-
Gujarati-English Code-Switching Speech Recognition using ensemble prediction of spoken language
Authors:
Yash Sharma,
Basil Abraham,
Preethi Jyothi
Abstract:
An important and difficult task in code-switched speech recognition is to recognize the language, as lots of words in two languages can sound similar, especially in some accents. We focus on improving performance of end-to-end Automatic Speech Recognition models by conditioning transformer layers on language ID of words and character in the output in an per layer supervised manner. To this end, we…
▽ More
An important and difficult task in code-switched speech recognition is to recognize the language, as lots of words in two languages can sound similar, especially in some accents. We focus on improving performance of end-to-end Automatic Speech Recognition models by conditioning transformer layers on language ID of words and character in the output in an per layer supervised manner. To this end, we propose two methods of introducing language specific parameters and explainability in the multi-head attention mechanism, and implement a Temporal Loss that helps maintain continuity in input alignment. Despite being unable to reduce WER significantly, our method shows promise in predicting the correct language from just spoken data. We introduce regularization in the language prediction by drop** LID in the sequence, which helps align long repeated output sequences.
△ Less
Submitted 12 March, 2024;
originally announced March 2024.
-
Surface fields and emergence of long-range couplings in classical uniaxial ferromagnetic arrays
Authors:
D. B. Abraham,
A. Maciolek,
A. Squarcini
Abstract:
Critical wetting is of crucial importance for the phase behaviour of a simple fluid or Ising magnet confined between walls that exert opposing surface fields so that one wall favours liquid (spin up), while the other favours gas (spin down). We show that arrays of boxes filled with fluid or Ising magnet and linked by channels with such ``opposing'' walls can exhibit long-range cooperative effects,…
▽ More
Critical wetting is of crucial importance for the phase behaviour of a simple fluid or Ising magnet confined between walls that exert opposing surface fields so that one wall favours liquid (spin up), while the other favours gas (spin down). We show that arrays of boxes filled with fluid or Ising magnet and linked by channels with such ``opposing'' walls can exhibit long-range cooperative effects, on a length scale far exceeding the bulk correlation length. We give the theoretical foundations of these long-range couplings by using a lattice gas (Ising model) description of a system.
△ Less
Submitted 29 December, 2023;
originally announced December 2023.
-
Fusing machine learning strategy with density functional theory to hasten the discovery of MXenes for hydrogen generation
Authors:
B. Moses Abraham,
Priyanka Sinha,
Prosun Halder,
Jayant K. Singh
Abstract:
The complexity of the topological and combinatorial configuration space of MXenes can give rise to gigantic design challenges that cannot be addressed through traditional experimental or routine theoretical approaches. To this end, we establish a robust and more broadly applicable multistep workflow from the toolbox of supervised machine learning (ML) algorithms for predicting the hydrogen evoluti…
▽ More
The complexity of the topological and combinatorial configuration space of MXenes can give rise to gigantic design challenges that cannot be addressed through traditional experimental or routine theoretical approaches. To this end, we establish a robust and more broadly applicable multistep workflow from the toolbox of supervised machine learning (ML) algorithms for predicting the hydrogen evolution reaction (HER) activity over 4,500 MM$^{\prime}$XT$_2$-type MXenes, where 25\% of the material space (1125 systems) is randomly selected to evaluate the HER performance using density functional theory (DFT) calculations. As the most desirable ML model, the random forest regression method with recursive feature elimination and hyperparameter optimization accurately and rapidly predicts the Gibbs free energy of hydrogen adsorption ($Δ$G$_{H}$) with a low predictive mean absolute error of 0.374 eV. Based on these observations, the H-atom adsorbed directly on top of the outermost metal atomic layer of the MM$^{\prime}$XT$_2$-type MXenes (site-2) with Nb, V, Mo, Cr and Ti metals composed of carbon based O-functionalization are discovered to be highly stable and active catalysts, surpassing that of commercially available platinum based counterparts. Overall, the physically meaningful predictions and insights of the developed ML/DFT-based multistep workflow will open new avenues for accelerated screening, rational design and discovery of potential HER catalysts.
△ Less
Submitted 10 December, 2022;
originally announced December 2022.
-
Thermal energy budget of electrons in the inner heliosphere: Parker Solar Probe Observations
Authors:
Joel B. Abraham,
Daniel Verscharen,
Robert T. Wicks,
Jefferson A. Agudelo Rueda,
Christopher J. Owen,
Georgios Nicolaou,
Seong-Yeop Jeong
Abstract:
We present an observational analysis of the electron thermal energy budget using data from Parker Solar Probe. We use the macroscopic moments, obtained from our fits to the measured electron distribution function, to evaluate the thermal energy budget based on the second moment of the Boltzmann equation. We separate contributions to the overall budget from reversible and irreversible processes. We…
▽ More
We present an observational analysis of the electron thermal energy budget using data from Parker Solar Probe. We use the macroscopic moments, obtained from our fits to the measured electron distribution function, to evaluate the thermal energy budget based on the second moment of the Boltzmann equation. We separate contributions to the overall budget from reversible and irreversible processes. We find that an irreversible thermal energy source must be present in the inner heliosphere over the heliocentric distance range from 0.15 to 0.47 au. The divergence of the heat flux is positive at heliocentric distances below 0.33 au, while beyond 0.33 au, there is a measurable degradation of the heat flux. Expansion effects dominate the thermal energy budget below 0.3 au. Under our steady-state assumption, the free streaming of the electrons is not sufficient to explain the observed thermal energy density budget. We conjecture that the most likely driver for the required heating process is turbulence. Our results are consistent with the known nonadiabatic polytropic index of the electrons, which we measure as 1.18 in the explored range of heliocentric distances.
