scispace - formally typeset
Search or ask a question
Institution

University of Crete

EducationRethymno, Greece
About: University of Crete is a education organization based out in Rethymno, Greece. It is known for research contribution in the topics: Population & Galaxy. The organization has 8681 authors who have published 21684 publications receiving 709078 citations. The organization is also known as: Panepistimio Kritis.


Papers
More filters
Journal ArticleDOI
TL;DR: Through an analysis of age-related acoustic characteristics of children's speech in the context of automatic speech recognition (ASR), effects such as frequency scaling of spectral envelope parameters are demonstrated and speaker normalization algorithm that combines frequency warping and model transformation is shown to reduce acoustic variability.
Abstract: Developmental changes in speech production introduce age-dependent spectral and temporal variability in the speech signal produced by children. Such variabilities pose challenges for robust automatic recognition of children's speech. Through an analysis of age-related acoustic characteristics of children's speech in the context of automatic speech recognition (ASR), effects such as frequency scaling of spectral envelope parameters are demonstrated. Recognition experiments using acoustic models trained from adult speech and tested against speech from children of various ages clearly show performance degradation with decreasing age. On average, the word error rates are two to five times worse for children speech than for adult speech. Various techniques for improving ASR performance on children's speech are reported. A speaker normalization algorithm that combines frequency warping and model transformation is shown to reduce acoustic variability and significantly improve ASR performance for children speakers (by 25-45% under various model training and testing conditions). The use of age-dependent acoustic models further reduces word error rate by 10%. The potential of using piece-wise linear and phoneme-dependent frequency warping algorithms for reducing the variability in the acoustic feature space of children is also investigated.

213 citations

Journal ArticleDOI
TL;DR: In this article, the authors present a statistical analysis of 248 luminous infrared galaxies (LIRGs) which comprise the Great Observatories All-sky LIRG Survey (GOALS) observed with the Infrared Spectrograph (IRS) on-board Spitzer in the rest-frame wavelength range between 5 and 38 µm.
Abstract: We present a statistical analysis of 248 luminous infrared galaxies (LIRGs) which comprise the Great Observatories All-sky LIRG Survey (GOALS) observed with the Infrared Spectrograph (IRS) on-board Spitzer in the rest-frame wavelength range between 5 and 38 µm. The GOALS sample enables a direct measurement of the relative contributions of star-formation and active galactic nuclei (AGN) to the total infrared (IR) emission from a large, statistically complete sample of LIRGs in the local Universe.Several diagnostics effective at isolating the AGN contribution to the Mid-infrared (MIR) emission using [NeV], [OIV] and [NeII] gas emission lines, the 6.2 µm PAH equivalent width (EQW) and the shape of the MIR continuum are compared. The [NeV] line which indicates the presence of an AGN is detected in 22% of all LIRGs. The 6.2 µm PAH EQW, [NeV]/L_(IR), [NeV]/[NeII] and [OIV]/[NeII] ratios, and the ratios of 6.2 µm PAH flux to the integrated continuum flux between 5.3 and 5.8 µm suggest values of around 10% for the fractional AGN contribution to the total IR luminosity of LIRGs. The median of these estimates suggests that for local LIRGs the fractional AGN contribution to the total IR luminosity is ~12%. AGN dominated LIRGs have higher global and nuclear IR luminosities, warmer MIR colors and are interacting more than starburst (SB) dominated LIRGs. However there are no obvious linear correlations between these properties, suggesting that none of these properties alone can determine the activity and evolution of an individual LIRG. A study of the IRAC colors of LIRGs confirms that methods of finding AGN on the basis of their MIR colors are effective at choosing AGN but 50% to 40% of AGN dominated LIRGs are not selected as such with these methods.

