Search

Emotions (e.g., Joy, Anger) are prevalent in daily software engineering (SE) activities, and are known to be significant indicators of work productivity (e.g., bug fixing efficiency). Recent studies have shown that directly applying general purpose emotion classification tools to SE corpora is not effective. Even within the SE domain, tool performance degrades significantly when trained on one communication channel and evaluated on another (e.g, StackOverflow vs. GitHub comments). Retraining a tool with channel-specific data takes significant effort since manually annotating large datasets of ground truth data is expensive. In this paper, we address this data scarcity problem by automatically creating new training data using a data augmentation technique. Based on an analysis of the types of errors made by popular SE-specific emotion recognition tools, we specifically target our data augmentation strategy in order to improve the performance of emotion recognition. Our results show an average improvement of 9.3% in micro F1-Score for three existing emotion classification tools (ESEM-E, EMTk, SEntiMoji) when trained with our best augmentation strategy.

Although machine learning (ML) has shown promise in numerous domains, there are concerns about generalizability to out-of-sample data. This is currently addressed by centrally sharing ample, and importantly diverse, data from multiple sites. However, such centralization is challenging to scale (or even not feasible) due to various limitations. Federated ML (FL) provides an alternative to train accurate and generalizable ML models, by only sharing numerical model updates. Here we present findings from the largest FL study to-date, involving data from 71 healthcare institutions across 6 continents, to generate an automatic tumor boundary detector for the rare disease of glioblastoma, utilizing the largest dataset of such patients ever used in the literature (25,256 MRI scans from 6,314 patients). We demonstrate a 33% improvement over a publicly trained model to delineate the surgically targetable tumor, and 23% improvement over the tumor's entire extent. We anticipate our study to: 1) enable more studies in healthcare informed by large and diverse data, ensuring meaningful results for rare diseases and underrepresented populations, 2) facilitate further quantitative analyses for glioblastoma via performance optimization of our consensus model for eventual public release, and 3) demonstrate the effectiveness of FL at such scale and task complexity as a paradigm shift for multi-site collaborations, alleviating the need for data sharing., Comment: federated learning, deep learning, convolutional neural network, segmentation, brain tumor, glioma, glioblastoma, FeTS, BraTS

The purpose of scholarly peer review is to evaluate the quality of scientific manuscripts. However, study after study demonstrates that peer review neither effectively nor reliably assesses research quality. Empirical standards attempt to address this problem by modelling a scientific community's expectations for each kind of empirical study conducted in that community. This should enhance not only the quality of research but also the reliability and predictability of peer review, as scientists adopt the standards in both their researcher and reviewer roles. However, these improvements depend on the quality and adoption of the standards. This tutorial will therefore present the empirical standard for mining software repositories, both to communicate its contents and to get feedback from the attendees. The tutorial will be organized into three parts: (1) brief overview of the empirical standards project; (2) detailed presentation of the repository mining standard; (3) discussion and suggestions for improvement.

Wireless Sensor networks are used to monitor remote areas. Wireless sensor network can be applied to monitor a facility by considering each camera as sensor nodes. Cameras are used as nodes in a wireless sensor network with a central server or a gateway node for all the monitoring and analysis of the information retrieved from the nodes. Identification and authentication of users in any organization is quite difficult due to high movement. Face recognition can be used detect faces and identify them continuously in a video feed which can be deployed to continuously monitor an area. Feeding from camera to base station uses Multi-task Cascaded Convolutional Neural Networks (MCTNN) and FaceNet algorithms for face recognition. Further information about the person is sent to all the end-user nodes present in the wireless network. This approach has been implemented and evaluated on a prototype wired camera network called FaceNet. A method for tracking people in 2D world coordinates and acquiring canonical frontal face images that fits the sensor network paradigm. The approach evaluates and demonstrates the tasking algorithm in action on data acquired from the FaceNet camera network. In this paper, face recognition algorithm FaceNet is used to implement security monitoring network, Comment: 18 pages

Codeathon activity is a practical approach for enduring the principles of Software Engineering and Object Oriented Modelling. Real world domain problem's solution was accomplished through team work. Analysing the problem and designing a feasible solution through a one day activity was achieved through virtual connection. There are three different sections in a semester, 13 teams were framed and assigned one problem statement. Individual team were supposed to prototype a solution which was further used to build one feasible solution. The feedback from students showed different sentiments associated with day long activity. Vivid emotions and expressions of students were analysed., Comment: 10 pages, 15 figures, 1 table