△ Less
Submitted 20 December, 2022; v1 submitted 3 November, 2022;
originally announced November 2022.
-
Accelerating the Discovery of g-C$_3$N$_4$-Supported Single Atom Catalysts for Hydrogen Evolution Reaction: A Combined DFT and Machine Learning Strategy
Authors:
M. V. Jyothirmai,
D. Roshini,
B. Moses Abraham,
Jayant K. Singh
Abstract:
Two-dimensional materials supported by single atom catalysis (SACs) are foreseen to replace platinum for large-scale industrial scalability of sustainable hydrogen generation. Here, a series of metal (Al, Sc, Ti, V, Cr, Mn, Fe, Ni, Cu, Zn) and non-metal (B, C, N, O, F, Si, P, S, Cl) single atoms embedded on various active sites of g-C$_3$N$_4$ are screened by DFT calculations and six machine learn…
▽ More
Two-dimensional materials supported by single atom catalysis (SACs) are foreseen to replace platinum for large-scale industrial scalability of sustainable hydrogen generation. Here, a series of metal (Al, Sc, Ti, V, Cr, Mn, Fe, Ni, Cu, Zn) and non-metal (B, C, N, O, F, Si, P, S, Cl) single atoms embedded on various active sites of g-C$_3$N$_4$ are screened by DFT calculations and six machine learning (ML) algorithms (support vector regression, gradient boosting regression, random forest regression, AdaBoost regression, multilayer perceptron regression, ridge regression). Our results based on formation energy, Gibbs free energy and bandgap analysis demonstrate that the single atoms of B, Mn and Co anchored on g-C$_3$N$_4$ can serve as highly efficient active sites for hydrogen production. The ML model based on support vector regression (SVR) exhibits the best performance to accurately and rapidly predict the Gibbs free energy of hydrogen adsorption ($Δ$GH ) for the test set with a lower mean absolute error (MAE) and a high coefficient of determination (R$^2$) of 0.45 and 0.81, respectively. Feature selection based on the SVR model highlights the top five primary features: formation energy, bond length, boiling point, melting point, and valance electron as key descriptors. Overall, the multistep work-flow employed through DFT calculations combined with ML models for efficient screening of potential hydrogen evolution reaction (HER) from g-C$_3$N$_4$-based single atom catalysis can significantly contribute to the catalyst design and fabrication.
△ Less
Submitted 3 November, 2022;
originally announced November 2022.
-
Radial evolution of thermal and suprathermal electron populations in the slow solar wind from 0.13 to 0.5 au : Parker Solar Probe Observations
Authors:
Joel B. Abraham,
Christopher J Owen,
Daniel Verscharen,
Mayur Bakrania,
David Stansby,
Robert T. Wicks,
Georgios Nicolaou,
Phyllis L Whittlesey,
Jefferson A. Agudelo Rueda,
Seong-Yeop Jeong,
Laura Bercic
Abstract:
We develop and apply a bespoke fitting routine to a large volume of solar wind electron distribution data measured by Parker Solar Probe (PSP) over its first five orbits, covering radial distances from 0.13 to 0.5 au. We characterise the radial evolution of the electron core, halo and strahl populations in the slow solar wind during these orbits. The fractional densities of these three electron po…
▽ More
We develop and apply a bespoke fitting routine to a large volume of solar wind electron distribution data measured by Parker Solar Probe (PSP) over its first five orbits, covering radial distances from 0.13 to 0.5 au. We characterise the radial evolution of the electron core, halo and strahl populations in the slow solar wind during these orbits. The fractional densities of these three electron populations provide evidence for the growth of the combined suprathermal halo and strahl populations from 0.13 to 0.17 au. Moreover, the growth in the halo population is not matched by a decrease of the strahl population at these distances, as has been reported for previous observations at distances greater than 0.3 au. We also find that the halo is negligible at small heliocentric distances. The fractional strahl density remains relatively constant ~1 % below 0.2 au, suggesting that the rise in the relative halo density is not solely due to the transfer of strahl electrons into the halo.
△ Less
Submitted 1 June, 2022; v1 submitted 11 April, 2022;
originally announced April 2022.
-
The Stability of the Electron Strahl against the Oblique Fast-magnetosonic/Whistler Instability in the Inner Heliosphere
Authors:
Seong-Yeop Jeong,
Joel B. Abraham,
Daniel Verscharen,
Laura Berčič,
David Stansby,
Georgios Nicolaou,
Christopher J. Owen,
Robert T. Wicks,
Andrew N. Fazakerley,
Jeffersson A. Agudelo Rueda,
Mayur Bakrania
Abstract:
We analyze the micro-kinetic stability of the electron strahl in the solar wind depending on heliocentric distance. The oblique fast-magnetosonic/whistler (FM/W) instability has emerged in the literature as a key candidate mechanism for the effective scattering of the electron strahl into the electron halo population. Using data from Parker Solar Probe (PSP) and Helios, we compare the measured str…
▽ More
We analyze the micro-kinetic stability of the electron strahl in the solar wind depending on heliocentric distance. The oblique fast-magnetosonic/whistler (FM/W) instability has emerged in the literature as a key candidate mechanism for the effective scattering of the electron strahl into the electron halo population. Using data from Parker Solar Probe (PSP) and Helios, we compare the measured strahl properties with the analytical thresholds for the oblique FM/W instability in the low- and high-$β_{\parallel c}$ regimes, where $β_{\parallel c}$ is the ratio of the core parallel thermal pressure to the magnetic pressure. Our PSP and Helios data show that the electron strahl is on average stable against the oblique FM/W instability in the inner heliosphere. Our analysis suggests that the instability, if at all, can only be excited sporadically and on short timescales. We discuss the caveats of our analysis and potential alternative explanations for the observed scattering of the electron strahl in the solar wind. Furthermore, we recommend the numerical evaluation of the stability of individual distributions in the future to account for any uncertainties in the validity of the analytical expressions for the instability thresholds.
