquetas, dándole aplicación directa sobre el tema que se trata. Se proponen además una serie de características con el objetivo de ampliar y mejorar las funcionalidades del programa diseñado. Dentro de dichas características destaca la propuesta de un modo de verificación híbrido aplicable a cualquier rama de la biometría y un programa de evaluación capaz de medir, graficar, y comparar las configuraciones de cada uno de los sistemas de reconocimiento implementados. Otra característica destacable es la herramienta programada para la creación de grafos personalizados y generación de modelos, aplicable a reconocimiento de objetos en general. En la cuarta y última parte, se presentan al principio los resultados obtenidos. En ellos se contemplan y analizan las comparaciones entre las distintas configuraciones de los sistemas de reconocimiento implementados para diferentes bases de datos (una de ellas formada con imágenes con condiciones de adquisición no controladas). También se miden las tasas de error del modo de verificación híbrido propuesto. Finalmente, se extraen conclusiones, y se proponen líneas futuras de investigación. ABSTRACT The main goal of this project is to develop a facial recognition system. To meet this end, it was necessary to accomplish a series of specific objectives, which were: researching on the existing face recognition technics nowadays, choosing an application where face recognition might be useful, design and develop a face recognition system using MATLAB, and measure the performance of the implemented system. This document is divided into four parts: INTRODUCTION, THEORTICAL FRAMEWORK, IMPLEMENTATION, and RESULTS, CONCLUSSIONS AND FUTURE RESEARCH STUDIES. In the first part, an introduction is made in relation to facial recognition nowadays, and the techniques used to develop a biometric system of this kind. Furthermore, the techniques chosen to be part of the implementation are justified. In the second part, the general structure and the two basic modes of a biometric system are explained. The error rates used to evaluate and compare the performance of a biometric system are explained as well. Moreover, a description of the concepts and methods used to detect and recognize faces in the third part is made. The design, characteristics, and applications of the systems put into practice are explained in the third part. The implementation consists in developing a program with graphical user interface made in MATLAB. This program uses four face recognition systems, each of them based on a different technique: Principal Component Analysis (PCA), Fisher’s Linear Discriminant (FLD), Gabor wavelets, and Elastic Graph Matching (EGM). In addition, with this implementation it is possible to create and edit one´s tagged database, giving it a direct application. Also, a group of characteristics are proposed to enhance the functionalities of the program designed. Among these characteristics, three of them should be emphasized in this summary: A proposal of an hybrid verification mode of a biometric system; and an evaluation program capable of measuring, plotting curves, and comparing different configurations of each implemented recognition system; and a tool programmed to create personalized graphs and models (tagged graph associated to an image of a person), which can be used generally in object recognition. In the fourth and last part of the project, the results of the comparisons between different configurations of the systems implemented are shown for three databases (One of them created with pictures taken under non-controlled environments). The error rates of the proposed hybrid verification mode are measured as well. Finally, conclusions are extracted and future research studies are proposed.

A connectionist computational method for face recognition

In this work, a modified version of the elastic bunch graph matching (EBGM) algorithm for face recognition is introduced. First, faces are detected by using a fuzzy skin detector based on the RGB color space. Then, the fiducial points for the facial graph are extracted automatically by adjusting a grid of points to the result of an edge detector. After that, the position of the nodes, their relation with their neighbors and their Gabor jets are calculated in order to obtain the feature vector defining each face. A self-organizing map (SOM) framework is shown afterwards. Thus, the calculation of the winning neuron and the recognition process are performed by using a similarity function that takes into account both the geometric and texture information of the facial graph. The set of experiments carried out for our SOM-EBGM method shows the accuracy of our proposal when compared with other state-of the-art methods.

The average mixing matrix signature

Laplacian-based descriptors, such as the Heat Kernel Signature and the Wave Kernel Signature, allow one to embed the vertices of a graph onto a vectorial space, and have been successfully used to find the optimal matching between a pair of input graphs. While the HKS uses a heat di↵usion process to probe the local structure of a graph, the WKS attempts to do the same through wave propagation. In this paper, we propose an alternative structural descriptor that is based on continuoustime quantum walks. More specifically, we characterise the structure of a graph using its average mixing matrix. The average mixing matrix is a doubly-stochastic matrix that encodes the time-averaged behaviour of a continuous-time quantum walk on the graph. We propose to use the rows of the average mixing matrix for increasing stopping times to develop a novel signature, the Average Mixing Matrix Signature (AMMS). We perform an extensive range of experiments and we show that the proposed signature is robust under structural perturbations of the original graphs and it outperforms both the HKS and WKS when used as a node descriptor in a graph matching task.

