-
TeLeS: Temporal Lexeme Similarity Score to Estimate Confidence in End-to-End ASR
Authors:
Nagarathna Ravi,
Thishyan Raj T,
Vipul Arora
Abstract:
Confidence estimation of predictions from an End-to-End (E2E) Automatic Speech Recognition (ASR) model benefits ASR's downstream and upstream tasks. Class-probability-based confidence scores do not accurately represent the quality of overconfident ASR predictions. An ancillary Confidence Estimation Model (CEM) calibrates the predictions. State-of-the-art (SOTA) solutions use binary target scores f…
▽ More
Confidence estimation of predictions from an End-to-End (E2E) Automatic Speech Recognition (ASR) model benefits ASR's downstream and upstream tasks. Class-probability-based confidence scores do not accurately represent the quality of overconfident ASR predictions. An ancillary Confidence Estimation Model (CEM) calibrates the predictions. State-of-the-art (SOTA) solutions use binary target scores for CEM training. However, the binary labels do not reveal the granular information of predicted words, such as temporal alignment between reference and hypothesis and whether the predicted word is entirely incorrect or contains spelling errors. Addressing this issue, we propose a novel Temporal-Lexeme Similarity (TeLeS) confidence score to train CEM. To address the data imbalance of target scores while training CEM, we use shrinkage loss to focus on hard-to-learn data points and minimise the impact of easily learned data points. We conduct experiments with ASR models trained in three languages, namely Hindi, Tamil, and Kannada, with varying training data sizes. Experiments show that TeLeS generalises well across domains. To demonstrate the applicability of the proposed method, we formulate a TeLeS-based Acquisition (TeLeS-A) function for sampling uncertainty in active learning. We observe a significant reduction in the Word Error Rate (WER) as compared to SOTA methods.
△ Less
Submitted 6 January, 2024;
originally announced January 2024.
-
The total neutron production from the alpha induced reaction on natural Zirconium
Authors:
Vafiya Thaslim T T,
M M Musthafa,
Midhun C V,
S. Ghugre,
Gokul Das H,
Swapna B,
Najmunnisa T,
F S Shana,
Ri** N T,
S Dasgupta,
J Datta
Abstract:
A significant amount of alpha particles, upto $\sim$ 35 MeV are produced in the reactor environment. Alpha induced reaction on natural Zirconium, a reactor component, upto 40 MeV has been measured using stacked foil activation technique. The total neutron production cross section from all possible channels for $α$ energies upto 35 MeV is also estimated using TALYS 1.96. The isomeric cross section…
▽ More
A significant amount of alpha particles, upto $\sim$ 35 MeV are produced in the reactor environment. Alpha induced reaction on natural Zirconium, a reactor component, upto 40 MeV has been measured using stacked foil activation technique. The total neutron production cross section from all possible channels for $α$ energies upto 35 MeV is also estimated using TALYS 1.96. The isomeric cross section ratio for the production of the radionuclide $^{95}Nb$ has been measured and reported for the first time.
△ Less
Submitted 1 March, 2023;
originally announced March 2023.
-
Explicit criterion of uniform LP duality for linear problems of copositive optimization
Authors:
Kostyukova O. I.,
Tchemisova T. V,
Dudina O. S
Abstract:
An uniform LP duality is an useful property of conic matrix systems. A consistent linear conic optimization problem yields uniform LP duality if for any linear cost function, for which the primal problem has finite optimal value, the corresponding Lagrange dual problem is attainable and the duality gap vanishes.
In this paper, we establish new necessary and sufficient conditions guaranteing the…
▽ More
An uniform LP duality is an useful property of conic matrix systems. A consistent linear conic optimization problem yields uniform LP duality if for any linear cost function, for which the primal problem has finite optimal value, the corresponding Lagrange dual problem is attainable and the duality gap vanishes.
In this paper, we establish new necessary and sufficient conditions guaranteing the uniform LP duality for linear problems of Copositive Programming and formulate these conditions in different equivalent forms. The main results are obtained using an approach developed in previous papers of the authors and based on a concept of immobile indices that permits alternative representations of the set of feasible solutions.
△ Less
Submitted 18 February, 2023;
originally announced February 2023.
