MD-SOAR Maryland Shared Open Access Repository
Not a member yet
34521 research outputs found
Sort by
XRISM/Xtend Transient Search (XTS) detected an X-ray flare from XRISM J0057+6021
No full textAuthors: N. Nagashima (Chuo U.), Y. Kanemaru, T. Yoshida, K. Fukushima, K. Hayashi, S. Ogawa (JAXA), M. Audard (U. de Geneve), E. Behar (Technion), S. Inoue (Kyoto U.), Y. Ishihara (Chuo U.), T. Kohmura (TUS), Y. Maeda (JAXA), M. Mizumoto (UTEF), M. Nobukawa (NUE), K. Pottschmidt (UMBC, NASA GSFC, CRESST), M. Shidatsu (Ehime U.), H. Sugai (Chuo U.), Y. Terada (Saitama U.), Y. Terashima (Ehime U.), Y. Tsuboi (Chuo U.), H. Uchida (Kyoto U.), T. Yanagi (Chuo U.), T. Yoneyama (Chuo U.), M. Yoshimoto (Osaka U.)XRISM/Xtend Transient Search (XTS) detected an X-ray flare from an X-ray source XRISM J0057+6021 on 2024-12-29 TT. The source position is determined to be (R.A., Dec.) = (14.195, 60.356), with a systematic error of ∼ 40 arcsec. A plausible counterpart is a multiple star HD 5408 (B7Vn+B9VHgMn+A1V). HD 5408 is located ∼ 20 arcsec apart from the position of XRISM J0057+6021. All statistical uncertainties in this report will be provided as a 90% confidence level unless stated otherwise.
The flare started on 2024-12-29 at ∼ 11h TT. The flare exponentially decayed in 5 × 10³ sec. In order to estimate the source flux, we fit the spectrum in the flare phase with an absorbed APEC model with a temperature of kT = 0.5 keV and hydrogen column density NH = 6 × 10²¹ cm⁻². Then, the model peak flux is calculated as 1 × 10⁻¹² erg s⁻¹ cm⁻² (0.4 – 10.0 keV). A systematic error of roughly 20% should be added to the statistical error. Corresponding luminosity is 3 × D₁₆₅ₚ꜀² × 10³⁰ erg s⁻¹ by assuming the distance to XRISM J0057+6021 of D₁₆₅ₚ꜀. We derived the above systematic error for the flux by comparing our derived values for the sources detected with XTS in several observations with those for the corresponding X-ray counterparts. We estimated the systematic error for the source position from the separations between the detected sources with the corresponding counterparts in the same field of view.https://www.astronomerstelegram.org/?read=1696
Parenting and Coping During a Crisis: A Qualitative Cross-Cultural Study Two Years After COVID-19
No full textThe COVID-19 pandemic unprecedentedly challenged families worldwide, yet little is known about how parents from diverse cultural contexts retrospectively interpret their parenting roles and coping strategies. This study explores parenting adjustments two years after the pandemic’s onset among five cultural groups: Bulgarian and Spanish (Eastern and Western Europe), Israeli Arabs and Jews (Middle East), and U.S. families. Fifty parents, primarily mothers of children aged 2–8, were recruited through snowball sampling. Semi-structured interviews were conducted using the Parenting Pentagon Model (PPM), which includes five constructs: Partnership, Parental Leadership, Love, Encouraging Independence, and Adherence to Rules. Data were analyzed using grounded theory and directed content analysis. Across cultures, Love and Parental Leadership were central to maintaining emotional stability and family cohesion. Partnership showed cultural variation: Bulgarian and Spanish parents often shared responsibilities, while U.S. mothers reported handling childcare alone, heightening work–life tension. Israeli-Arab fathers became more involved in caregiving, while Israeli-Jewish mothers described both strengthened and strained partnerships. Coping strategies were shaped by cultural values and family demographics (e.g., family size). The findings emphasize parents’ vital role in fostering family resilience during crises and stress the importance of culturally sensitive support to enhance families’ adaptive capacity for future challenges.Funding: This research received no external funding.https://www.mdpi.com/2227-7102/15/9/111
MaGrIP: Magnitude and Gradient-Informed Pruning for Task-Agnostic Large Language Models
