Página 1 dos resultados de 456 itens digitais encontrados em 0.015 segundos

Novelty detection and segmentation based on Gaussian mixture models: a case study in 3D robotic laser mapping

Drews-Jr., Paulo; Núñez, Pedro; Rocha, Rui P.; Campos, Mario; Dias, Jorge
Fonte: Elsevier Publicador: Elsevier
Tipo: Artigo de Revista Científica
ENG
Relevância na Pesquisa
86.04%
This article proposes a framework to detect and segment changes in robotics datasets, using 3D robotic mapping as a case study. The problem is very relevant in several application domains, not necessarily related with mobile robotics, including security, health, industry and military applications. The aim is to identify significant changes by comparing current data with previous data provided by sensors. This feature is extremely challenging because large amounts of noisy data must be processed in a feasible way. The proposed framework deals with novelty detection and segmentation in robotic maps using clusters provided by Gaussian Mixture Models (GMMs). GMMs provides a feature space that enables data compression and effective processing. Two alternative criteria to detect changes in the GMM space are compared: a greedy technique based on the Earth Mover’s Distance (EMD); and a structural matching algorithm that fulfills both absolute (global matching) and relative constraints (structural matching). The proposed framework is evaluated with real robotic datasets and compared with other methods known from literature. With this purpose, 3D mapping experiments are carried out with both simulated data and real data from a mobile robot equipped with a 3D range sensor.

Representações hierárquicas de vocábulos de línguas indígenas brasileiras: modelos baseados em mistura de Gaussianas; Hierarchical representations of words of brazilian indigenous languages: models based on Gaussian mixture

Sepúlveda Torres, Lianet
Fonte: Biblioteca Digitais de Teses e Dissertações da USP Publicador: Biblioteca Digitais de Teses e Dissertações da USP
Tipo: Dissertação de Mestrado Formato: application/pdf
Publicado em 08/12/2010 PT
Relevância na Pesquisa
86.12%
Apesar da ampla diversidade de línguas indígenas no Brasil, poucas pesquisas estudam estas línguas e suas relações. Inúmeros esforços têm sido dedicados a procurar similaridades entre as palavras das línguas indígenas e classificá-las em famílias de línguas. Seguindo a classificação mais aceita das línguas indígenas do Brasil, esta pesquisa propõe comparar palavras de 10 línguas indígenas brasileiras. Para isso, considera-se que estas palavras são sinais de fala e estima-se a função de distribuição de probabilidade (PDF) de cada palavra, usando um modelo de mistura de gaussianas (GMM). A PDF foi considerada um modelo para representar as palavras. Os modelos foram comparados utilizando medidas de distância para construir estruturas hierárquicas que evidenciaram possíveis relações entre as palavras. Seguindo esta linha, a hipótese levantada nesta pesquisa é que as PDFs baseadas em GMM conseguem caracterizar as palavras das línguas indígenas, permitindo o emprego de medidas de distância entre elas para estabelecer relações entre as palavras, de forma que tais relações confirmem algumas das classificações. Os parâmetros do GMM foram calculados utilizando o algoritmo Maximização da Expectância (em inglês...

A connectionist approach for incremental function approximation and on-line tasks; Uma abordagem conexionista para a aproximação incremental de funções e tarefas de tempo real