-
On equivalent representations and properties of faces of the cone of copositive matrice
Authors:
Kostyukova O. I.,
Tchemisova T. V
Abstract:
The paper is devoted to a study of the cone $\cop$ of copositive matrices. Based on the known from semi-infinite optimization concept of immobile indices, we define zero and minimal zero vectors of a subset of the cone $\cop$ and use them to obtain different representations of faces of $\cop$ and the corresponding dual cones. We describe the minimal face of $\cop$ containing a given convex subset…
▽ More
The paper is devoted to a study of the cone $\cop$ of copositive matrices. Based on the known from semi-infinite optimization concept of immobile indices, we define zero and minimal zero vectors of a subset of the cone $\cop$ and use them to obtain different representations of faces of $\cop$ and the corresponding dual cones. We describe the minimal face of $\cop$ containing a given convex subset of this cone and prove some propositions that allow to obtain equivalent descriptions of the feasible sets of a copositive problems and may be useful for creating new numerical methods based on their regularization.
△ Less
Submitted 7 December, 2020;
originally announced December 2020.
-
Photocatalytic Degradation of Methylene Blue by ZnO/NiFe2O4 Nanoparticles
Authors:
AdelekeJ. T,
Theivasanthi. T,
Thiruppathi. M,
Swaminathan. M,
Akomolafe. T,
AlabiA. B
Abstract:
This work reports the successful solid state synthesis of ZnO/NiFe2O4 nanocomposite by calcinations of both green synthesized ZnO nanorods and NiFe2O4 prismatic nanorods at high temperature (850 °C) for ten hours. These nanomaterials were characterized for structural, functional groups and morphological studies using XRD, FTIR and SEM respectively. The photocatalytic activity of the catalysts was…
▽ More
This work reports the successful solid state synthesis of ZnO/NiFe2O4 nanocomposite by calcinations of both green synthesized ZnO nanorods and NiFe2O4 prismatic nanorods at high temperature (850 °C) for ten hours. These nanomaterials were characterized for structural, functional groups and morphological studies using XRD, FTIR and SEM respectively. The photocatalytic activity of the catalysts was assessed through photodegradation of methylene blue (MB) dye. The photocatalytic studies reveal that the ZnO/NiFe2O4 nanocomposite exhibits high degradation efficiency under UV light. The study shows that in the removal of MB solution by ZnO/NiFe2O4 nanoparticles, OH and photoproduced holes h+ are the main species. Preparation of ZnO/NiFe2O4 nanocomposite by coupling of NiFe2O4 with green synthesized ZnO in solid state method is the novelty of this report. The synthesized recyclable ZnO/NiFe2O4 nanocomposite has potential applications in the treatment of both domestic and industrial waste water.
△ Less
Submitted 26 December, 2018;
originally announced May 2019.
-
A generalized family of anisotropic compact object in general relativity
Authors:
S. K. Maurya,
Saibal Ray,
Shounak Ghosh,
Sarbajit Manna,
Smitha T. T
Abstract:
We present model for anisotropic compact star under the general theory of relativity of Einstein. In the study a 4-dimensional spacetime has been considered which is embedded into the 5-dimensional flat metric so that the spherically symmetric metric has class 1 when the condition $e^λ=\left(\,1+C\,e^ν \,{ν'}^2\,\right)$ is satisfied ($λ$ and $ν$ being the metric potentials along with a constant…
▽ More
We present model for anisotropic compact star under the general theory of relativity of Einstein. In the study a 4-dimensional spacetime has been considered which is embedded into the 5-dimensional flat metric so that the spherically symmetric metric has class 1 when the condition $e^λ=\left(\,1+C\,e^ν \,{ν'}^2\,\right)$ is satisfied ($λ$ and $ν$ being the metric potentials along with a constant $C$). A set of solutions for the field equations are found depending on the index $n$ involved in the physical parameters. The interior solutions have been matched smoothly at the boundary of the spherical distribution to the exterior Schwarzschild solution which necessarily provides values of the unknown constants. We have chosen the values of $n$ as $n=2$ and $n$=10 to 20000 for which interesting and physically viable results can be found out. The numerical values of the parameters and arbitrary constants for different compact stars are assumed in the graphical plots and tables as follows: (i) LMC X-4 : $a=0.0075$, $b=0.000821$ for $n=2$ and $a=0.0075$, $nb=0.00164$ for $n\ge 10$, (ii) SMC X-1: $a=0.00681$, $b=0.00078$ for $n=2$, and $a=0.00681$, $nb=0.00159$ for $n \ge 10$. The investigations on the physical features of the model include several astrophysical issues, like (i) regularity behavior of stars at the centre, (ii) well behaved condition for velocity of sound, (iii) energy conditions, (iv) stabilty of the system via the following three techniques - adiabatic index, Herrera cracking concept and TOV equation, (v) total mass, effective mass and compactification factor and (vi) surface redshift. Specific numerical values of the compact star candidates LMC X-4 and SMC X-1 are calculated for central and surface densities as well as central pressure to compare the model value with actual observational data.