213 citations

Journal ArticleDOI
TL;DR: In this paper, the authors compared the properties of Type-1 and Type-2 Seyfert tori using clumpy torus models and a Bayesian approach to fit the infrared nuclear spectral energy distributions (SEDs).
Abstract: We present new mid-infrared (MIR) imaging data for three Type-1 Seyfert galaxies obtained with T-ReCS on the Gemini-South Telescope at subarcsecond resolution. Our aim is to enlarge the sample studied in a previous work to compare the properties of Type-1 and Type-2 Seyfert tori using clumpy torus models and a Bayesian approach to fit the infrared nuclear spectral energy distributions (SEDs). Thus, the sample considered here comprises 7 Type-1, 11 Type-2, and 3 intermediate-type Seyferts. The unresolved IR emission of the Seyfert 1 galaxies can be reproduced by a combination of dust heated by the central engine and direct AGN emission, while for the Seyfert 2 nuclei only dust emission is considered. These dusty tori have physical sizes smaller than 6 pc radius, as derived from our fits. Unification schemes of AGN account for a variety of observational differences in terms of viewing geometry. However, we find evidence that strong unification may not hold, and that the immediate dusty surroundings of Type-1 and Type-2 Seyfert nuclei are intrinsically different. The Type-2 tori studied here are broader, have more clumps, and these clumps have lower optical depths than those of Type-1 tori. The larger the covering factor of the torus, the smaller the probability of having direct view of the AGN, and vice-versa. In our sample, Seyfert 2 tori have larger covering factors (CT=0.95� 0.02) and smaller escape probabilities (Pesc=0.05� 0.08 0.03 %) than those of Seyfert 1 (CT=0.5� 0.1; Pesc=18� 3 %). All the previous differences are significant according to the KullbackLeibler divergence. Thus, on the basis of the results presented here, the classification of a Seyfert galaxy as a Type-1 or Type-2 depends more on the intrinsic properties of the torus rather than on its mere inclination towards us, in contradiction with the simplest unification model. Subject headings: galaxies: active – galaxies: nuclei – galaxies: Seyfert – infrared: galaxies

213 citations

Journal ArticleDOI
TL;DR: In this article, detailed dynamic equations for the power system and wind energy conversion system (WECS) components and their synthesis to a unified model are presented, which is the basis for creating simulation software able to perform the transient stability analysis of isolated diesel-wind turbine power systems for accurate assessment of their interaction.
Abstract: In the first part of this two-part paper, detailed dynamic equations for the power system and wind energy conversion system (WECS) components and their synthesis to a unified model are presented. This model is the basis for creating simulation software able to perform the transient stability analysis of isolated diesel-wind turbine power systems for accurate assessment of their interaction. Approximations in the various component models, when necessary, remain between limits that do not affect the accuracy of the analysis performed. A new general multimachine power system model is also developed which describes the topology and the complexity of wind-diesel power systems in a compact form which is easy to implement in the simulation software. >

212 citations

Journal ArticleDOI
TL;DR: In this article, the technical efficiency of organic and conventional olive-growing farms using a stochastic production frontier methodology and a translog functional specification is analyzed. But, both input-and output-oriented technical efficiency scores are still relatively low for both types of olive-farming.

212 citations


Authors

Showing all 8725 results

NameH-indexPapersCitations
Mercouri G. Kanatzidis1521854113022
T. J. Pearson150895126533
Stylianos E. Antonarakis13874693605
William Wijns12775295517
Andrea Comastri11170649119
Costas M. Soukoulis10864450208
Elias Anaissie10737242808
Jian Zhang107306469715
Emmanouil T. Dermitzakis10129482496
Andreas Engel9944833494
Nikos C. Kyrpides9671162360
David J. Kerr9554439408
Manolis Kogevinas9562328521
Thomas Walz9225529981
Jean-Paul Latgé9134329152
Network Information
Related Institutions (5)
University of Amsterdam
140.8K papers, 5.9M citations

94% related

University of Helsinki
113.1K papers, 4.6M citations

94% related

Heidelberg University
119.1K papers, 4.6M citations

94% related

University of Paris
174.1K papers, 5M citations

93% related

Rutgers University
159.4K papers, 6.7M citations

93% related

Performance
Metrics
No. of papers from the Institution in previous years
YearPapers
202328
2022103
20211,380
20201,288
20191,180
20181,131