△ Less
Submitted 25 January, 2022; v1 submitted 24 January, 2022;
originally announced January 2022.
-
The Kinetic Expansion of Solar-Wind Electrons: Transport Theory and Predictions for the very Inner Heliosphere
Authors:
Seong-Yeop Jeong,
Daniel Verscharen,
Christian Vocks,
Joel B. Abraham,
Christopher J. Owen,
Robert T. Wicks,
Andrew N. Fazakerley,
David Stansby,
Laura Berčič,
Georgios Nicolaou,
Jeffersson A. Agudelo Rueda,
Mayur Bakrania
Abstract:
We propose a transport theory for the kinetic evolution of solar-wind electrons in the heliosphere. We derive a gyro-averaged kinetic transport equation that accounts for the spherical expansion of the solar wind and the geometry of the Parker-spiral magnetic field. To solve our three-dimensional kinetic equation, we develop a mathematical approach that combines the Crank--Nicolson scheme in veloc…
▽ More
We propose a transport theory for the kinetic evolution of solar-wind electrons in the heliosphere. We derive a gyro-averaged kinetic transport equation that accounts for the spherical expansion of the solar wind and the geometry of the Parker-spiral magnetic field. To solve our three-dimensional kinetic equation, we develop a mathematical approach that combines the Crank--Nicolson scheme in velocity space and a finite-difference Euler scheme in configuration space. We initialize our model with isotropic electron distribution functions and calculate the kinetic expansion at heliocentric distances from 5 to 20 solar radii. In our kinetic model, the electrons evolve mainly through the combination of the ballistic particle streaming, the magnetic mirror force, and the electric field. By applying fits to our numerical results, we quantify the parameters of the electron strahl and core part of the electron velocity distributions. The strahl fit parameters show that the density of the electron strahl is around 7% of the total electron density at a distance of 20 solar radii, the strahl bulk velocity and strahl temperature parallel to the background magnetic field stay approximately constant beyond a distance of 15 solar radii, and $β_{\parallel s}$ (i.e., the ratio between strahl parallel thermal pressure to the magnetic pressure) is approximately constant with heliocentric distance at a value of about 0.02. We compare our results with data measured by Parker Solar Probe. Furthermore, we provide theoretical evidence that the electron strahl is not scattered by the oblique fast-magnetosonic/whistler instability in the near-Sun environment.
△ Less
Submitted 3 January, 2022; v1 submitted 31 December, 2021;
originally announced January 2022.
-
Zero-modified Count Time Series with Markovian Intensities
Authors:
N. Balakrishna,
Muhammed Anvar,
Bovas Abraham
Abstract:
This paper proposes a method for analyzing count time series with inflation or deflation of zeros. In particular, zero-modified Poisson and zero-modified negative binomial series with intensities generated by non-negative Markov sequences are studied in detail. Parameters of the model are estimated by the method of estimating equations which is facilitated by expressing the model in a generalized…
▽ More
This paper proposes a method for analyzing count time series with inflation or deflation of zeros. In particular, zero-modified Poisson and zero-modified negative binomial series with intensities generated by non-negative Markov sequences are studied in detail. Parameters of the model are estimated by the method of estimating equations which is facilitated by expressing the model in a generalized state space form. The latent intensities required for estimation are extracted using generalized Kalman filter. The applications of proposed model and its estimation methods are illustrated using simulated and real data sets.
△ Less
Submitted 5 July, 2021;
originally announced July 2021.
-
Multilingual and code-switching ASR challenges for low resource Indian languages
Authors:
Anuj Diwan,
Rakesh Vaideeswaran,
Sanket Shah,
Ankita Singh,
Srinivasa Raghavan,
Shreya Khare,
Vinit Unni,
Saurabh Vyas,
Akash Rajpuria,
Chiranjeevi Yarra,
Ashish Mittal,
Prasanta Kumar Ghosh,
Preethi Jyothi,
Kalika Bali,
Vivek Seshadri,
Sunayana Sitaram,
Samarth Bharadwaj,
Jai Nanavati,
Raoul Nanavati,
Karthik Sankaranarayanan,
Tejaswi Seeram,
Basil Abraham
Abstract:
Recently, there is increasing interest in multilingual automatic speech recognition (ASR) where a speech recognition system caters to multiple low resource languages by taking advantage of low amounts of labeled corpora in multiple languages. With multilingualism becoming common in today's world, there has been increasing interest in code-switching ASR as well. In code-switching, multiple language…
▽ More
Recently, there is increasing interest in multilingual automatic speech recognition (ASR) where a speech recognition system caters to multiple low resource languages by taking advantage of low amounts of labeled corpora in multiple languages. With multilingualism becoming common in today's world, there has been increasing interest in code-switching ASR as well. In code-switching, multiple languages are freely interchanged within a single sentence or between sentences. The success of low-resource multilingual and code-switching ASR often depends on the variety of languages in terms of their acoustics, linguistic characteristics as well as the amount of data available and how these are carefully considered in building the ASR system. In this challenge, we would like to focus on building multilingual and code-switching ASR systems through two different subtasks related to a total of seven Indian languages, namely Hindi, Marathi, Odia, Tamil, Telugu, Gujarati and Bengali. For this purpose, we provide a total of ~600 hours of transcribed speech data, comprising train and test sets, in these languages including two code-switched language pairs, Hindi-English and Bengali-English. We also provide a baseline recipe for both the tasks with a WER of 30.73% and 32.45% on the test sets of multilingual and code-switching subtasks, respectively.