No full textLarge Language Models (LLMs) have become foundational tools in natural language processing, achieving state-of-the-art performance across a variety of tasks. However, their immense size and computational requirements make them impractical for deployment in resource-constrained environments, such as edge devices and embedded systems. In this work, we introduce Magnitude and Gradient-Informed Pruning (MaGrIP), a novel framework for task-agnostic pruning and compression of LLMs. MaGrIP employs a dual-threshold strategy combining magnitude- and gradient-based saliency measures to efficiently prune redundant neurons while retaining task performance. Our results demonstrate the effectiveness of MaGrIP in compressing state-of-the-art models. The compression reduced the total computational complexity of the FFN layers from O (d · ℎ) to O ( (d − q) · ℎ) . In terms of model size, our pruning approach significantly reduces both model parameters and storage requirements while maintaining competitive perplexity scores evaluated on WikiText-2. For the Gemma 7B model, our method reduces the total size from 28 GB to 5 GB, while for Gemma 2B, MaGrIP achieves a size reduction from 8 GB to 1.5 GB. MaGrIP furthermore exhibits robust performance across multiple benchmarks, such as BOOLQ, ARC-E, and CSQA. Specifically, the pruned Gemma 7B model at 50% pruning achieved 59.26% accuracy on ARC-E compared to 81.06% for the baseline, and 64.74% accuracy on BoolQ compared to 59.98% for the baseline. Similarly, the pruned Llama 3 8B at 50% pruning achieved 46.76% accuracy on ARC-E compared to 77.57% for the baseline, reflecting the trade-off between compression and accuracy. LLMs compressed using MaGrIP, when deployed on the Nvidia Jetson Orin Nano, achieved a 2.16 × improvement in throughput and a 2.3 × improvement in performance compared to baseline LLMs.Research was partly sponsored by the Army Research Laboratory and was accomplished under Cooperative Agreement Number W911NF-24-2-0080. The views and conclusions contained in this document are those of the authors and should not be interpreted as representing the oicial policies, either expressed or implied, of the Army Research Laboratory or the U.S. Government. The U.S. Government is authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation herein.https://dl.acm.org/doi/10.1145/376606
BanglaTalk: Towards Real-Time Speech Assistance for Bengali Regional Dialects
No full textReal-time speech assistants are becoming increasingly popular for ensuring improved accessibility to information. Bengali, being a low-resource language with a high regional dialectal diversity, has seen limited progress in developing such systems. Existing systems are not optimized for real-time use and focus only on standard Bengali. In this work, we present BanglaTalk, the first real-time speech assistance system for Bengali regional dialects. BanglaTalk follows the client-server architecture and uses the Real-time Transport Protocol (RTP) to ensure low-latency communication. To address dialectal variation, we introduce a dialect-aware ASR system, BRDialect, developed by fine-tuning the IndicWav2Vec model in ten Bengali regional dialects. It outperforms the baseline ASR models by 12.41-33.98% on the RegSpeech12 dataset. Furthermore, BanglaTalk can operate at a low bandwidth of 24 kbps while maintaining an average end-to-end delay of 4.9 seconds. Low bandwidth usage and minimal end-to-end delay make the system both cost-effective and interactive for real-time use cases, enabling inclusive and accessible speech technology for the diverse community of Bengali speakers.http://arxiv.org/abs/2510.0618
B-TGAT: A Bi-directional Temporal Graph Attention Transformer for Clustering Multivariate Spatiotemporal Data
No full textClustering high-dimensional multivariate spatiotemporal climate data is challenging due to complex temporal dependencies, evolving spatial interactions, and non-stationary dynamics. Conventional clustering methods, including recurrent and convolutional models, often struggle to capture both local and global temporal relationships while preserving spatial context. We present a time-distributed hybrid U-Net autoencoder that integrates a Bi-directional Temporal Graph Attention Transformer (B-TGAT) to guide efficient temporal clustering of multidimensional spatiotemporal climate datasets. The encoder and decoder are equipped with ConvLSTM2D modules that extract joint spatial--temporal features by modeling localized dynamics and spatial correlations over time, and skip connections that preserve multiscale spatial details during feature compression and reconstruction. At the