Heinen, Milton Roberto
Fonte: Universidade Federal do Rio Grande do Sul Publicador: Universidade Federal do Rio Grande do Sul
Tipo: Tese de Doutorado Formato: application/pdf
ENG
Relevância na Pesquisa
66.16%
Este trabalho propõe uma nova abordagem conexionista, chamada de IGMN (do inglês Incremental Gaussian Mixture Network), para aproximação incremental de funções e tarefas de tempo real. Ela é inspirada em recentes teorias do cérebro, especialmente o MPF (do inglês Memory-Prediction Framework) e a Inteligência Artificial Construtivista, que fazem com que o modelo proposto possua características especiais que não estão presentes na maioria dos modelos de redes neurais existentes. Além disso, IGMN é baseado em sólidos princípios estatísticos (modelos de mistura gaussianos) e assintoticamente converge para a superfície de regressão ótima a medida que os dados de treinamento chegam. As principais vantagens do IGMN em relação a outros modelos de redes neurais são: (i) IGMN aprende instantaneamente analisando cada padrão de treinamento apenas uma vez (cada dado pode ser imediatamente utilizado e descartado); (ii) o modelo proposto produz estimativas razoáveis baseado em poucos dados de treinamento; (iii) IGMN aprende de forma contínua e perpétua a medida que novos dados de treinamento chegam (não existem fases separadas de treinamento e utilização); (iv) o modelo proposto resolve o dilema da estabilidade-plasticidade e não sofre de interferência catastrófica; (v) a topologia da rede neural é definida automaticamente e de forma incremental (novas unidades são adicionadas sempre que necessário); (vi) IGMN não é sensível às condições de inicialização (de fato IGMN não utiliza nenhuma decisão e/ou inicialização aleatória); (vii) a mesma rede neural IGMN pode ser utilizada em problemas diretos e inversos (o fluxo de informações é bidirecional) mesmo em regiões onde a função alvo tem múltiplas soluções; e (viii) IGMN fornece o nível de confiança de suas estimativas. Outra contribuição relevante desta tese é o uso do IGMN em importantes tarefas nas áreas de robótica e aprendizado de máquina...

On the determination of epsilon during discriminative GMM training

Guido, Rodrigo Capobianco; Chen, Shi-Huang; Junior, Sylvio Barbon; Souza, Leonardo Mendes; Vieira, Lucimar Sasso; Rodrigues, Luciene Cavalcanti; Escola, Joao Paulo Lemos; Zulato, Paulo Ricardo Franchi; Lacerda, Michel Alves; Ribeiro, Jussara
Fonte: Universidade Estadual Paulista Publicador: Universidade Estadual Paulista
Tipo: Conferência ou Objeto de Conferência Formato: 362-364
ENG
Relevância na Pesquisa
66.12%
Discriminative training of Gaussian Mixture Models (GMMs) for speech or speaker recognition purposes is usually based on the gradient descent method, in which the iteration step-size, ε, uses to be defined experimentally. In this letter, we derive an equation to adaptively determine ε, by showing that the second-order Newton-Raphson iterative method to find roots of equations is equivalent to the gradient descent algorithm. © 2010 IEEE.

Gaussian mixture models and semantic gating improve reconstructions from human brain activity

Schoenmakers, Sanne; Güçlü, Umut; van Gerven, Marcel; Heskes, Tom
Fonte: Frontiers Media S.A. Publicador: Frontiers Media S.A.
Tipo: Artigo de Revista Científica
Publicado em 30/01/2015 EN
Relevância na Pesquisa
66.04%
Better acquisition protocols and analysis techniques are making it possible to use fMRI to obtain highly detailed visualizations of brain processes. In particular we focus on the reconstruction of natural images from BOLD responses in visual cortex. We expand our linear Gaussian framework for percept decoding with Gaussian mixture models to better represent the prior distribution of natural images. Reconstruction of such images then boils down to probabilistic inference in a hybrid Bayesian network. In our set-up, different mixture components correspond to different character categories. Our framework can automatically infer higher-order semantic categories from lower-level brain areas. Furthermore, the framework can gate semantic information from higher-order brain areas to enforce the correct category during reconstruction. When categorical information is not available, we show that automatically learned clusters in the data give a similar improvement in reconstruction. The hybrid Bayesian network leads to highly accurate reconstructions in both supervised and unsupervised settings.

On Convergence Properties of the EM Algorithm for Gaussian Mixtures

Jordan, Michael; Xu, Lei
Fonte: MIT - Massachusetts Institute of Technology Publicador: MIT - Massachusetts Institute of Technology
Formato: 9 p.; 291671 bytes; 476864 bytes; application/postscript; application/pdf
EN_US
Relevância na Pesquisa
65.99%
"Expectation-Maximization' (EM) algorithm and gradient-based approaches for maximum likelihood learning of finite Gaussian mixtures. We show that the EM step in parameter space is obtained from the gradient via a projection matrix $P$, and we provide an explicit expression for the matrix. We then analyze the convergence of EM in terms of special properties of $P$ and provide new results analyzing the effect that $P$ has on the likelihood surface. Based on these mathematical results, we present a comparative discussion of the advantages and disadvantages of EM and other algorithms for the learning of Gaussian mixture models.