COVID-19 pandemic has introduced a new lifestyle due to lockdown. The impact was on food habits, working hours, and sleeping patterns. The goal of this study is to detect lifestyle changes caused by confinement during the COVID-19 pandemic, such as dietary habits, physical activities, and to explore changes in the body weight. A structured questionnaire was used in the study to collect anthropometric data; daily consumption of particular foods, water intake, food frequency, and number of meals/day. The data is presented in a graph illustration to show health , lifestyle trends and the friendship among engineering students., Comment: 11pages,5figures,5tables

Activity-based learning helps students to learn through participation. A virtual codeathon activity, as part of this learning scheme, was conducted for 180 undergraduate students to focus on analysis and design of solutions to crucial real-world problems in the existing Covid-19 pandemic situation. In this paper, an analysis is made to know the problem solving skills of students given a single problem statement. Evaluators can further collate these multiple solutions into one optimal solution. This Codeathon activity impacts their practical approach towards the analysis and design., Comment: 6 pages, 3 tables

We show that a magnetic field in the high-symmetry direction lifts the macroscopic classical ground-state degeneracy of the honeycomb $\Gamma$ model and induces a long-range magnetic order. While a simple spin-polarized state is stabilized for the ferromagnetic $\Gamma$-exchange, a periodic $\sqrt{3}\times \sqrt{3}$ magnetic order is selected by magnetic field for the antiferromagnetic interaction. We show that the complex spin structure of the tripled unit cell can be described by the magnetization vector and a N\'eel order parameter, similar to those for the spin-flop state of a bipartite antiferromagnet. Indeed, the transition from the low-field plaquette-ordered spin liquid to the field-induced magnetic order can be viewed as a generalized spin-flop transition. An accidental O(2) degeneracy associated with rotation symmetry of the N\'eel vector is broken by either quantum or thermal fluctuations, leaving a six-fold degenerate ground state. At high fields, the breaking of the ground-state $Z_6$ symmetry is through two Berezinskii-Kosterlitz-Thouless transitions that enclose a critical XY phase., Comment: 6 pages, 5 figures

The shell type supernova remnant (SNR) Cas A exhibits structures at nearly all angular scales. Previous studies show the angular power spectrum $(C_{\ell})$ of the radio emission to be a broken power law, consistent with MHD turbulence. The break has been identified with the transition from 2D to 3D turbulence at the angular scale corresponding to the shell thickness. Alternatively, this can also be explained as 2D inverse cascade driven by energy injection from knot-shock interactions. Here we present $C_{\ell}$ measured from archival VLA $5$GHz (C band) data, and Chandra X-ray data in the energy ranges ${\rm A}=0.6-1.0 \, \, {\rm keV}$ and ${\rm B} =4.2-6.0 \, \, {\rm keV}$, both of which are continuum dominated. The different emissions all trace fluctuations in the underlying plasma and possibly also the magnetic field, and we expect them to be correlated. We quantify this using the cross $C_{\ell}$ between the different emissions. We find that X-ray B is strongly correlated with both radio and X-ray A, however X-ray A is only very weakly correlated with radio. This supports a picture where X-ray A is predominantly thermal bremsstrahlung whereas X-ray B is a composite of thermal bremsstrahlung and non-thermal synchrotron emission. The various $C_{\ell}$ measured here, all show a broken power law behaviour. However, the slopes are typically shallower than those in radio and the position of the break also corresponds to smaller angular scales. These findings provide observational inputs regarding the nature of turbulence and the emission mechanisms in Cas A., Comment: Accepted for publication in MNRAS