△ Less
Submitted 31 March, 2021;
originally announced April 2021.
-
nanoTesla magnetometry with the silicon vacancy in silicon carbide
Authors:
John B. S. Abraham,
Cameron Gutgsell,
Dalibor Todorovski,
Scott Sperling,
Jacob E. Epstein,
Brian S. Tien-Street,
Timothy M. Sweeney,
Jeremiah J. Wathen,
Elizabeth A. Pogue,
Peter G. Brereton,
Tyrel M. McQueen,
Wesley Frey,
B. D. Clader,
Robert Osiander
Abstract:
Silicon Carbide is a promising host material for spin defect based quantum sensors owing to its commercial availability and established techniques for electrical and optical microfabricated device integration. The negatively charged silicon vacancy is one of the leading spin defects studied in silicon carbide owing to its near telecom photoemission, high spin number, and nearly temperature indepen…
▽ More
Silicon Carbide is a promising host material for spin defect based quantum sensors owing to its commercial availability and established techniques for electrical and optical microfabricated device integration. The negatively charged silicon vacancy is one of the leading spin defects studied in silicon carbide owing to its near telecom photoemission, high spin number, and nearly temperature independent ground state zero field splitting. We report the realization of nanoTesla shot-noise limited ensemble magnetometry based on optically detected magnetic resonance with the silicon vacancy in 4H silicon carbide. By coarsely optimizing the anneal parameters and minimizing power broadening, we achieved a sensitivity of 3.5 nT/$\sqrt{Hz}$. This was accomplished without utilizing complex photonic engineering, control protocols, or applying excitation powers greater than a Watt. This work demonstrates that the silicon vacancy in silicon carbide provides a low-cost and simple approach to quantum sensing of magnetic fields.
△ Less
Submitted 17 June, 2021; v1 submitted 2 November, 2020;
originally announced November 2020.
-
Improving Low Resource Code-switched ASR using Augmented Code-switched TTS
Authors:
Yash Sharma,
Basil Abraham,
Karan Taneja,
Preethi Jyothi
Abstract:
Building Automatic Speech Recognition (ASR) systems for code-switched speech has recently gained renewed attention due to the widespread use of speech technologies in multilingual communities worldwide. End-to-end ASR systems are a natural modeling choice due to their ease of use and superior performance in monolingual settings. However, it is well known that end-to-end systems require large amoun…
▽ More
Building Automatic Speech Recognition (ASR) systems for code-switched speech has recently gained renewed attention due to the widespread use of speech technologies in multilingual communities worldwide. End-to-end ASR systems are a natural modeling choice due to their ease of use and superior performance in monolingual settings. However, it is well known that end-to-end systems require large amounts of labeled speech. In this work, we investigate improving code-switched ASR in low resource settings via data augmentation using code-switched text-to-speech (TTS) synthesis. We propose two targeted techniques to effectively leverage TTS speech samples: 1) Mixup, an existing technique to create new training samples via linear interpolation of existing samples, applied to TTS and real speech samples, and 2) a new loss function, used in conjunction with TTS samples, to encourage code-switched predictions. We report significant improvements in ASR performance achieving absolute word error rate (WER) reductions of up to 5%, and measurable improvement in code switching using our proposed techniques on a Hindi-English code-switched ASR task.
△ Less
Submitted 12 October, 2020;
originally announced October 2020.
-
Learning not to Discriminate: Task Agnostic Learning for Improving Monolingual and Code-switched Speech Recognition
Authors:
Gurunath Reddy Madhumani,
Sanket Shah,
Basil Abraham,
Vikas Joshi,
Sunayana Sitaram
Abstract:
Recognizing code-switched speech is challenging for Automatic Speech Recognition (ASR) for a variety of reasons, including the lack of code-switched training data. Recently, we showed that monolingual ASR systems fine-tuned on code-switched data deteriorate in performance on monolingual speech recognition, which is not desirable as ASR systems deployed in multilingual scenarios should recognize bo…
▽ More
Recognizing code-switched speech is challenging for Automatic Speech Recognition (ASR) for a variety of reasons, including the lack of code-switched training data. Recently, we showed that monolingual ASR systems fine-tuned on code-switched data deteriorate in performance on monolingual speech recognition, which is not desirable as ASR systems deployed in multilingual scenarios should recognize both monolingual and code-switched speech with high accuracy. Our experiments indicated that this loss in performance could be mitigated by using certain strategies for fine-tuning and regularization, leading to improvements in both monolingual and code-switched ASR. In this work, we present further improvements over our previous work by using domain adversarial learning to train task agnostic models. We evaluate the classification accuracy of an adversarial discriminator and show that it can learn shared layer parameters that are task agnostic. We train end-to-end ASR systems starting with a pooled model that uses monolingual and code-switched data along with the adversarial discriminator. Our proposed technique leads to reductions in Word Error Rates (WER) in monolingual and code-switched test sets across three language pairs.
△ Less
Submitted 9 June, 2020;
originally announced June 2020.
