We describe experiments in building HMM text-to-speech voices on professional broadcast news data from multiple speakers. We build on earlier work comparing techniques for selecting utterances from the corpus and voice adaptation to produce the most natural-sounding voices. While our ultimate goal is to develop intelligible and natural-sounding synthetic voices in low-resource languages rapidly...

In this study, spaceborne brightness temperature measurements of snow-covered sparse forest are simulated from a daily data set collected at Sodankylä, Finland during winter 2006-2007. The data set consists of measurement data of soil, snow and air. The atmospheric effects are simulated with two different methods: from measurements near the ground and from balloon-borne radio sounding profiles....

Can infants, in the very first stages of word learning, use their perceptual sensitivity to the phonetics of speech while learning words? Research to date suggests that infants of 14 months cannot learn two similar-sounding words unless there is substantial contextual support. The current experiment advances our understanding of this failure by testing whether the source of infants' difficulty ...

Multi-channel speech enhancement with ad-hoc sensors has been a challenging task. Speech model guided beamforming algorithms are able to recover natural sounding speech, but the speech models tend to be oversimplified or the inference would otherwise be too complicated. On the other hand, deep learning based enhancement approaches are able to learn complicated speech distributions and perform e...

The Sounding Analog Retrieval System (SARS) is a forecasting algorithm that uses sounding derived parameters to find historical severe weather events using a database of proximity soundings associated with a particular type of severe weather. From these matches, SARS arrives at probabilistic forecasts of the severe phenomenon. SARS finds analog matches based on a short list of thermodynamic and...

Nowadays in solving geological problems, the technologies of UAV-geophysics, primarily magnetic and gamma surveys, are being increasingly used. However, for formation classical triad airborne geophysics methods UAV version, there was not enough technology UAV-electromagnetic sounding, which would allow studying environment at depths tens hundreds meters with high detail. This article describes ...

Carbon dioxide, generally considered as the second most abundant volatile component in silicate magmas, is expected to significantly influence various melt properties. In particular, our knowledge about its dynamical effects is lacking over most of Earth's mantle pressure regime. Here, we report the first-principles molecular dynamics results on the transport properties of carbonated MgSiO3 liq...

An overview of the current status of research on the electromagnetic induction sounding of Europa’s ocean and deep interior is provided. After briefly reviewing the history of electromagnetic induction methods used for sounding the interiors of Earth and the Moon, we provide a basic theoretical foundation of electromagnetic wave theory for spherical bodies. Next, evidence of electromagnetic ind...

This paper introduces the technique of retrieving the profiles of vertical distribution of ozone considering temperature and aerosol correction in DIAL sounding of the atmosphere. The authors determine wavelengths, which are promising for measurements of ozone profiles in the upper troposphere–lower stratosphere. An ozone differential absorption lidar is designed for the measurements. The resul...

A magnetotelluric monitoring station has been installed in the Val d’Agri area (southern Italy), to investigate the physics underlying the generation mechanisms of the electrokinetic effect, due to rapid pore pressure changes and fluid flows near the focal area of incoming earthquakes. It is well known that the magnetotelluric method reveals variations in electrical resistivity within the Earth...