A people-to-people matching system using graph mining techniques

A people-to-people matching system (or a match-making system) refers to a system in which users join with the objective of meeting other users with the common need. Some real-world examples of these systems are employer-employee (in job search networks), mentor-student (in university social networks), consume-to-consumer (in marketplaces) and male-female (in an online dating network). The network underlying in these systems consists of two groups of users, and the relationships between users need to be captured for developing an efficient match-making system. Most of the existing studies utilize information either about each of the users in isolation or their interaction separately, and develop recommender systems using the one form of information only. It is imperative to understand the linkages among the users in the network and use them in developing a match-making system. This study utilizes several social network analysis methods such as graph theory, small world phenomenon, centrality analysis, density analysis to gain insight into the entities and their relationships present in this network. This paper also proposes a new type of graph called “attributed bipartite graph”. By using these analyses and the proposed type of graph, an efficient hybrid recommender system is developed which generates recommendation for new users as well as shows improvement in accuracy over the baseline methods.

A stable graph-based representation for object recognition through high-order matching

Many Object recognition techniques perform some flavour of point pattern matching between a model and a scene. Such points are usually selected through a feature detection algorithm that is robust to a class of image transformations and a suitable descriptor is computed over them in order to get a reliable matching. Moreover, some approaches take an additional step by casting the correspondence problem into a matching between graphs defined over feature points. The motivation is that the relational model would add more discriminative power, however the overall effectiveness strongly depends on the ability to build a graph that is stable with respect to both changes in the object appearance and spatial distribution of interest points. In fact, widely used graph-based representations, have shown to suffer some limitations, especially with respect to changes in the Euclidean organization of the feature points. In this paper we introduce a technique to build relational structures over corner points that does not depend on the spatial distribution of the features. © 2012 ICPR Org Committee.

CAT-GRAPH+ : towards odometry-driven place consolidation in changing environments

Changing environments present a number of challenges to mobile robots, one of the most significant being mapping and localisation. This problem is particularly significant in vision-based systems where illumination and weather changes can cause feature-based techniques to fail. In many applications only sections of an environment undergo extreme perceptual change. Some range-based sensor mapping approaches exploit this property by combining occasional place recognition with the assumption that odometry is accurate over short periods of time. In this paper, we develop this idea in the visual domain, by using occasional vision-driven loop closures to infer loop closures in nearby locations where visual recognition is difficult due to extreme change. We demonstrate successful map creation in an environment in which change is significant but constrained to one area, where both the vanilla CAT-Graph and a Sum of Absolute Differences matcher fails, use the described techniques to link dissimilar images from matching locations, and test the robustness of the system against false inferences.

Automated topometric graph generation from floor plan analysis

The world is rich with information such as signage and maps to assist humans to navigate. We present a method to extract topological spatial information from a generic bitmap floor plan and build a topometric graph that can be used by a mobile robot for tasks such as path planning and guided exploration. The algorithm first detects and extracts text in an image of the floor plan. Using the locations of the extracted text, flood fill is used to find the rooms and hallways. Doors are found by matching SURF features and these form the connections between rooms, which are the edges of the topological graph. Our system is able to automatically detect doors and differentiate between hallways and rooms, which is important for effective navigation. We show that our method can extract a topometric graph from a floor plan and is robust against ambiguous cases most commonly seen in floor plans including elevators and stairwells.

Exact bounds for distributed graph colouring

A distributed system is a collection of networked autonomous processing units which must work in a cooperative manner. Currently, large-scale distributed systems, such as various telecommunication and computer networks, are abundant and used in a multitude of tasks. The field of distributed computing studies what can be computed efficiently in such systems. Distributed systems are usually modelled as graphs where nodes represent the processors and edges denote communication links between processors. This thesis concentrates on the computational complexity of the distributed graph colouring problem. The objective of the graph colouring problem is to assign a colour to each node in such a way that no two nodes connected by an edge share the same colour. In particular, it is often desirable to use only a small number of colours. This task is a fundamental symmetry-breaking primitive in various distributed algorithms. A graph that has been coloured in this manner using at most k different colours is said to be k-coloured. This work examines the synchronous message-passing model of distributed computation: every node runs the same algorithm, and the system operates in discrete synchronous communication rounds. During each round, a node can communicate with its neighbours and perform local computation. In this model, the time complexity of a problem is the number of synchronous communication rounds required to solve the problem. It is known that 3-colouring any k-coloured directed cycle requires at least ½(log* k - 3) communication rounds and is possible in ½(log* k + 7) communication rounds for all k ≥ 3. This work shows that for any k ≥ 3, colouring a k-coloured directed cycle with at most three colours is possible in ½(log* k + 3) rounds. In contrast, it is also shown that for some values of k, colouring a directed cycle with at most three colours requires at least ½(log* k + 1) communication rounds. Furthermore, in the case of directed rooted trees, reducing a k-colouring into a 3-colouring requires at least log* k + 1 rounds for some k and possible in log* k + 3 rounds for all k ≥ 3. The new positive and negative results are derived using computational methods, as the existence of distributed colouring algorithms corresponds to the colourability of so-called neighbourhood graphs. The colourability of these graphs is analysed using Boolean satisfiability (SAT) solvers. Finally, this thesis shows that similar methods are applicable in capturing the existence of distributed algorithms for other graph problems, such as the maximal matching problem.