-
Learning to Recognize Code-switched Speech Without Forgetting Monolingual Speech Recognition
Authors:
Sanket Shah,
Basil Abraham,
Gurunath Reddy M,
Sunayana Sitaram,
Vikas Joshi
Abstract:
Recently, there has been significant progress made in Automatic Speech Recognition (ASR) of code-switched speech, leading to gains in accuracy on code-switched datasets in many language pairs. Code-switched speech co-occurs with monolingual speech in one or both languages being mixed. In this work, we show that fine-tuning ASR models on code-switched speech harms performance on monolingual speech.…
▽ More
Recently, there has been significant progress made in Automatic Speech Recognition (ASR) of code-switched speech, leading to gains in accuracy on code-switched datasets in many language pairs. Code-switched speech co-occurs with monolingual speech in one or both languages being mixed. In this work, we show that fine-tuning ASR models on code-switched speech harms performance on monolingual speech. We point out the need to optimize models for code-switching while also ensuring that monolingual performance is not sacrificed. Monolingual models may be trained on thousands of hours of speech which may not be available for re-training a new model. We propose using the Learning Without Forgetting (LWF) framework for code-switched ASR when we only have access to a monolingual model and do not have the data it was trained on. We show that it is possible to train models using this framework that perform well on both code-switched and monolingual test sets. In cases where we have access to monolingual training data as well, we propose regularization strategies for fine-tuning models for code-switching without sacrificing monolingual accuracy. We report improvements in Word Error Rate (WER) in monolingual and code-switched test sets compared to baselines that use pooled data and simple fine-tuning.
△ Less
Submitted 1 June, 2020;
originally announced June 2020.
-
Canada and the SKA from 2020-2030
Authors:
Kristine Spekkens,
Cynthia Chiang,
Roland Kothes,
Erik Rosolowsky,
Michael Rupen,
Samar Safi-Harb,
Jonathan Sievers,
Greg Sivakoff,
Ingrid Stairs,
Nienke van der Marel,
Bob Abraham,
Rachel Alexandroff,
Norbert Bartel,
Stefi Baum,
Michael Bietenholz,
Aaron Boley,
Dick Bond,
Joanne Brown,
Toby Brown,
Gary Davis,
Jayanne English,
Greg Fahlman,
Laura Ferrarese,
James Di Francesco,
Bryan Gaensler
, et al. (35 additional authors not shown)
Abstract:
This white paper submitted for the 2020 Canadian Long-Range Planning process (LRP2020) presents the prospects for Canada and the Square Kilometre Array (SKA) from 2020-2030, focussing on the first phase of the project (SKA1) scheduled to begin construction early in the next decade. SKA1 will make transformational advances in our understanding of the Universe across a wide range of fields, and Cana…
▽ More
This white paper submitted for the 2020 Canadian Long-Range Planning process (LRP2020) presents the prospects for Canada and the Square Kilometre Array (SKA) from 2020-2030, focussing on the first phase of the project (SKA1) scheduled to begin construction early in the next decade. SKA1 will make transformational advances in our understanding of the Universe across a wide range of fields, and Canadians are poised to play leadership roles in several. Canadian key SKA technologies will ensure a good return on capital investment in addition to strong scientific returns, positioning Canadian astronomy for future opportunities well beyond 2030. We therefore advocate for Canada's continued scientific and technological engagement in the SKA from 2020-2030 through participation in the construction and operations phases of SKA1.
△ Less
Submitted 8 November, 2019;
originally announced November 2019.
-
Plasmodium Detection Using Simple CNN and Clustered GLCM Features
Authors:
Julisa Bana Abraham
Abstract:
Malaria is a serious disease caused by the Plasmodium parasite that transmitted through the bite of a female Anopheles mosquito and invades human erythrocytes. Malaria must be recognized precisely in order to treat the patient in time and to prevent further spread of infection. The standard diagnostic technique using microscopic examination is inefficient, the quality of the diagnosis depends on t…
▽ More
Malaria is a serious disease caused by the Plasmodium parasite that transmitted through the bite of a female Anopheles mosquito and invades human erythrocytes. Malaria must be recognized precisely in order to treat the patient in time and to prevent further spread of infection. The standard diagnostic technique using microscopic examination is inefficient, the quality of the diagnosis depends on the quality of blood smears and experience of microscopists in classifying and counting infected and non-infected cells. Convolutional Neural Networks (CNN) is one of deep learning class that able to automate feature engineering and learn effective features that could be very effective in diagnosing malaria. This study proposes an intelligent system based on simple CNN for detecting malaria parasites through images of thin blood smears. The CNN model obtained high sensitivity of 97% and relatively high PPV of 81%. This study also proposes a false positive reduction method using feature clustering extracted from the gray level co-occurrence matrix (GLCM) from the Region of Interests (ROIs). Adding the GLCM feature can significantly reduce false positives. However, this technique requires manual set up of silhouette and euclidean distance limits to ensure cluster quality, so it does not adversely affect sensitivity.
△ Less
Submitted 28 September, 2019;
originally announced September 2019.
-
End-to-End ASR for Code-switched Hindi-English Speech
Authors:
Brij Mohan Lal Srivastava,
Basil Abraham,
Sunayana Sitaram,
Rupesh Mehta,
Preethi Jyothi
Abstract:
End-to-end (E2E) models have been explored for large speech corpora and have been found to match or outperform traditional pipeline-based systems in some languages. However, most prior work on end-to-end models use speech corpora exceeding hundreds or thousands of hours. In this study, we explore end-to-end models for code-switched Hindi-English language with less than 50 hours of data. We utilize…
▽ More
End-to-end (E2E) models have been explored for large speech corpora and have been found to match or outperform traditional pipeline-based systems in some languages. However, most prior work on end-to-end models use speech corpora exceeding hundreds or thousands of hours. In this study, we explore end-to-end models for code-switched Hindi-English language with less than 50 hours of data. We utilize two specific measures to improve network performance in the low-resource setting, namely multi-task learning (MTL) and balancing the corpus to deal with the inherent class imbalance problem i.e. the skewed frequency distribution over graphemes. We compare the results of the proposed approaches with traditional, cascaded ASR systems. While the lack of data adversely affects the performance of end-to-end models, we see promising improvements with MTL and balancing the corpus.