The study of the intensity fluctuation power spectrum of individual supernova remnants (SNRs) can reveal the structures present at sub-pc scales, and also constrain the physical process that generates those structures. There are various effects, such as the remnant shell thickness, projection of a three-dimensional structure onto a two-dimensional observational plane, and the presence of diffuse "foreground" emission, which causes the observed power spectrum to deviate from the intrinsic power spectrum of the fluctuations. Here, we report results from a systematic study of these effects, using direct numerical simulations, in the measured power spectrum. For an input power-law power spectrum, independent of the power-law index, we see a break in the observed power law at a scale which depends on the shell thickness of a shell-type SNR, and the three-dimensional turbulence changes to two-dimensional turbulence beyond that scale. We also report how the estimated power spectrum is expected to deviate from the intrinsic SNR power spectrum in the presence of additional diffuse Galactic synchrotron emission (DGSE) around the remnant shell. For a reasonable choice of the parameters, if the intrinsic SNR power spectrum is shallower than the DGSE power spectrum, the SNR contribution dominates at small angular scales of the estimated power spectra. On the other hand, if the SNR power spectrum is relatively steeper, the original power spectra is recovered only over a small window of angular scales. This study shows how detailed modeling may be used to infer the true power spectrum from the observed SNR intensity fluctuations power spectrum, which in turn can be used to constrain the nature of the turbulence that gives rise to these small scale structures., Comment: 8 pages, 5 figures, Accepted for publication in MNRAS

We demonstrate machine-learning enabled large-scale dynamical simulations of electronic phase separation in double-exchange system. This model, also known as the ferromagnetic Kondo lattice model, is believed to be relevant for the colossal magnetoresistance phenomenon. Real-space simulations of such inhomogeneous states with exchange forces computed from the electron Hamiltonian can be prohibitively expensive for large systems. Here we show that linear-scaling exchange field computation can be achieved using neural networks trained by datasets from exact calculation on small lattices. Our Landau-Lifshitz dynamics simulations based on machine-learning potentials nicely reproduce not only the nonequilibrium relaxation process, but also correlation functions that agree quantitatively with exact simulations. Our work paves the way for large-scale dynamical simulations of correlated electron systems using machine-learning models., Comment: 6 pages, 4 figures

Segmentation of cardiac images, particularly late gadolinium-enhanced magnetic resonance imaging (LGE-MRI) widely used for visualizing diseased cardiac structures, is a crucial first step for clinical diagnosis and treatment. However, direct segmentation of LGE-MRIs is challenging due to its attenuated contrast. Since most clinical studies have relied on manual and labor-intensive approaches, automatic methods are of high interest, particularly optimized machine learning approaches. To address this, we organized the "2018 Left Atrium Segmentation Challenge" using 154 3D LGE-MRIs, currently the world's largest cardiac LGE-MRI dataset, and associated labels of the left atrium segmented by three medical experts, ultimately attracting the participation of 27 international teams. In this paper, extensive analysis of the submitted algorithms using technical and biological metrics was performed by undergoing subgroup analysis and conducting hyper-parameter analysis, offering an overall picture of the major design choices of convolutional neural networks (CNNs) and practical considerations for achieving state-of-the-art left atrium segmentation. Results show the top method achieved a dice score of 93.2% and a mean surface to a surface distance of 0.7 mm, significantly outperforming prior state-of-the-art. Particularly, our analysis demonstrated that double, sequentially used CNNs, in which a first CNN is used for automatic region-of-interest localization and a subsequent CNN is used for refined regional segmentation, achieved far superior results than traditional methods and pipelines containing single CNNs. This large-scale benchmarking study makes a significant step towards much-improved segmentation methods for cardiac LGE-MRIs, and will serve as an important benchmark for evaluating and comparing the future works in the field.

Minimally Invasive Surgery (MIS) techniques have gained rapid popularity among surgeons since they offer significant clinical benefits including reduced recovery time and diminished post-operative adverse effects. However, conventional endoscopic systems output monocular video which compromises depth perception, spatial orientation and field of view. Suturing is one of the most complex tasks performed under these circumstances. Key components of this tasks are the interplay between needle holder and the surgical needle. Reliable 3D localization of needle and instruments in real time could be used to augment the scene with additional parameters that describe their quantitative geometric relation, e.g. the relation between the estimated needle plane and its rotation center and the instrument. This could contribute towards standardization and training of basic skills and operative techniques, enhance overall surgical performance, and reduce the risk of complications. The paper proposes an Augmented Reality environment with quantitative and qualitative visual representations to enhance laparoscopic training outcomes performed on a silicone pad. This is enabled by a multi-task supervised deep neural network which performs multi-class segmentation and depth map prediction. Scarcity of labels has been conquered by creating a virtual environment which resembles the surgical training scenario to generate dense depth maps and segmentation maps. The proposed convolutional neural network was tested on real surgical training scenarios and showed to be robust to occlusion of the needle. The network achieves a dice score of 0.67 for surgical needle segmentation, 0.81 for needle holder instrument segmentation and a mean absolute error of 6.5 mm for depth estimation., Comment: Accepted for SPIE Medical Imaging 2020