A novel robust scene change detection algorithm for autonomous robots using mixtures of gaussians

Manso, Luis; Núñez, Pedro; Silva, Sidnei da; Drews Junior, Paulo Lilles Jorge
Fonte: Universidade Federal do Rio Grande Publicador: Universidade Federal do Rio Grande
Tipo: Artigo de Revista Científica
ENG
Relevância na Pesquisa
65.99%
Interest in change detection techniques has considerably increased during recent years in the field of autonomous robotics. This is partly because changes in a robot’s working environment are useful for several robotic skills (e.g., spatial cognition, modelling or navigation) and applications (e.g., surveillance or guidance robots). Changes are usually detected by comparing current data provided by the robot’s sensors with a previously known map or model of the environment. When the data consists of a large point cloud, dealing with it is a computationally expensive task, mainly due to the amount of points and the redundancy. Using Gaussian Mixture Models (GMM) instead of raw point clouds leads to a more compact feature space that can be used to efficiently process the input data. This allows us to successfully segment the set of 3D points acquired by the sensor and reduce the computational load of the change detection algorithm. However, the segmentation of the environment as a Mixture of Gaussians has some problems that need to be properly addressed. In this paper, a novel change detection algorithm is described in order to improve the robustness and computational cost of previous approaches. The proposal is based on the classic Expectation Maximization (EM) algorithm...

Novelty detection and segmentation based on gaussian mixture models: a case study in 3D robotic laser mapping

Drews Junior, Paulo Lilles Jorge; Núñez, Pedro; Rocha, Rui Paulo Pinto da; Campos, Mario Fernando Montenegro; Dias, Jorge
Fonte: Universidade Federal do Rio Grande Publicador: Universidade Federal do Rio Grande
Tipo: Artigo de Revista Científica
ENG
Relevância na Pesquisa
86.04%
This article proposes a framework to detect and segment changes in robotics datasets, using 3D robotic mapping as a case study. The problem is very relevant in several application domains, not necessarily related with mobile robotics, including security, health, industry and military applications. The aim is to identify significant changes by comparing current data with previous data provided by sensors. This feature is extremely challenging because large amounts of noisy data must be processed in a feasible way. The proposed framework deals with novelty detection and segmentation in robotic maps using clusters provided by Gaussian Mixture Models (GMMs). GMMs provides a feature space that enables data compression and effective processing. Two alternative criteria to detect changes in the GMM space are compared: a greedy technique based on the Earth Mover’s Distance (EMD); and a structural matching algorithm that fulfills both absolute (global matching) and relative constraints (structural matching). The proposed framework is evaluated with real robotic datasets and compared with other methods known from literature. With this purpose, 3D mapping experiments are carried out with both simulated data and real data from a mobile robot equipped with a 3D range sensor.

Visual learning and recognition of sequential data manifolds with applications to human movement analysis

Wang, L.; Suter, D.
Fonte: Academic Press Inc Publicador: Academic Press Inc
Tipo: Artigo de Revista Científica
Publicado em //2008 EN
Relevância na Pesquisa
65.99%
Human motion analysis is increasingly attracting much attention from computer vision researchers. This paper aims to address the task of human gait and activity analysis from image sequences by learning and recognition of sequential data under a general integrated framework. Human movements generally exhibit intrinsically nonlinear spatiotemporal characteristics in the high-dimensional ambient space. An attractive framework, which we explore here, is to: (1) Extract simple and reliable features from image sequences. (2) Find a low-dimensional feature representation embedded in high-dimensional image data. (3) Then characterize/classify the motions in this low-dimensional feature space. We examine two simple alternatives for step 1: silhouette and a distance transformed silhouette; and three quite different methods for step 3: Gaussian mixture models (GMM) based classification, a matching-based approach with the mean Hausdorff distance, and continuous hidden Markov models (HMM) based modelling and recognition. The core is step 2 where we choose to use LPP (locality preserving projections), an optimal linear approximation to a nonlinear spectral embedding technique (i.e., Laplacian eigenmap). In essence our aim is to see whether this core...

Bivariate models for the analysis of internal nitrogen use efficiency: mixture models as an exploratory tool.

