Model based learning of sigma points in unscented Kalman filtering

The unscented Kalman filter (UKF) is a widely used method in control and time series applications. The UKF suffers from arbitrary parameters necessary for a step known as sigma point placement, causing it to perform poorly in nonlinear problems. We show how to treat sigma point placement in a UKF as a learning problem in a model based view. We demonstrate that learning to place the sigma points correctly from data can make sigma point collapse much less likely. Learning can result in a significant increase in predictive performance over default settings of the parameters in the UKF and other filters designed to avoid the problems of the UKF, such as the GP-ADF. At the same time, we maintain a lower computational complexity than the other methods. We call our method UKF-L. ©2010 IEEE.

The fish stocks in Uganda aquatic systems: opportunities and challenges for transformation

The status of fish stocks in a water body at any one time is a function of several factors affecting the production of fish in that water body. These include: total number (abundance) and biomass(weight) present, growth (size and age), recruitment (the quantity of fish entering the fishery) including reproduction, mortality which is caused by fishing or natural causes, Other indirect factors of major importance to the status of the stocks include production factors (water quality and availability of natural food for fish), the life history parameters of the different species making up the stocks (e.g. sex ratios, condition of the fish, reproductive potential (i.e. fecundity) etc), Changes in fish stocks do occur when any of the above listed factors directly influence aspects of growth, reproduction and mortality and therefore, numbers and standing stock (biomass). In the exploited fisheries, major research concerns regarding stocks relate to the listed factors especially: estimates of stock abundance/biomass, the quantity of fish being caught,where the fish are caught, which species are caught (relative abundance)when the fish are caught, how the fish are caught. The balance between stock abundance and amount of fish caught provides the basis for intervention. Due to the diverse characteristics of the physical water environment, fishes are in general, not evenly distributed throughout a water body. Shallow and vegetated areas tend to support higher abundance and diversity of fish species. In addition, seasonal variations in fish abundance are so strong that fluctuations in catch have to be expected at fish landings.

Model based learning of sigma points in unscented Kalman filtering

The unscented Kalman filter (UKF) is a widely used method in control and time series applications. The UKF suffers from arbitrary parameters necessary for sigma point placement, potentially causing it to perform poorly in nonlinear problems. We show how to treat sigma point placement in a UKF as a learning problem in a model based view. We demonstrate that learning to place the sigma points correctly from data can make sigma point collapse much less likely. Learning can result in a significant increase in predictive performance over default settings of the parameters in the UKF and other filters designed to avoid the problems of the UKF, such as the GP-ADF. At the same time, we maintain a lower computational complexity than the other methods. We call our method UKF-L. © 2011 Elsevier B.V.

Mixed source model and its adapted vocal tract filter estimate for voice transformation and synthesis

In current methods for voice transformation and speech synthesis, the vocal tract filter is usually assumed to be excited by a flat amplitude spectrum. In this article, we present a method using a mixed source model defined as a mixture of the Liljencrants-Fant (LF) model and Gaussian noise. Using the LF model, the base approach used in this presented work is therefore close to a vocoder using exogenous input like ARX-based methods or the Glottal Spectral Separation (GSS) method. Such approaches are therefore dedicated to voice processing promising an improved naturalness compared to generic signal models. To estimate the Vocal Tract Filter (VTF), using spectral division like in GSS, we show that a glottal source model can be used with any envelope estimation method conversely to ARX approach where a least square AR solution is used. We therefore derive a VTF estimate which takes into account the amplitude spectra of both deterministic and random components of the glottal source. The proposed mixed source model is controlled by a small set of intuitive and independent parameters. The relevance of this voice production model is evaluated, through listening tests, in the context of resynthesis, HMM-based speech synthesis, breathiness modification and pitch transposition. © 2012 Elsevier B.V. All rights reserved.

Hydrogen-induced morphotropic phase transformation of single-crystalline vanadium dioxide nanobeams.

We report a morphotropic phase transformation in vanadium dioxide (VO2) nanobeams annealed in a high-pressure hydrogen gas, which leads to the stabilization of metallic phases. Structural analyses show that the annealed VO2 nanobeams are hexagonal-close-packed structures with roughened surfaces at room temperature, unlike as-grown VO2 nanobeams with the monoclinic structure and with clean surfaces. Quantitative chemical examination reveals that the hydrogen significantly reduces oxygen in the nanobeams with characteristic nonlinear reduction kinetics which depend on the annealing time. Surprisingly, the work function and the electrical resistance of the reduced nanobeams follow a similar trend to the compositional variation due mainly to the oxygen-deficiency-related defects formed at the roughened surfaces. The electronic transport characteristics indicate that the reduced nanobeams are metallic over a large range of temperatures (room temperature to 383 K). Our results demonstrate the interplay between oxygen deficiency and structural/electronic phase transitions, with implications for engineering electronic properties in vanadium oxide systems.

Programmable transformation of vertically aligned carbon nanotubes into 3D microstructures

Capillary forming of carbon nanotubes (CNTs) enables the fabrication of unique 3D microstructures over large areas. In this paper we focus on the simulation as well as on the integration of these structures in MEMS devices. We developed finite element models (FEM) that enables qualitative prediction of shape transformations caused by capillary forming; and show how capillary formed CNT structured can be integrated with conventional lithographic processing for patterning of polymers and metals in concert with CNTs. © 2011 IEEE.