△ Less
Submitted 22 June, 2019;
originally announced June 2019.
-
Action at a distance in classical uniaxial ferromagnetic arrays
Authors:
D. B. Abraham,
A. Maciołek,
A. Squarcini,
O. Vasilyev
Abstract:
We examine in detail the theoretical foundations of striking long-range couplings emerging in arrays of fluid cells connected by narrow channels by using a lattice gas (Ising model) description of a system. We present a reexamination of the well known exact determination of the two-point correlation function along the edge of a channel using the transfer matrix technique and a new interpretation i…
▽ More
We examine in detail the theoretical foundations of striking long-range couplings emerging in arrays of fluid cells connected by narrow channels by using a lattice gas (Ising model) description of a system. We present a reexamination of the well known exact determination of the two-point correlation function along the edge of a channel using the transfer matrix technique and a new interpretation is provided. The explicit form of the correlation length is found to grow exponentially with the cross section of the channels at the bulk two-phase coexistence. The aforementioned result is recaptured by a refined version of the Fisher-Privman theory of first order phase transitions in which the Boltzmann factor for a domain wall is decorated with a contribution stemming from the point tension originated at its endpoints. The Boltzmann factor for a domain wall together with the point tension is then identified exactly thanks to two independent analytical techniques, providing a critical test of the Fisher-Privman theory. We then illustrate how to build up the network model from its elementary constituents, the cells and the channels. Moreover, we are able to extract the strength of the coupling between cells and express them in terms of the length and width and coarse grained quantities such as surface and point tensions. We then support our theoretical investigation with a series of corroborating results based on Monte Carlo simulations. We illustrate how the long range ordering occurs and how the latter is signaled by the thermodynamic quantities corresponding to both planar and three-dimensional Ising arrays.
△ Less
Submitted 3 July, 2017;
originally announced July 2017.
-
Detection and Analysis of Emotion From Speech Signals
Authors:
Assel Davletcharova,
Sherin Sugathan,
Bibia Abraham,
Alex Pappachen James
Abstract:
Recognizing emotion from speech has become one the active research themes in speech processing and in applications based on human-computer interaction. This paper conducts an experimental study on recognizing emotions from human speech. The emotions considered for the experiments include neutral, anger, joy and sadness. The distinuishability of emotional features in speech were studied first follo…
▽ More
Recognizing emotion from speech has become one the active research themes in speech processing and in applications based on human-computer interaction. This paper conducts an experimental study on recognizing emotions from human speech. The emotions considered for the experiments include neutral, anger, joy and sadness. The distinuishability of emotional features in speech were studied first followed by emotion classification performed on a custom dataset. The classification was performed for different classifiers. One of the main feature attribute considered in the prepared dataset was the peak-to-peak distance obtained from the graphical representation of the speech signals. After performing the classification tests on a dataset formed from 30 different subjects, it was found that for getting better accuracy, one should consider the data collected from one person rather than considering the data from a group of people.
△ Less
Submitted 22 June, 2015;
originally announced June 2015.
-
Third sound measurements of superfluid $^4$He films on multiwall carbon nanotubes below 1K
Authors:
Emin Menachekanian,
John B. S. Abraham,
Bob Chen,
Vito Iaia,
Andrew Li,
Gary A. Williams
Abstract:
Third sound is studied for superfluid films of 4He adsorbed on multiwall carbon nanotubes packed into an annular resonator. The third sound is generated with mechanical oscillation of the cell, and detected with carbon bolometers. A filling curve at temperatures near 250 mK shows oscillations in the third sound velocity, with maxima at the completion of the 4th and 5th atomic layers. Sharp changes…
▽ More
Third sound is studied for superfluid films of 4He adsorbed on multiwall carbon nanotubes packed into an annular resonator. The third sound is generated with mechanical oscillation of the cell, and detected with carbon bolometers. A filling curve at temperatures near 250 mK shows oscillations in the third sound velocity, with maxima at the completion of the 4th and 5th atomic layers. Sharp changes in the Q factor of the third sound are found at partial layer fillings. Temperature sweeps at a number of fill points show strong broadening effects on the Kosterlitz-Thouless (KT) transition, and rapidly increasing dissipation, in qualitative agreement with the predictions of Machta and Guyer. At the 4th layer completion there is a sudden reduction of the transition temperature $T_{KT}$, and then a recovery back to linear variation with temperature, although the slope is considerably smaller than the KT prediction. Some of these effects may be related to changes in the gas-liquid coexistence regions.
△ Less
Submitted 18 November, 2014;
originally announced November 2014.
-
Apparent first-order wetting and anomalous scaling in the two-dimensional Ising model
Authors:
X. T. Wu,
D. B. Abraham,
J. O. Indekeu
Abstract:
The global phase diagram of wetting in the two-dimensional (2d) Ising model is obtained through exact calculation of the surface excess free energy. Besides a surface field for inducing wetting, a surface-coupling enhancement is included. The wetting transition is critical (second order) for any finite ratio of surface coupling J_s to bulk coupling J, and turns first order in the limit J_s/J to in…
▽ More
The global phase diagram of wetting in the two-dimensional (2d) Ising model is obtained through exact calculation of the surface excess free energy. Besides a surface field for inducing wetting, a surface-coupling enhancement is included. The wetting transition is critical (second order) for any finite ratio of surface coupling J_s to bulk coupling J, and turns first order in the limit J_s/J to infinity. However, for J_s/J much larger than 1 the critical region is exponentially small and practically invisible to numerical studies. A distinct pre-asymptotic regime exists in which the transition displays first-order character. Surprisingly, in this regime the surface susceptibility and surface specific heat develop a divergence and show anomalous scaling with an exponent equal to 3/2.