Munoz Santa, Isabel
Fonte: Universidade de Adelaide Publicador: Universidade de Adelaide
Tipo: Tese de Doutorado
Publicado em //2014
Relevância na Pesquisa
66.08%
Ratios are commonly used among plant and soil scientists, in particular to express the plant nutrient utilisation efficiency of macro- and micro-nutrients. The internal nutrient efficiency can be understood in terms of maximising yield per a unit of nutrient in the plant. At present, IEɴ data are usually collected from designed field trials where different treatments are applied (e.g. fertiliser treatments) and analysed by univariate linear mixed models. However, univariate linear models on the ratio do not maintain information on the original traits, including their correlation, which presents a challenge when interpreting the effect of agronomic practices or environmental conditions on the process of nutrient conversion into grain. Moreover, the distributional properties of ratios do not comply with the assumptions of these linear models favoured in the area of soil and plant science research. A more suitable approach is to collect the traits of interest and to use bivariate analyses. These analyses preserve the information on the original traits and avoid issues associated with the ratio distributional properties. If the data comes from field studies, different experimental and environmental conditions may lead to the presence of patterns (groups) in the data in addition or concurrently with designed treatments. Researchers in plant and soil sciences may be interested in identifying those conditions...

Gaussian Mixture Models for Affordance Learning Using Bayesian Networks

Osório, Pedro; Bernardino, Alexandre; Martínez-Cantín, Rubén; Santos-Victor, José
Fonte: IEEE Publicador: IEEE
Tipo: info:eu-repo/semantics/acceptedVersion; info:eu-repo/semantics/conferenceObject Formato: application/pdf
Publicado em /12/2010 ENG
Relevância na Pesquisa
65.97%
Affordances are fundamental descriptors of relationships between actions, objects and effects. They provide the means whereby a robot can predict effects, recognize actions, select objects and plan its behavior according to desired goals. This paper approaches the problem of an embodied agent exploring the world and learning these affordances autonomously from its sensory experiences. Models exist for learning the structure and the parameters of a Bayesian Network encoding this knowledge. Although Bayesian Networks are capable of dealing with uncertainty and redundancy, previous work considered complete observability of the discrete sensory data, which may lead to hard errors in the presence of noise. In this paper we consider a probabilistic representation of the sensors by Gaussian Mixture Models (GMMs) and explicitly taking into account the probability distribution contained in each discrete affordance concept, which can lead to a more correct learning.; European Community's Seventh Framework Program; Proceedings of: IEEE/RSJ 2010 International Conference on Intelligent Robots and Systems (IROS 2010), October 18-22, 2010, Taipe, Taiwan

Bayesian Gaussian Mixture Models for High-Density Genotyping Arrays

Sabatti, Chiara; Lange, Kenneth
Fonte: PubMed Publicador: PubMed
Tipo: Artigo de Revista Científica
Publicado em 01/03/2008 EN
Relevância na Pesquisa
65.99%
Affymetrix's SNP (single-nucleotide polymorphism) genotyping chips have increased the scope and decreased the cost of gene-mapping studies. Because each SNP is queried by multiple DNA probes, the chips present interesting challenges in genotype calling. Traditional clustering methods distinguish the three genotypes of an SNP fairly well given a large enough sample of unrelated individuals or a training sample of known genotypes. This article describes our attempt to improve genotype calling by constructing Gaussian mixture models with empirically derived priors. The priors stabilize parameter estimation and borrow information collectively gathered on tens of thousands of SNPs. When data from related family members are available, our models capture the correlations in signals between relatives. With these advantages in mind, we apply the models to Affymetrix probe intensity data on 10,000 SNPs gathered on 63 genotyped individuals spread over eight pedigrees. We integrate the genotype-calling model with pedigree analysis and examine a sequence of symmetry hypotheses involving the correlated probe signals. The symmetry hypotheses raise novel mathematical issues of parameterization. Using the Bayesian information criterion, we select the best combination of symmetry assumptions. Compared to Affymetrix's software...

Automatic speaker recognition with Multi-resolution Gaussian Mixture models (MR-GMMs)

D’Almeida, Frederico Quadros; Nascimento, Francisco Assis de Oliveira; Berger, Pedro de Azevedo; Silva, Lúcio Martins da
Fonte: Brazilian Association of High Technology Experts (ABEAT) Publicador: Brazilian Association of High Technology Experts (ABEAT)
Tipo: Artigo de Revista Científica
ENG
Relevância na Pesquisa
65.97%
Gaussian Mixture Models (GMMs) are the most widely used technique for voice modeling in automatic speaker recognition systems. In this paper, we introduce a variation of the traditional GMM approach that uses models with variable complexity (resolution). Termed Multi-resolution GMMs (MR-GMMs); this new approach yields more than a 50% reduction in the computational costs associated with proper speaker identification, as compared to the traditional GMM approach. We also explore the noise robustness of the new method by investigating MR-GMM performance under noisy audio conditions using a series of practical identification tests.