△ Less
Submitted 26 February, 2018; v1 submitted 29 September, 2017;
originally announced October 2017.
-
Modeling of charged anisotropic compact stars in general relativity
Authors:
Baiju Dayanandan,
S. K. Maurya,
Smitha T. T
Abstract:
A charged compact star models have been determined for anisotropic fluid distribution. We have solved the Einstein's- Maxwell field equations to construct the charged compact star models by using radial pressure, metric function $e^λ$ and electric charge function. The generic charged anisotropic solution is verified by exploring different physical conditions like, causality condition, mass-radius…
▽ More
A charged compact star models have been determined for anisotropic fluid distribution. We have solved the Einstein's- Maxwell field equations to construct the charged compact star models by using radial pressure, metric function $e^λ$ and electric charge function. The generic charged anisotropic solution is verified by exploring different physical conditions like, causality condition, mass-radius relation and stability of the solution (via. adiabatic index, TOV equations and Herrera cracking concept). It is observed that the present charged anisotropic compact star is compatible with the star PSR 1937+21. However we also presented the EOS $ρ=f(p)$ for present charged compact star model.
△ Less
Submitted 20 April, 2017; v1 submitted 1 November, 2016;
originally announced November 2016.
-
A new exact solution for anisotropic compact stars of embedding class one
Authors:
S. K. Maurya,
Smitha T. T.,
Y. K. Gupta,
Farook Rahaman
Abstract:
We have presented a new anisotropic solution of Einstein's field equations for compact star models. The Einstein's field equations are solved by using the class one condition \cite{1}. After that we constructed the physically valid expression of anisotropy factor by the help of metric potentials and there after we obtained the physical parameters like energy density, radial and transverse pressure…
▽ More
We have presented a new anisotropic solution of Einstein's field equations for compact star models. The Einstein's field equations are solved by using the class one condition \cite{1}. After that we constructed the physically valid expression of anisotropy factor by the help of metric potentials and there after we obtained the physical parameters like energy density, radial and transverse pressure. These models parameters are well behaved inside the compact star and satisfy all the required physical conditions. Also we observed a very interesting result that all the physical parameters are depend on the anisotropy factor. The present compact star models are quite compatible with the observational astrophysical compact stellar objects like Her X-1, RXJ 1856-37, SAX J1808.4-3658(SS1), SAX J1808.4-3658(SS2)
△ Less
Submitted 28 July, 2016; v1 submitted 5 December, 2015;
originally announced December 2015.
-
Study on two Methods for Nonlinear Force-free Extrapolation Based on Semi-Analytical Field
Authors:
Liu S.,
Zhang H. Q.,
Su J. T. Song M. T
Abstract:
In this paper, two semi-analytical solutions of force free fields (\citeauthor{low90}, \citeyear{low90}) have been used to test two nonlinear force-free extrapolation methods. One is the boundary integral equation (BIE) method developed by \citeauthor{yan00} (\citeyear{yan00}), and another is the approximate vertical integration (AVI) method developed by \citeauthor{son06} (\citeyear{son06}). Some…
▽ More
In this paper, two semi-analytical solutions of force free fields (\citeauthor{low90}, \citeyear{low90}) have been used to test two nonlinear force-free extrapolation methods. One is the boundary integral equation (BIE) method developed by \citeauthor{yan00} (\citeyear{yan00}), and another is the approximate vertical integration (AVI) method developed by \citeauthor{son06} (\citeyear{son06}). Some improvements for the AVI method have been taken to avoid the singular points in the process of calculation. It is found that the correlation coefficients between the first semi-analytical field and extrapolated field by BIE, and also that by improved AVI, are greater than $90\%$ below a height 10 of the $64 \times 64$ lower boundary. While for the second semi-analytical field, these correlation coefficients are greater than $80\%$ below the same relative height. Although the differences between the semi-analytical solutions and the extrapolated fields exist for both BIE and AVI methods, these two methods can give reliable results for the height of about 15$\%$ of the extent of the lower boundary.