△ Less
Submitted 3 January, 2016; v1 submitted 30 October, 2014;
originally announced October 2014.
-
Action at the distance
Authors:
D. B. Abraham,
A. Maciolek,
O. Vasilyev
Abstract:
We present a system exhibiting giant proximity effects which parallel observations in superfluid helium (Perron et al, Nature Physics V. 6, 499 (2010)) and give a theoretical explanation of these phenomena based on the mesoscopic picture of phase coexistence in finite systems. Our theory is confirmed by MC simulation studies. Our work demonstrates that such action-at-a-distance can occur in classi…
▽ More
We present a system exhibiting giant proximity effects which parallel observations in superfluid helium (Perron et al, Nature Physics V. 6, 499 (2010)) and give a theoretical explanation of these phenomena based on the mesoscopic picture of phase coexistence in finite systems. Our theory is confirmed by MC simulation studies. Our work demonstrates that such action-at-a-distance can occur in classical systems involving simple or complex fluids, such as colloid-polymer mixtures, or ferromagnets.
△ Less
Submitted 21 May, 2014;
originally announced May 2014.
-
Non-universal Casimir Effect in Saturated Superfluid $^4$He Films at T$_λ$
Authors:
John B. S. Abraham,
Gary A. Williams,
Konstantin Penanen
Abstract:
Measurements of Casimir effects in $^4$He films in the vicinity of the bulk superfluid transition temperature $T_λ$ have been carried out, where changes in the film thickness and the superfluid density are both monitored as a function of temperature. The Kosterlitz-Thouless superfluid onset temperature in the film is found to occur just as the Casimir dip in the film thickness from critical fluctu…
▽ More
Measurements of Casimir effects in $^4$He films in the vicinity of the bulk superfluid transition temperature $T_λ$ have been carried out, where changes in the film thickness and the superfluid density are both monitored as a function of temperature. The Kosterlitz-Thouless superfluid onset temperature in the film is found to occur just as the Casimir dip in the film thickness from critical fluctuations becomes evident. Additionally, a new film-thickening effect is observed precisely at $T_λ$ when the temperature is swept extremely slowly. We propose that this is a non-universal Casimir effect arising from the viscous suppression of second sound modes in the film.
△ Less
Submitted 16 December, 2013; v1 submitted 3 September, 2013;
originally announced September 2013.
-
Collective Effects in Linear Spectroscopy of Dipole-Coupled Molecular Arrays
Authors:
A. A. Kocherzhenko,
J. Dawlaty,
B. P. Abolins,
F. Herrera,
D. B. Abraham,
K. B. Whaley
Abstract:
We present a consistent analysis of linear spectroscopy for arrays of nearest neighbor dipole-coupled two-level molecules that reveals distinct signatures of weak and strong coupling regimes separated for infinite size arrays by a quantum critical point. In the weak coupling regime, the ground state of the molecular array is disordered, but in the strong coupling regime it has (anti)ferroelectric…
▽ More
We present a consistent analysis of linear spectroscopy for arrays of nearest neighbor dipole-coupled two-level molecules that reveals distinct signatures of weak and strong coupling regimes separated for infinite size arrays by a quantum critical point. In the weak coupling regime, the ground state of the molecular array is disordered, but in the strong coupling regime it has (anti)ferroelectric ordering. We show that multiple molecular excitations (odd/even in weak/strong coupling regime) can be accessed directly from the ground state. We analyze the scaling of absorption and emission with system size and find that the oscillator strengths show enhanced superradiant behavior in both ordered and disordered phases. As the coupling increases, the single excitation oscillator strength rapidly exceeds the well known Heitler-London value. In the strong coupling regime we show the existence of a unique spectral transition with excitation energy that can be tuned by varying the system size and that asymptotically approaches zero for large systems. The oscillator strength for this transition scales quadratically with system size, showing an anomalous one-photon superradiance. For systems of infinite size, we find a novel, singular spectroscopic signature of the quantum phase transition between disordered and ordered ground states. We outline how arrays of ultra cold dipolar molecules trapped in an optical lattice can be used to access the strong coupling regime and observe the anomalous superradiant effects associated with this regime.
△ Less
Submitted 16 October, 2014; v1 submitted 10 July, 2013;
originally announced July 2013.
-
Surface states and the thermal Casimir interaction
Authors:
Douglas B. Abraham,
Anna Maciolek
Abstract:
Using exact calculations, we elucidate the significance of the surface (bound) states for the thermal Casimir interactions for an Ising strip with a finite width. The surface state arises whenever an imaginary wavenumber mode appears in the spectrum of the transfer matrix, taken in the direction parallel to the edges of the strip. Depending on the boundary conditions, the imaginary modes emerge be…
▽ More
Using exact calculations, we elucidate the significance of the surface (bound) states for the thermal Casimir interactions for an Ising strip with a finite width. The surface state arises whenever an imaginary wavenumber mode appears in the spectrum of the transfer matrix, taken in the direction parallel to the edges of the strip. Depending on the boundary conditions, the imaginary modes emerge below or above the bulk critical temperature, or below the wetting temperature of a single surface with surface magnetic field. The bound states are responsible for the strong asymmetry of the Casimir forces between the super- and sub-critical regimes and for their sign. Our analysis uses the fact that the Casimir forces have two mathematical forms. We show that these very different representation are the same and in the process find the origin of the asymmetry.
△ Less
Submitted 25 June, 2012;
originally announced June 2012.