Supernova remnants (SNRs) have a variety of overall morphology as well as rich structures over a wide range of scales. Quantitative study of these structures can potentially reveal fluctuations of density and magnetic field originating from the interaction with ambient medium and turbulence in the expanding ejecta. We have used $1.5$GHz (L band) and $5$GHz (C band) VLA data to estimate the angular power spectrum $C_{\ell}$ of the synchrotron emission fluctuations of the Kepler SNR. This is done using the novel, visibility based, Tapered Gridded Estimator of $C_{\ell}$. We have found that, for $\ell = (1.9 - 6.9) \times 10^{4}$, the power spectrum is a broken power law with a break at $\ell = 3.3 \times 10^{4}$, and power law index of $-2.84\pm 0.07$ and $-4.39\pm 0.04$ before and after the break respectively. The slope $-2.84$ is consistent with 2D Kolmogorov turbulence and earlier measurements for the Tycho SNR. We interpret the break to be related to the shell thickness of the SNR ($0.35 $ pc) which approximately matches $\ell = 3.3 \times 10^{4}$ (i.e., $0.48$ pc). However, for $\ell > 6.9 \times 10^{4}$, the estimated $C_{\ell}$ of L band is likely to have dominant contribution from the foregrounds while for C band the power law slope $-3.07\pm 0.02$ is roughly consistent with $3$D Kolmogorov turbulence like that observed at large $\ell$ for Cas A and Crab SNRs., Comment: 10 pages, 6 figures, 8 tables. Accepted for publication in MNRAS Main journal

We study the spin dynamics of classical Heisenberg antiferromagnet with nearest neighbor interactions on a quasi-two-dimensional kagome bilayer. This geometrically frustrated lattice consists of two kagome layers connected by a triangular-lattice linking layer. By combining Monte Carlo with precessional spin dynamics simulations, we compute the dynamical structure factor of the classical spin liquid in kagome bilayer and investigate the thermal and dilution effects. While the low frequency and long wavelength dynamics of the cooperative paramagnetic phase is dominated by spin diffusion, weak magnon excitations persist at higher energies, giving rise the half moon pattern in the dynamical structure factor. In the presence of spin vacancies, the dynamical properties of the diluted system can be understood within the two population picture. The spin diffusion of the "correlated" spin clusters is mainly driven by the zero-energy weather-van modes, giving rise to an autocorrelation function that decays exponentially with time. On the other hand, the diffusive dynamics of the quasi-free "orphan" spins leads to a distinctive longer time power-law tail in the autocorrelation function. We discuss the implications of our work for the glassy behaviors observed in the archetypal frustrated magnet SrCr$_{9p}$Ga$_{12-9p}$O$_{19}$ (SCGO)., Comment: 10 pages, 12 figures

Many algorithmic steps require more than one statement to implement, but not big enough to be a method (e.g., add element, find the maximum, determine a value, etc.). These steps are generally implemented by loops. Internal comments for the loops often describe these intermediary steps, however, unfortunately a very small percentage of code is well documented to help new users/coders. As a result, information at levels of abstraction between the individual statement and the whole method is not leveraged by current source code analyses, as that information is not easily available beyond any internal comments describing the code blocks. Hence, this project explores the generality of an approach to automatically determine the high level actions of loop constructs. The approach is to mine loop characteristics of a given loop structure over the repository of the Quorum language source code, map it to an (already developed for Java) action identification model, and thus identify the action performed by the specified loop. The results are promising enough to conclude that this approach could be applied to other programming languages too., Comment: This is a report of a preliminary research I conducted in 2015, where I extended upon the research of Wang et al. ("Developing a Model of Loop Actions by Mining Loop Characteristics from a Large Code Corpus" In Proceedings of 31st International Conference on Software Maintenance and Evolution, Bremen, Germany, September 2015. )