No abstract

No abstract

No abstract

No abstract

No abstract

A major challenge in Automatic Speech Recognition(ASR) systems is to handle speech from a diverse set of accents. A model trained using a single accent performs rather poorly when confronted with different accents. One of the solutions is a multicondition model trained on all the accents. However the performance improvement in this approach might be rather limited. Otherwise, accent-specific models might be trained but they become impractical as number of accents increases. In this paper, we propose a novel acoustic model architecture based on Mixture of Experts (MoE) which works well on multiple accents without having the overhead of training separate models for separate accents. The work is based on our earlier work, termed as MixNet, where we showed performance improvement by separation of phonetic class distributions in the feature space. In this paper, we propose an architecture that helps to compensate phonetic and accent variabilities which helps in even better discrimination among the classes. These variabilities are learned in a joint framework , and produce consistent improvements over all the individual accents, amounting to an overall 18% relative improvement in accuracy compared to baseline trained in multi-condition style.

No abstract

Effective atomic numbers (ZPIeff) and electron densities of eighteen alcohols such as wood alcohol, CH3OH; grain alcohol, C2H5OH; rubbing alcohol, C3H7OH; butanol, C4H9OH; amyl alcohol, C5H11OH; cetyl alcohol, C16H33OH; ethylene glycol, C2H4(OH)2; glycerin, C3H5(OH)3; PVA, C2H4O; erythritol, C4H6(OH)4; xylitol, C5H7(OH)5; sorbitol, C6H8(OH)6; volemitol, C7H9(OH)7; allyl alcohol, C3H5OH; geraniol, C10H17OH; propargyl alcohol, C3H3OH; inositol, C6H6(OH)6, and menthol, C10H19OH have been calculated in the photon energy region of 1 keV-100 GeV. The estimated values have been compared with experimental values wherever possible. The comparison of ZPIeff of the alcohols with water phantom and PMMA phantom indicate that the ethylene glycol, glycerin, and PVA are substitute for PMMA phantom and PVA is substitute of water phantom. ZPIeff of alcohols have also been compared with human organs and tissues. Ethylene glycol, glycerin and PVA, allyl alcohol, and wood alcohols are found tissue substitutes for most of human organs. Kerma which is the product of the energy fluence and mass energy-absorption coefficient, have been calculated in the energy region from 1 keV to 20 MeV for the alcohols. The results show the kerma is more or less independent of energy above 100 keV