On the characterization of flowering curves using Gaussian mixture models

Proïa, Frédéric; Pernet, Alix; Thouroude, Tatiana; Michel, Gilles; Clotault, Jérémy
Fonte: Universidade Cornell Publicador: Universidade Cornell
Tipo: Artigo de Revista Científica
Publicado em 18/10/2015
Relevância na Pesquisa
66.07%
In this paper, we develop a statistical methodology applied to the characterization of flowering curves using Gaussian mixture models. Our study relies on a set of rosebushes flowering data, and Gaussian mixture models are mainly used to quantify the reblooming behavior of each one. In this regard, we also suggest our own selection criterion to take into account the lack of symmetry of most of the flowering curves. Three classes are created on the basis of the reblooming indicators, and a subclassification is made using a longitudinal $k$--means algorithm which highlights the role also played by the precocity of the flowering. A principal component analysis is finally conducted on a set of indicators derived from our statistical approach to get an overview of the correlations between the features that we have decided to retain on each curve. Results suggest the lack of correlation between reblooming and flowering precocity. The pertinent indicators obtained in this study will be a first step towards the comprehension of the environmental and genetic control of these biological processes.; Comment: 26 pages, 25 figures, 1 table

Model Selection for Gaussian Mixture Models

Huang, Tao; Peng, Heng; Zhang, Kun
Fonte: Universidade Cornell Publicador: Universidade Cornell
Tipo: Artigo de Revista Científica
Publicado em 15/01/2013
Relevância na Pesquisa
66.04%
This paper is concerned with an important issue in finite mixture modelling, the selection of the number of mixing components. We propose a new penalized likelihood method for model selection of finite multivariate Gaussian mixture models. The proposed method is shown to be statistically consistent in determining of the number of components. A modified EM algorithm is developed to simultaneously select the number of components and to estimate the mixing weights, i.e. the mixing probabilities, and unknown parameters of Gaussian distributions. Simulations and a real data analysis are presented to illustrate the performance of the proposed method.

Simple Methods for Initializing the EM Algorithm for Gaussian Mixture Models

Blömer, Johannes; Bujna, Kathrin
Fonte: Universidade Cornell Publicador: Universidade Cornell
Tipo: Artigo de Revista Científica
Publicado em 20/12/2013
Relevância na Pesquisa
66.07%
In this paper, we consider simple and fast approaches to initialize the Expectation-Maximization algorithm (EM) for multivariate Gaussian mixture models. We present new initialization methods based on the well-known $K$-means++ algorithm and the Gonzalez algorithm. These methods close the gap between simple uniform initialization techniques and complex methods, that have been specifically designed for Gaussian mixture models and depend on the right choice of hyperparameters. In our evaluation we compare our methods with a commonly used random initialization method, an approach based on agglomerative hierarchical clustering, and a known, plain adaption of the Gonzalez algorithm. Our results indicate that algorithms based on $K$-means++ outperform the other methods.

Hypothesis Testing for Parsimonious Gaussian Mixture Models

Punzo, Antonio; Browne, Ryan P.; McNicholas, Paul D.
Fonte: Universidade Cornell Publicador: Universidade Cornell
Tipo: Artigo de Revista Científica
Publicado em 02/05/2014
Relevância na Pesquisa
66.18%
Gaussian mixture models with eigen-decomposed covariance structures make up the most popular family of mixture models for clustering and classification, i.e., the Gaussian parsimonious clustering models (GPCM). Although the GPCM family has been used for almost 20 years, selecting the best member of the family in a given situation remains a troublesome problem. Likelihood ratio tests are developed to tackle this problems. These likelihood ratio tests use the heteroscedastic model under the alternative hypothesis but provide much more flexibility and real-world applicability than previous approaches that compare the homoscedastic Gaussian mixture versus the heteroscedastic one. Along the way, a novel maximum likelihood estimation procedure is developed for two members of the GPCM family. Simulations show that the $\chi^2$ reference distribution gives reasonable approximation for the LR statistics only when the sample size is considerable and when the mixture components are well separated; accordingly, following Lo (2008), a parametric bootstrap is adopted. Furthermore, by generalizing the idea of Greselin and Punzo (2013) to the clustering context, a closed testing procedure, having the defined likelihood ratio tests as local tests, is introduced to assess a unique model in the general family. The advantages of this likelihood ratio testing procedure are illustrated via an application to the well-known Iris data set.