-
Casimir interactions in Ising strips with boundary fields: exact results
Authors:
Douglas B. Abraham,
Anna Maciolek
Abstract:
An exact statistical mechanical derivation is given of the critical Casimir forces for Ising strips with arbitrary surface fields applied to edges. Our results show that the strength as well as the sign of the force can be controled by varying the temperature or the fields. An interpretation of the results is given in terms of a linked cluster expansion. This suggests a systematic approach for d…
▽ More
An exact statistical mechanical derivation is given of the critical Casimir forces for Ising strips with arbitrary surface fields applied to edges. Our results show that the strength as well as the sign of the force can be controled by varying the temperature or the fields. An interpretation of the results is given in terms of a linked cluster expansion. This suggests a systematic approach for deriving the critical Casimir force which can be used in more general models.
△ Less
Submitted 1 December, 2009;
originally announced December 2009.
-
Interfaces in driven Ising models: shear enhances confinement
Authors:
Thomas H. R. Smith,
Oleg Vasilyev,
Douglas B. Abraham,
Anna Maciołek,
Matthias Schmidt
Abstract:
We use a phase-separated driven two-dimensional Ising lattice gas to study fluid interfaces exposed to shear flow parallel to the interface. The interface is stabilized by two parallel walls with opposing surface fields and a driving field parallel to the walls is applied which (i) either acts locally at the walls or (ii) varies linearly with distance across the strip. Using computer simulations…
▽ More
We use a phase-separated driven two-dimensional Ising lattice gas to study fluid interfaces exposed to shear flow parallel to the interface. The interface is stabilized by two parallel walls with opposing surface fields and a driving field parallel to the walls is applied which (i) either acts locally at the walls or (ii) varies linearly with distance across the strip. Using computer simulations with Kawasaki dynamics, we find that the system reaches a steady state in which the magnetisation profile is the same as that in equilibrium, but with a rescaled length implying a reduction of the interfacial width. An analogous effect was recently observed in sheared phase-separated colloidal dispersions. Pair correlation functions along the interface decay more rapidly with distance under drive than in equilibrium and for cases of weak drive can be rescaled to the equilibrium result.
△ Less
Submitted 15 July, 2008; v1 submitted 26 February, 2008;
originally announced February 2008.
-
Effective forces induced by fluctuating interface: exact results
Authors:
D. B. Abraham,
F. H. L. Essler,
A. Maciolek,
;
Abstract:
We present exact derivations of the effective capillary wave fluctuation induced forces resulting from pinning of an interface between two coexisting phases at two points separated by a distance r. In two dimensions the Ising ferromagnet calculations based on the transfer matrix approach give an attractive force decaying as 1/r for large distances. In three dimensions map** of the body-centere…
▽ More
We present exact derivations of the effective capillary wave fluctuation induced forces resulting from pinning of an interface between two coexisting phases at two points separated by a distance r. In two dimensions the Ising ferromagnet calculations based on the transfer matrix approach give an attractive force decaying as 1/r for large distances. In three dimensions map** of the body-centered solid-on-solid model onto the 6-vertex model allows for exact solution using the bosonization analysis of the equivalent XXZ Heisenberg quantum chain. The exact result gives the attractive force which decays asymptotically as 1/(r\log r).
△ Less
Submitted 30 March, 2007;
originally announced March 2007.
-
A microscopic model for thin film spreading
Authors:
Douglas B. Abraham,
Rodolfo Cuerno,
Esteban Moro
Abstract:
A microscopic, driven lattice gas model is proposed for the dynamics and spatio-temporal fluctuations of the precursor film observed in spreading experiments. Matter is transported both by holes and particles, and the distribution of each can be described by driven diffusion with a moving boundary. This picture leads to a stochastic partial differential equation for the shape of the boundary, wh…
▽ More
A microscopic, driven lattice gas model is proposed for the dynamics and spatio-temporal fluctuations of the precursor film observed in spreading experiments. Matter is transported both by holes and particles, and the distribution of each can be described by driven diffusion with a moving boundary. This picture leads to a stochastic partial differential equation for the shape of the boundary, which agrees with the simulations of the lattice gas. Preliminary results for flow in a thermal gradient are discussed.
△ Less
Submitted 14 January, 2002;
originally announced January 2002.
-
Faceting Transition in an Exactly Solvable Terrace-Ledge-Kink model
Authors:
D. B. Abraham,
F. H. L. Essler,
F. T. Latremoliere
Abstract:
We solve exactly a Terrace-Ledge-Kink (TLK) model describing a crystal surface at a microscopic level. We show that there is a faceting transition driven either by temperature or by the chemical potential that controls the slope of the surface. In the rough phase we investigate thermal fluctuations of the surface using Conformal Field Theory.
We solve exactly a Terrace-Ledge-Kink (TLK) model describing a crystal surface at a microscopic level. We show that there is a faceting transition driven either by temperature or by the chemical potential that controls the slope of the surface. In the rough phase we investigate thermal fluctuations of the surface using Conformal Field Theory.
△ Less
Submitted 24 August, 1998;
originally announced August 1998.
-
Non-equilibrium Dynamics of Finite Interfaces
Authors:
D. B. Abraham,
T. J. Newman,
G. M. Schütz
Abstract:
We present an exact solution to an interface model representing the dynamics of a domain wall in a two-phase Ising system. The model is microscopically motivated, yet we find that in the scaling regime our results are consistent with those obtained previously from a phenomenological, coarse-grained Langevin approach.
We present an exact solution to an interface model representing the dynamics of a domain wall in a two-phase Ising system. The model is microscopically motivated, yet we find that in the scaling regime our results are consistent with those obtained previously from a phenomenological, coarse-grained Langevin approach.
△ Less
Submitted 20 April, 1994;
originally announced April 1994.
