The performance of GPS is strongly degraded in a multipath environment. The multipath impact the distribution of the additive noise corrupting the distance measurements between the satellites and the GPS receiver. In this paper, this distribution is assumed unknown and modeled in a flexible way by using the Bayesian non parametric framework and more precisely the Dirichlet process mixtures. Nevertheless, these latter depend on the so-called scale parameter which can be difficult to tune a priori. The originality of our approach consists in adapting a recent version of the online EM algorithm, developed by Cappé for hidden Markov models, to compute a maximum a posteriori estimate of the scale parameter. Then, as the proposed model is non linear and non Gaussian, the EM-based scale parameter estimation is coupled with a Rao-Blackwellized particle filter for the joint estimation of the mobile location and the distance measurement noise distribution.

A nonparametric approach combining generative models and functional data analysis is presented in this paper for classifying functional data which arise naturally in a wide variety of signal processing applications, such as brain computer interfacing, speech recognition, or image classification. Based on a new and improved family of Bayesian classifiers, we extend hierarchical Bayesian classification methodology from vector to functional settings. We provide theoretical and practical motivations to our approach which relies on Dirichlet process mixtures and Gaussian processes. The performance is evaluated on phoneme recognition task, and compared to that of Functional Support Vector Machines (FSVMs).

The Dirichlet Process Mixture (DPM) models represent an attractive approach to modeling latent distributions parametrically. In DPM models the Dirichlet process (DP) is applied especially when the distribution of latent parameters is to be considered as multimodal. DPMs allow for uncertainty in the choice of parametric forms and in the number of mixing components (clusters). The parameters of a DP include the precision a and the base probability measure G0(μ, Σ). In most applications, the choice of priors and posteriors computation for the hyperparameters (α, μ, Σ) clearly influences inferences about the level of clustering in the mixture. This is the main focus of this paper. We consider the problem of density estimation of an observation noise distribution in a dynamic nonlinear model from a Bayesian nonparametric view-point. Our approach is illustrated in a real-world data analysis task dealing with the estimation of pseudorange errors in a GNSS based localization context.

This paper presents a method aimed at recognizing environmental sounds for surveillance and security applications. We propose to apply one-class support vector machines (1-SVMs) together with a sophisticated dissimilarity measure in order to address audio classification, and more specifically, sound recognition. We illustrate the performance of this method on an audio database, which consists of 1015 sounds belonging to nine classes. The database used presents high intraclass diversity in temps of signal properties and some kind of interclass similarities. A large discrepancy in the number of items in each class implies nonuniform probability of sound appearances. The method proceeds as follows: first, the use of a set of state-of-the-art audio features is studied. Then, we introduce a set of novel features obtained by combining elementary features. Experiments conducted on a nine-class classification problem show the superiority of this novel sound recognition method. The best recognition accuracy (96.89%) is obtained when combining wavelet-based features, MFCCs, and individual temporal and frequency features. Our 1-SVM-based multiclass classification approach overperforms the conventional hidden Markov model-based system in the experiments conducted, the improvement in the error rate can reach 50%. Besides, we provide empirical results showing that the single-class SVM outperforms a combination of binary SVMs. Additional experiments demonstrate our method is robust to environmental noise.

This paper proposes to apply optimized one-class support vector machines (1-SVMs) as a discriminative framework in order to address a specific audio classification problem. First, since SVM-based classifier with gaussian RBF kernel is sensitive to the kernel width, the width will be scaled in a distribution-dependent way permitting to avoid under-fitting and over-fitting problems. Moreover, an advanced dissimilarity measure will be introduced. We illustrate the performance of these methods on an audio database containing environmental sounds that may be of great importance for surveillance and security applications. The experiments conducted on a multi-class problem show that by choosing adequately the SVM parameters, we can efficiently address a sounds classification problem characterized by complex real-world datasets.