Tractable Measure of Component Overlap for Gaussian Mixture Models

Nowakowska, Ewa; Koronacki, Jacek; Lipovetsky, Stan
Fonte: Universidade Cornell Publicador: Universidade Cornell
Tipo: Artigo de Revista Científica
Publicado em 26/07/2014
Relevância na Pesquisa
66.11%
The ability to quantify distinctness of a cluster structure is fundamental for certain simulation studies, in particular for those comparing performance of different classification algorithms. The intrinsic integral measure based on the overlap of corresponding mixture components is often analytically intractable. This is also the case for Gaussian mixture models with unequal covariance matrices when space dimension $d > 1$. In this work we focus on Gaussian mixture models and at the sample level we assume the class assignments to be known. We derive a measure of component overlap based on eigenvalues of a generalized eigenproblem that represents Fisher's discriminant task. We explain rationale behind it and present simulation results that show how well it can reflect the behavior of the integral measure in its linear approximation. The analyzed coefficient possesses the advantage of being analytically tractable and numerically computable even in complex setups.

Statistical Compressive Sensing of Gaussian Mixture Models

Yu, Guoshen; Sapiro, Guillermo
Fonte: Universidade Cornell Publicador: Universidade Cornell
Tipo: Artigo de Revista Científica
Publicado em 20/10/2010
Relevância na Pesquisa
66.08%
A new framework of compressive sensing (CS), namely statistical compressive sensing (SCS), that aims at efficiently sampling a collection of signals that follow a statistical distribution and achieving accurate reconstruction on average, is introduced. For signals following a Gaussian distribution, with Gaussian or Bernoulli sensing matrices of O(k) measurements, considerably smaller than the O(k log(N/k)) required by conventional CS, where N is the signal dimension, and with an optimal decoder implemented with linear filtering, significantly faster than the pursuit decoders applied in conventional CS, the error of SCS is shown tightly upper bounded by a constant times the k-best term approximation error, with overwhelming probability. The failure probability is also significantly smaller than that of conventional CS. Stronger yet simpler results further show that for any sensing matrix, the error of Gaussian SCS is upper bounded by a constant times the k-best term approximation with probability one, and the bound constant can be efficiently calculated. For signals following Gaussian mixture models, SCS with a piecewise linear decoder is introduced and shown to produce for real images better results than conventional CS based on sparse models.

Statistical Compressed Sensing of Gaussian Mixture Models

Yu, Guoshen; Sapiro, Guillermo
Fonte: Universidade Cornell Publicador: Universidade Cornell
Tipo: Artigo de Revista Científica
Publicado em 30/01/2011
Relevância na Pesquisa
66.15%
A novel framework of compressed sensing, namely statistical compressed sensing (SCS), that aims at efficiently sampling a collection of signals that follow a statistical distribution, and achieving accurate reconstruction on average, is introduced. SCS based on Gaussian models is investigated in depth. For signals that follow a single Gaussian model, with Gaussian or Bernoulli sensing matrices of O(k) measurements, considerably smaller than the O(k log(N/k)) required by conventional CS based on sparse models, where N is the signal dimension, and with an optimal decoder implemented via linear filtering, significantly faster than the pursuit decoders applied in conventional CS, the error of SCS is shown tightly upper bounded by a constant times the best k-term approximation error, with overwhelming probability. The failure probability is also significantly smaller than that of conventional sparsity-oriented CS. Stronger yet simpler results further show that for any sensing matrix, the error of Gaussian SCS is upper bounded by a constant times the best k-term approximation with probability one, and the bound constant can be efficiently calculated. For Gaussian mixture models (GMMs), that assume multiple Gaussian distributions and that each signal follows one of them with an unknown index...