Nonparallel training for voice conversion based on a parameter adaptation approach

The objective of voice conversion algorithms is to modify the speech by a particular source speaker so that it sounds as if spoken by a different target speaker. Current conversion algorithms employ a training procedure, during which the same utterances spoken by both the source and target speakers are needed for deriving the desired conversion parameters. Such a (parallel) corpus, is often difficult or impossible to collect. Here, we propose an algorithm that relaxes this constraint, i.e., the training corpus does not necessarily contain the same utterances from both speakers. The proposed algorithm is based on speaker adaptation techniques, adapting the conversion parameters derived for a particular pair of speakers to a different pair, for which only a nonparallel corpus is available. We show that adaptation reduces the error obtained when simply applying the conversion parameters of one pair of speakers to another by a factor that can reach 30%. A speaker identification measure is also employed that more insightfully portrays the importance of adaptation, while listening tests confirm the success of our method. Both the objective and subjective tests employed, demonstrate that the proposed algorithm achieves comparable results with the ideal case when a parallel corpus is available.     

Use of topographic correction in Landsat TM-based forest interpretation in Nepal

The land use of two districts in western Nepal was classified into forest and non-forest by satellite imagery with different topographical corrections. The accuracy of the forest biomass and timber volume estimates was also tested. The forest areas were interpretated correctly, with an accuracy of 82-88%. The best result was derived when the Minnaert constants of forest were separated from those of other land uses which were combined. The timber volume and biomass estimates were computed using the weighted nearest neighbour method. The coefficients of determination were 0.06 for timber volume and 0.09 for biomass. The use of the existing land use data seems to have potential for improving the results of topographic normalization. Because the reflectance in mountainous areas depends on many vegetative and terrain features, the simple non-Lambertian correction over broad categories did not significantly improve the results.     

An emergent approach to systems of physical agents

For the past ten years the author has been involved in industrial research on applications of agent architectures to manufacturing, and is presently technical advisor to an industrial consortium that is fielding pilot tests of shop-floor agents. The specific problem addressed by most of their work is the management of material flow and resource utilization on the factory floor. Classically, these problems have been considered under the rubrics of ‘scheduling’ and ‘shop-floor control’, and have been addressed with top-down control that does not offer the robustness and agility required in modern competitive environments. The approach has been to reify parts and equipment as agents, and seek architectures that permit the overall shop schedule to emerge dynamically from their interaction, rather than being imposed top-down.     

Evaluating expectations about negative emotional states of aggressive boys using Bayesian model selection.

Researchers often have expectations about the research outcomes in regard to inequality constraints between, e.g., group means. Consider the example of researchers who investigated the effects of inducing a negative emotional state in aggressive boys. It was expected that highly aggressive boys would, on average, score higher on aggressive responses toward other peers than moderately aggressive boys, who would in turn score higher than nonaggressive boys. In most cases, null hypothesis testing is used to evaluate such hypotheses. We show, however, that hypotheses formulated using inequality constraints between the group means are generally not evaluated properly. The wrong hypotheses are tested, i.e.. the null hypothesis that group means are equal. In this article, we propose an innovative solution to these above-mentioned issues using Bayesian model selection, which we illustrate using a case study. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Convergent and divergent responses to emotional displays of ingroup and outgroup.

In the present research, we test the assumption that emotional mimicry and contagion are moderated by group membership. We report two studies using facial electromyography (EMG; Study 1), Facial Action Coding System (FACS; Study 2), and self-reported emotions (Study 2) as dependent measures. As predicted, both studies show that ingroup anger and fear displays were mimicked to a greater extent than outgroup displays of these emotions. The self-report data in Study 2 further showed specific divergent reactions to outgroup anger and fear displays. Outgroup anger evoked fear, and outgroup fear evoked aversion. Interestingly, mimicry increased liking for ingroup models but not for outgroup models. The findings are discussed in terms of the social functions of emotions in group contexts. (PsycINFO Database Record (c) 2011 APA, all rights reserved)     

Exact solutions to the modified 2D Burgers vortex equation: a general result for the unsteady case

Exact solutions to the unsteady modified 2D Burgers vortex equation are obtained for arbitrary forms of the flow parameter γ(t) via Fourier analysis. The result extends the available literature, where only very specific forms of the flow parameter γ(t) have been considered in the unsteady flow cases. One strength of the method is that we need not assume separable solutions.     

Zirconia-modified exfoliated graphite

Zirconia has been incorporated into exfoliated graphite (EG) through the anodic polarization in the natural graphite-ZrO(NO₃)₂-HNO₃-H₂O system, followed by flash heating. The thermal properties of the oxidized graphites employed as precursors to EG have been studied by thermogravimetry in combination with differential scanning calorimetry, and the distribution of ZrO₂ particles in the EG has been assessed by scanning and transmission electron microscopy. Conditions are described for the preparation of EG with bulk densities in the range 1.3–4.7 g/l and ZrO₂ contents in the range 4–34 wt %.     

Variational density matrix optimization using semidefinite programming

We discuss how semidefinite programming can be used to determine the second-order density matrix directly through a variational optimization. We show how the problem of characterizing a physical or N-representable density matrix leads to matrix-positivity constraints on the density matrix. We then formulate this in a standard semidefinite programming form, after which two interior point methods are discussed to solve the SDP. As an example we show the results of an application of the method on the isoelectronic series of Beryllium.     

A Revised Criterion for the Portevin–Le Chatelier Effect Based on the Strain-Rate Sensitivity of the Work-Hardening Rate

An improved analysis is presented of the stability of plastic deformation under conditions where dynamic strain aging (DSA) occurs, which leads to instabilities known as the Portevin–Le Chatelier (PLC) effect. It is shown that PLC instabilities can occur for conditions that are not covered by the currently prevailing criterion presented by Estrin and Kubin (1991), which focuses on a negative strain rate sensitivity of the flow stress, caused by interactions of solutes with thermally activated glide of mobile dislocations. The current analysis recognizes that the strain-rate sensitivity of the flow stress consists of two contributions, one associated with glide of mobile dislocations and the second with work hardening, related to storage of immobile dislocations. In this paper, an instability criterion is proposed that takes into account the possibility of a negative strain-rate sensitivity of the work-hardening rate, which is caused by diffusion of solutes to immobile dislocations. The latter contribution leads to an extended instability criterion. This criterion also provides an explanation for the existence of a critical strain above which instabilities occur. In this article, previously published tensile test data are used to show that a negative strain-rate sensitivity of the work-hardening rate, which influences significantly the occurrence of the PLC effect, can indeed occur under DSA conditions.