△ Less
Submitted 9 May, 2014;
originally announced May 2014.
-
The RMS Survey: Galactic distribution of massive star formation
Authors:
J. S. Urquhart,
C. C. Figura,
T. J. T.,
Moore,
M. G. Hoare,
S. L. Lumsden,
J. C. Mottram,
M. A. Thompson,
R. D. Oudmaijer
Abstract:
Abridged: We have used the well-selected sample of ~1750 embedded, young, massive stars identified by the RMS survey to investigate the Galactic distribution of recent massive star formation. We describe the various methods used to assign distances extracted from the literature, and solve the distance ambiguities towards ~200 sources located within the Solar circle using archival HI data. These di…
▽ More
Abridged: We have used the well-selected sample of ~1750 embedded, young, massive stars identified by the RMS survey to investigate the Galactic distribution of recent massive star formation. We describe the various methods used to assign distances extracted from the literature, and solve the distance ambiguities towards ~200 sources located within the Solar circle using archival HI data. These distances are used to calculate bolometric luminosities and estimate the survey completeness (~2x10^4 lsun). In total, we calculate the distance and luminosity of ~1650 sources, one third of which are above the survey's completeness threshold. Examination of the sample's longitude, latitude, radial velocities and mid-infrared images has identified ~120 small groups of sources, many of which are associated with well known star formation complexes, such as W43, W49 and W51.
We compare the positional distribution of the sample with the expected locations of the spiral arms, assuming a model of the Galaxy consisting of four gaseous arms. The distribution of young massive stars in the Milky Way is spatially correlated with the spiral arms, with strong peaks in the source position and luminosity distributions at the arms' Galactocentric radii. The overall source and luminosity surface densities are both well correlated with the surface density of the molecular gas, which suggests that the massive star formation rate per unit molecular mass is approximately constant across the Galaxy.
We estimate the total luminosity of the embedded massive star population to be ~0.76x10^8 lsun, 30% of which is associated with the ten most active star forming complexes. We measure the scale height as a function of Galactocentric distance and find that it increases only modestly from ~20-30 pc between 4 and 8 kpc, but much more rapidly at larger distances.
△ Less
Submitted 17 October, 2013;
originally announced October 2013.
-
The NIFFTE project
Authors:
Ruz J.,
Asner D. M.,
Baker R. G.,
Bundgaard J.,
Burgett E.,
Cunningham M.,
Deaven J.,
Duke D. L.,
Greife U.,
Grimes S.,
Heffner M.,
Hill T.,
Isenhower D.,
Klay J. L.,
Kleinrath V.,
Kornilov N.,
Laptev A. B.,
Loveland W.,
Masseyf T. N.,
Meharchand R.,
Qu H.,
Sangiorgio S.,
Seilhan B.,
Snyder L.,
Stave S.
, et al. (8 additional authors not shown)
Abstract:
The Neutron Induced Fission Fragment Tracking Experiment (NIFFTE) is a double-sided Time Projection Chamber (TPC) with micromegas readout designed to measure the energy-dependent neutron-induced fission cross sections of the major and minor actinides with unprecedented accuracy. The NIFFTE project addresses the challenge of minimizing major sources of systematic uncertainties from previous fission…
▽ More
The Neutron Induced Fission Fragment Tracking Experiment (NIFFTE) is a double-sided Time Projection Chamber (TPC) with micromegas readout designed to measure the energy-dependent neutron-induced fission cross sections of the major and minor actinides with unprecedented accuracy. The NIFFTE project addresses the challenge of minimizing major sources of systematic uncertainties from previous fission chamber measurements such as: target and beam non-uniformities, misidentification of alpha and light charged particles as fission fragments, and uncertainties inherent to the reference standards used. In-beam tests of the NIFFTE TPC at the Los Alamos Neutron Science Center (LANSCE) started in 2010 and have continued in 2011, 2012 and 2013. An overview of the NIFFTE TPC status and performance at LANSCE will be presented.
△ Less
Submitted 6 November, 2013; v1 submitted 30 September, 2013;
originally announced September 2013.
