Abstract—When dealing with decentralized estimation, it is important to reduce the cost of communicating the distributed observations—a problem receiving revived interest in the context of wireless sensor networks. In this paper, we derive and analyze distributed state estimators of dynamical stochastic processes, whereby the low communication cost is effected by requiring the transmission of a single bit per observation. Following a Kalman filtering (KF) approach, we develop recursive algorithms for distributed state estimation based on the sign of innovations (SOI). Even though SOI-KF can afford minimal communication overhead, we prove that in terms of performance and complexity it comes very close to the clairvoyant KF which is based on the analog-amplitude observations. Reinforcing our conclusions, we show that the SOI-KF applied to distributed target tracking based on distance-only observations yields accurate estimates at low communication cost. Index Terms—Distributed state estimation, Kalman filter (KF), target tracking, wireless sensor networks.

Having accurate left ventricle (LV) segmentations across a cardiac cycle provides useful quantitative (e.g. ejection fraction) and qualitative information for diagnosis of certain heart conditions. Existing LV segmentation techniques are founded mostly upon algorithms for segmenting static images. In order to exploit the dynamic structure of the heart in a principled manner, we approach the problem of LV segmentation as a recursive estimation problem. In our framework, LV boundaries constitute the dynamic system state to be estimated, and a sequence of observed cardiac images constitute the data. By formulating the problem as one of state estimation, the segmentation at each particular time is based not only on the data observed at that instant, but also on predictions based on past segmentations. This requires a dynamical system model of the LV, which we propose to learn from training data through an information-theoretic approach. To incorporate the learned dynamic model into our segmentation framework and obtain predictions, we use ideas from particle filtering. Our framework uses a curve evolution method to combine such predictions with the observed images to estimate the LV boundaries at each time. We demonstrate the effectiveness of the proposed approach on a large set of cardiac images. We observe that our approach provides more accurate segmentations than those from static image segmentation techniques, especially when the observed data are of limited quality.

Abstract — In nonlinear Bayesian estimation it is generally inevitable to incorporate approximate descriptions of the exact estimation algorithm. There are two possible ways to involve approximations: Approximating the nonlinear stochastic system model or approximating the prior probability density function. The key idea of the introduced novel estimator called Hybrid Density Filter relies on approximating the nonlinear system, thus approximating conditional densities. These densities nonlinearly relate the current system state to the future system state at predictions or to potential measurements at measurement updates. A hybrid density consisting of both Dirac delta functions and Gaussian densities is used for an optimal approximation. This paper addresses the optimization problem for treating the conditional density approximation. Furthermore, efficient estimation algorithms are derived based upon the special structure of the hybrid density, which yield a Gaussian mixture representation of the system state’s density.

Abstract not found

Abstract not found

Abstract. To date, the neural decoding of time-evolving physical state – for example, the path of a foraging rat or arm movements – has been largely carried out using linear trajectory models, primarily due to their computational efficiency. The possibility of better capturing the statistics of the movements using nonlinear trajectory models, thereby yielding more accurate decoded trajectories, is enticing. However, nonlinear decoding usually carries a higher computational cost, which is an important consideration in real-time settings. In this paper, we present techniques for nonlinear decoding employing modal Gaussian approximations, expectatation propagation, and Gaussian quadrature. We compare their decoding accuracy versus computation time tradeoffs based on high-dimensional simulated neural spike counts. Key words: Nonlinear dynamical models, nonlinear state estimation, neural decoding, neural prosthetics, expectation-propagation, Gaussian quadrature

Abstract — We present a general probabilistic perspective on Gaussian filtering and smoothing. This allows us to show that common approaches to Gaussian filtering/smoothing can be distinguished solely by their methods of computing/approximating the means and covariances of joint probabilities. This implies that novel filters and smoothers can be derived straightforwardly by providing methods for computing these moments. Based on this insight, we derive the cubature Kalman smoother and propose a novel robust filtering and smoothing algorithm based on Gibbs sampling. I.

For several tasks in sensor networks, such as localization, information fusion, or sensor scheduling, Bayesian estimation is of paramount importance. Due to the limited computational and memory resources of the nodes in a sensor network, evaluation of the prediction step of the Bayesian estimator has to be performed very efficiently. An exact and closed-form representation of the predicted probability density function of the system state is typically impossible to obtain, since exactly solving the prediction step for nonlinear discrete-time dynamic systems in closed form is unfeasible. Assuming additive noise, we propose an accurate approximation of the predicted density, that can be calculated efficiently by optimally approximating the transition density using a hybrid density. A hybrid density consists of two different density types: Dirac delta functions that cover the domain of the current density of the system state, and another density type, e.g. Gaussian densities, that cover the domain of the predicted density. The freely selectable, second density type of the hybrid density depends strongly on the noise affecting the nonlinear system. So, the proposed approximation framework for nonlinear prediction is not restricted to a specific noise density. It further allows an analytical evaluation of the Chapman-Kolmogorov prediction equation and can be interpreted as a deterministic sampling estimation approach. In contrast to methods using random sampling like particle filters, a dramatic reduction in the number of components and a subsequent decrease in computation time for approximating the predicted density is gained.

Abstract – Nonlinear fusion of multi-dimensional densities is an important application in Bayesian state estimation. In the approach proposed here, a joint density over all considered densities is build, which is then approximated by means of a Dirac mixture density by partitioning the joint state space into regions that are represented by single Dirac components. This approximation procedure depends on the nonlinear fusion model and only areas relevant to this model are considered. The processing in joint state space has advantages, especially when fusing Dirac mixture densities. Within this approach, degeneration can be avoided and even densities without mutual support can be combined. Thus, this approach gives an alternative to multiplication of Dirac mixtures with a likelihood, as used in the particle filter. Furthermore, a nonlinear Bayesian estimator with filter and prediction step can be formulated, which is able to cope with both discrete and continuous densities.

Abstract—This paper presents a reconfigurable particle filter design methodology for a real-time bearings-only tracking application. The methodology provides the capability of selecting a single particle filter from multiple particle filter realizations with maximum resource sharing. The autonomous buffer controller mechanism for the architecture ensures correct operation of the particle filters. Parameter adaptation and algorithm reconfiguration can be accomplished with negligible reconfiguration overhead through buffer controllers and a set of switches for transforming dataflow structures such that any desired particle filter can be implemented. Two target particle filters, sample importance resample filter (SIRF) and Gaussian particle filter (GPF), are realized using field programmable gate array (FPGA) based on the proposed methodology. However, the architecture can be extended for a wide range of particle filters with different sets of dynamics. This paper successfully demonstrates that implementation of a domain specific processor for particle filters is feasible with performance that is much higher than that of commercially available digital signal processors (DSPs). Index Terms—Buffer controller, field programmable gate array (FPGA) design methodology, particle filter, reconfigurable design. I.