Modelling non-stationary noise with spectral factorisation in automatic speech recognition

Speech recognition systems intended for everyday use must be able to cope with a large variety of noise types and levels, including highly non-stationary multi-source mixtures. This study applies spectral factorisation algorithms and long temporal context for separating speech and noise from mixed signals. To adapt the system to varying environments, noise models are acquired from the context, or learnt from the mixture itself without prior information. We also propose methods for reducing the size of the bases used for speech and noise modelling by 20–40 times for better practical applicability. We evaluate the performance of the methods both as a standalone classifier and as a signal-enhancing front-end for external recognisers. For the CHiME noisy speech corpus containing non-stationary multi-source household noises at signal-to-noise ratios ranging from +9 to ?6 dB, we report average keyword recognition rates up to 87.8% using a single-stream sparse classification algorithm.     

Analysis of an efficient parallel implementation of active-set Newton algorithm

The Journal of Supercomputing - This paper presents an analysis of an efficient parallel implementation of the active-set Newton algorithm (ASNA), which is used to estimate the nonnegative weights...     

From acacia honey monosaccharide content to a new external binary plasticizer applicable in aqueous whey protein films

Plasticizer is an essential adjuvant in food and pharmaceutical film coatings affecting the appearance, mechanical and permeation properties of the final coat. In the present study, film formation and plasticization of native whey proteins (potential future “green” coating agents for pharmaceuticals and food products), were studied with free isolated films. Special attention was paid to the effects of plasticizer, preheating and film forming solution pH on the mechanical stress–strain and moisture permeation properties of whey protein films. Glycerol, binary mixtures of fructose and glucose, and non-crystalline acacia honey were studied as external plasticizers. The type and amount of plasticizer affected the mechanical stress–strain properties of the whey protein films. A short preheating treatment of whey proteins prior to film casting resulted in mechanically strong films with a reduced elongation. The film forming properties of aqueous whey proteins could be modified by adjusting the pH above the isoelectric point of β-lactoglobulin prior to film coating. For effective plasticization, whey protein films required a high amount of monosaccharide containing plasticizer ranging from 80% to 120% (calculated from the protein weight). A new external binary plasticizer having the same ratio of monosaccharides as non-crystalline acacia honey (fructose and glucose 1.67:1) was found to be applicable in aqueous whey protein films.     

Electromagnetic field computation in semiconductor laser resonators

An electromagnetic method based on rigorous diffraction theory of gratings is applied to the analysis of fields in semiconductor laser cavities. The method is based on the Fourier modal method; it is fully rigorous for infinitely periodic resonators and highly accurate for single resonators when absorbing boundary conditions are applied. Fundamental-mode intracavity and near-field distributions are evaluated for some selected geometries, and resonance frequencies are predicted.     

Project-based and temporary organizing: Reconnecting and rediscovering

In recent years, the linkages between project management and organization theory have become stronger. In an attempt to address this development, this paper analyzes the research on temporary and project-based organizing. It especially discusses the development associated with the EGOS sub-themes on project organizing and the potential avenues for future research. The paper also summarizes the key findings from the included papers in the special issue on project-based and temporary organizing, which is based on papers from the EGOS conference in 2013. One key argument is that project organizing needs to develop along three lines: new empirical contexts, new theoretical/conceptual issues, and new research methodologies.     

Surplus equivalence of leveled commitment contracts

In automated negotiation systems consisting of self-interested agents, contracts have traditionally been binding. Leveled commitment contracts—i.e., contracts where each party can decommit by paying a predetermined penalty—were recently shown to improve expected social welfare even if agents decommit strategically in Nash equilibrium. Such contracts differ based on whether agents have to declare their decommitting decisions sequentially or simultaneously, and whether or not agents have to pay the penalties if both decommit. For a given contract, these mechanisms lead to different decommitting thresholds, probabilities, and expected social welfare. However, this paper shows that each of these mechanisms leads to the same social welfare when the contract price and penalties are optimized for each mechanism separately. Our derivations allow agents to construct optimal leveled commitment contracts. We show that such integrative bargaining does not hinder distributive bargaining: the surplus can be divided arbitrarily (as long as each agent benefits), e.g., equally, without compromising optimality. Nonuniqueness questions are answered. We also show that surplus equivalence ceases to hold if agents are not risk neutral.     

Microfluidic Templating of Spatially Inhomogeneous Protein Microgels

3D scaffolds in the form of hydrogels and microgels have allowed for more native cell‐culture systems to be developed relative to flat substrates. Native biological tissues are, however, usually spatially inhomogeneous and anisotropic, but regulating the spatial density of hydrogels at the microscale to mimic this inhomogeneity has been challenging to achieve. Moreover, the development of biocompatible synthesis approaches for protein‐based microgels remains challenging, and typical gelation conditions include UV light, extreme pH, extreme temperature, or organic solvents, factors which can compromise the viability of cells. This study addresses these challenges by demonstrating an approach to fabricate protein microgels with controllable radial density through microfluidic mixing and physical and enzymatic crosslinking of gelatin precursor molecules. Microgels with a higher density in their cores and microgels with a higher density in their shells are demonstrated. The microgels have robust stability at 37 °C and different dissolution rates through enzymolysis, which can be further used for gradient scaffolds for 3D cell culture, enabling controlled degradability, and the release of biomolecules. The design principles of the microgels could also be exploited to generate other soft materials for applications ranging from novel protein‐only micro reactors to soft robots.     

Modeling anisotropic light propagation in a realistic model of the human head

A Monte Carlo model capable of describing photon migration in arbitrary three-dimensional geometry with spatially varying optical properties and tissue anisotropy is presented. We use the model to explore the effects of anisotropy for optical measurements of the human head. An anisotropic diffusion equation that corresponds to our Monte Carlo model is derived, and a comparison between the Monte Carlo model and the diffusion equation solution with finite elements is given.