Multiresolution information measures applied to speech recognition

Considerable advances in automatic speech recognition have been made in the last decades, thanks specially to the use of hidden Markov models. In the field of speech signal analysis, different techniques have been developed. However, deterioration in the performance of the speech recognizers has been observed when they are trained with clean signal and tested with noisy signals. This is still an open problem in this field. Continuous multiresolution entropy has been shown to be robust to additive noise in applications to different physiological signals. In previous works we have included Shannon and Tsallis entropies, and their corresponding divergences, in different speech analysis and recognition systems. In this paper we present an extension of the continuous multiresolution entropy to different divergences and we propose them as new dimensions for the pre-processing stage of a speech recognition system. This approach takes into account information about changes in the dynamics of speech signal at different scales. The methods proposed here are tested with speech signals corrupted with babble and white noise. Their performance is compared with classical mel cepstral parametrization. The results suggest that these continuous multiresolution entropy related measures provide valuable information to the speech recognition system and that they could be considered to be included as an extra component in the pre-processing stage.     

Coarsening-resistant Ag nanoparticles stabilized on amorphous TiO<Subscript>x</Subscript> nanoparticles

Bare Ag nanoparticles (～10 nm) and Ag nanoparticles (1–20 nm) on the surfaces of larger TiO_x nanoparticles were prepared by laser ablation of microparticle aerosols (LAMA). The behaviors of the nanoparticles during high temperature annealing were then studied with ex situ and in situ transmission electron microscopy. For the ex situ heating experiments, Ag and Ag-on-TiO_x NPs were collected onto gold TEM grids and subjected to annealing treatments at 500 °C in argon, vacuum, and air. At this temperature, bare Ag NPs on carbon TEM supports coarsened rapidly in both air and argon atmospheres. In contrast, Ag-on-TiO_x NPs that were heated to 500 °C in flowing argon or in a vacuum did not coarsen significantly and were remarkably stable. Ag-on-TiO_x NPs that were heated to 500 °C in air, however, behaved quite differently. The TiO_x crystallized upon heating and a significant loss of Ag were observed from the surfaces of the TiO_x, likely due to sublimation. These results demonstrate that the surface defect structure and chemistry of the oxide support strongly influence the thermal stability of Ag NPs produced by LAMA.     

A study of the Immune Epitope Database for some fungi species using network topological indices

In the last years, the encryption of system structure information with different network topological indices has been a very active field of research. In the present study, we assembled for the first time a complex network using data obtained from the Immune Epitope Database for fungi species, and we then considered the general topology, the node degree distribution, and the local structure of this network. We also calculated eight node centrality measures for the observed network and compared it with three theoretical models. In view of the results obtained, we may expect that the present approach can become a valuable tool to explore the complexity of this database, as well as for the storage, manipulation, comparison, and retrieval of information contained therein.     

High-energy picosecond Nd:YVO4 slab amplifier for OPCPA pumping

We demonstrate 12-ps pulses with up to 0.6-mJ pulse energy at repetition rates of 50 kHz and 100 kHz from a Nd:YVO₄ slab amplifier built in a simple four-pass configuration. Excellent noise performance with pulse energy fluctuations below 0.8% rms has been achieved by using 10-μJ seed pulses from a highly stable industrial laser system and moderate gain (30–46) in the slab amplifier.     

Detection of nanoparticles in edible plant tissues exposed to nano-copper using single-particle ICP-MS

The increasing use of nanopesticides has raised concerns about their effects on crop plants and the impact of human health as well as ecological effects. While increased uptake of metal ions has been observed before, to date, very few studies have demonstrated the presence of nanoparticles in edible tissues. Single-particle inductively coupled plasma–mass spectrometry (sp-ICP-MS) has been suggested as a powerful tool to detect inorganic nanoparticles (NPs) in environmental samples. Here, we exposed edible plant tissues from lettuce, kale, and collard green to nano-CuO, simulating its use as a nanopesticide. We applied sp-ICP-MS to demonstrate the presence of nanoparticles, both in the water used to rinse crop leaf surfaces exposed to nano-CuO and within the leaf tissues. Lettuces retained the highest amounts of nCuO NPs on the leaf surface, followed by collard green and then kale. Surface hydrophilicity and roughness of the leaf surfaces played an important role in retaining nano-CuO. The results indicate that most of the nanoparticles are removed via washing, but that a certain fraction is taken up by the leaves and can result in human exposure, albeit at low levels.

Open image in new window

Graphical abstract ?

     

Quantum cloning with an optical fiber amplifier

It has been shown theoretically that a light amplifier working on the physical principle of stimulated emission should achieve optimal quantum cloning of the polarization state of light. We demonstrate close-to-optimal universal quantum cloning of polarization in a standard fiber amplifier for telecom wavelengths. For cloning 1-->2 we find a fidelity of 0.82, the optimal value being 5 / 6=0.83.     

Unitary Correlations and the Fejér Kernel

Let M be a unitary matrix with eigenvalues t _j , and let f be a function on the unit circle. Define X _f (M)=f(t _j ). We derive exact and asymptotic formulae for the covariance of X _f and X _g with respect to the measures |(M)|²dM where dM is Haar measure and an irreducible character. The asymptotic results include an analysis of the Fejér kernel which may be of independent interest.     

The structure and dynamics of reacting plane mixing layers

The reacting two-dimensional plane mixing layer has been studied in two configurations: a rearward facing step and a two-stream mixing layer. Observations have been made of the steady state behavior, and the unsteady behavior when the flow was forced by a specific acoustic frequency. The steady behavior of the mean properties of the reacting flows is similar to that of non-reacting free shear flows except for the global effects of thermodynamic property changes. The structure of these flows is qualitatively similar to that of non-reacting flows. Vortices form by the two-dimensional Kelvin-Helmholtz instability and grow by subharmonic combination until the mixing layer interacts with the walls. Entrainment is dominated by the two-dimensional vortex motion. Three-dimensional instabilities give rise to secondary vortices which are coherent over several Kelvin-Helmholtz structures and dominate the fine scale mixing process. The mixing transition corresponds to a loss of coherence in the layer. Unsteady behavior occurs when there are resonant interactions with the Kelvin-Helmholtz instability or the instability associated with the recirculation vortex in the rearward facing step flow. Modeling efforts are reported which show promise of simulating the essential features of plane mixing layers.A version of this paper was presented at the ASME Winter Annual Meeting of 1984 and printed in AMD-Vol. 66