Fixed-orientation equilateral triangle matching of point sets

Given a point set P and a class C of geometric objects, G(C)(P) is a geometric graph with vertex set P such that any two vertices p and q are adjacent if and only if there is some C is an element of C containing both p and q but no other points from P. We study G(del)(P) graphs where del is the class of downward equilateral triangles (i.e., equilateral triangles with one of their sides parallel to the x-axis and the corner opposite to this side below that side). For point sets in general position, these graphs have been shown to be equivalent to half-Theta(6) graphs and TD-Delaunay graphs. The main result in our paper is that for point sets P in general position, G(del)(P) always contains a matching of size at least vertical bar P vertical bar-1/3] and this bound is tight. We also give some structural properties of G(star)(P) graphs, where is the class which contains both upward and downward equilateral triangles. We show that for point sets in general position, the block cut point graph of G(star)(P) is simply a path. Through the equivalence of G(star)(P) graphs with Theta(6) graphs, we also derive that any Theta(6) graph can have at most 5n-11 edges, for point sets in general position. (C) 2013 Elsevier B.V. All rights reserved.

Belief propagation for minimum weight many-to-one matchings in the random complete graph

In a complete bipartite graph with vertex sets of cardinalities n and n', assign random weights from exponential distribution with mean 1, independently to each edge. We show that, as n -> infinity, with n' = n/alpha] for any fixed alpha > 1, the minimum weight of many-to-one matchings converges to a constant (depending on alpha). Many-to-one matching arises as an optimization step in an algorithm for genome sequencing and as a measure of distance between finite sets. We prove that a belief propagation (BP) algorithm converges asymptotically to the optimal solution. We use the objective method of Aldous to prove our results. We build on previous works on minimum weight matching and minimum weight edge cover problems to extend the objective method and to further the applicability of belief propagation to random combinatorial optimization problems.

Divide and Conquer: Efficient Large-Scale Structure from Motion Using Graph Partitioning

Despite significant advances in recent years, structure-from-motion (SfM) pipelines suffer from two important drawbacks. Apart from requiring significant computational power to solve the large-scale computations involved, such pipelines sometimes fail to correctly reconstruct when the accumulated error in incremental reconstruction is large or when the number of 3D to 2D correspondences are insufficient. In this paper we present a novel approach to mitigate the above-mentioned drawbacks. Using an image match graph based on matching features we partition the image data set into smaller sets or components which are reconstructed independently. Following such reconstructions we utilise the available epipolar relationships that connect images across components to correctly align the individual reconstructions in a global frame of reference. This results in both a significant speed up of at least one order of magnitude and also mitigates the problems of reconstruction failures with a marginal loss in accuracy. The effectiveness of our approach is demonstrated on some large-scale real world data sets.

SiGMa: Simple Greedy Matching for Aligning Large Knowledge Bases

The Internet has enabled the creation of a growing number of large-scale knowledge bases in a variety of domains containing complementary information. Tools for automatically aligning these knowledge bases would make it possible to unify many sources of structured knowledge and answer complex queries. However, the efficient alignment of large-scale knowledge bases still poses a considerable challenge. Here, we present Simple Greedy Matching (SiGMa), a simple algorithm for aligning knowledge bases with millions of entities and facts. SiGMa is an iterative propagation algorithm which leverages both the structural information from the relationship graph as well as flexible similarity measures between entity properties in a greedy local search, thus making it scalable. Despite its greedy nature, our experiments indicate that SiGMa can efficiently match some of the world's largest knowledge bases with high precision. We provide additional experiments on benchmark datasets which demonstrate that SiGMa can outperform state-of-the-art approaches both in accuracy and efficiency.

Convex Quadratic Programming Approach to the Maximum Matching Problem

The problem of determining a maximum matching or whether there exists a perfect matching, is very common in a large variety of applications and as been extensively studied in graph theory. In this paper we start to introduce a characterisation of a family of graphs for which its stability number is determined by convex quadratic programming. The main results connected with the recognition of this family of graphs are also introduced. It follows a necessary and sufficient condition which characterise a graph with a perfect matching and an algorithmic strategy, based on the determination of the stability number of line graphs, by convex quadratic programming, applied to the determination of a perfect matching. A numerical example for the recognition of graphs with a perfect matching is described. Finally, the above algorithmic strategy is extended to the determination of a maximum matching of an arbitrary graph and some related results are presented.

An NP-complete matching problem /

"UILU-ENG 79-1713."