bottleneck, B-TGAT integrates graph-based spatial modeling with attention-driven temporal encoding, enabling adaptive weighting of temporal neighbors and capturing both short and long-range dependencies across regions. This architecture produces discriminative latent embeddings optimized for clustering. Experiments on three distinct spatiotemporal climate datasets demonstrate superior cluster separability, temporal stability, and alignment with known climate transitions compared to state-of-the-art baselines. The integration of ConvLSTM2D, U-Net skip connections, and B-TGAT enhances temporal clustering performance while providing interpretable insights into complex spatiotemporal variability, advancing both methodological development and climate science applications.This work is supported by HDR Institute: HARP - Harnessing Data and Model Revolution in the Polar Regions (OAC-2118285).http://arxiv.org/abs/2509.1320
Combined dark matter search towards dwarf spheroidal galaxies with Fermi-LAT, HAWC, H.E.S.S., MAGIC, and VERITAS
No full textDwarf spheroidal galaxies (dSphs) are excellent targets for indirect dark matter (DM) searches using gamma-ray telescopes because they are thought to have high DM content and a low astrophysical background. The sensitivity of these searches is improved by combining the observations of dSphs made by different gamma-ray telescopes. We present the results of a combined search by the most sensitive currently operating gamma-ray telescopes, namely: the satellite-borne Fermi-LAT telescope; the ground-based imaging atmospheric Cherenkov telescope arrays H.E.S.S., MAGIC, and VERITAS; and the HAWC water Cherenkov detector. Individual datasets were analyzed using a common statistical approach. Results were subsequently combined via a global joint likelihood analysis. We obtain constraints on the velocity-weighted cross section ⟨σv⟩ for DM self-annihilation as a function of the DM particle mass. This five-instrument combination allows the derivation of up to 2-3 times more constraining upper limits on ⟨σv⟩ than the individual results over a wide mass range spanning from 5 GeV to 100 TeV. Depending on the DM content modeling, the 95% confidence level observed limits reach 1.5X10⁻²⁴cm³ s⁻¹ and 3.2X10⁻²⁵cm³ s⁻¹, respectively, in the τ⁺τ⁻ annihilation channel for a DM mass of 2 TeV.The Fermi-LAT Collaboration acknowledges generous ongoing support from a number of agencies and institutes that have supported both the development and the operation of the LAT as well as scientific data analysis. These include the National Aeronautics and Space Administration and the Department of Energy in the United States, the Commissariat à l’Energie Atomique and the Centre National de la Recherche Scientifique / Institut National de Physique Nucléaire et de Physique des Particules in France, the Agenzia Spaziale Italiana and the Istituto Nazionale di Fisica Nucleare in Italy, the Ministry of Education, Culture, Sports, Science and Technology (MEXT), High Energy Accelerator Research Organization (KEK) and Japan Aerospace Exploration Agency(JAXA) in Japan, and the K. A. Wallenberg Foundation, the Swedish Research Council and the Swedish National Space Board in Sweden. Additional support for science analysis during the operations phase is gratefully acknowledged from the Istituto Nazionale di Astrofisica in Italy and the Centre National d’Études Spatiales in France. This work performed in part under DOE Contract DE-AC02-76SF00515. The HAWC collaboration acknowledges the support from: the US National Science Foundation (NSF); the US Department of Energy Office of High-Energy Physics; the Laboratory Directed Research and Development (LDRD) program of Los Alamos National Laboratory; Consejo Nacional de Ciencia y Tecnología (CONACyT), México, grants LNC-2023-117, 271051, 232656, 260378, 179588, 254964, 258865, 243290, 132197, A1-S-46288, A1-S-22784, CF-2023-I-645, cátedras 873, 1563, 341, 323, Red HAWC, México; DGAPA-UNAM grants IG101323, IN111716-3, IN111419, IA102019, IN106521, IN114924, IN110521 , IN102223; VIEP-BUAP; PIFI 2012, 2013, PROFOCIE 2014, 2015; the University of Wisconsin Alumni Research Foundation; the Institute of Geophysics, Planetary Physics, and Signatures at Los Alamos National Laboratory; Polish Science Centre grant, 2024/53/B/ST9/02671; Coordinación de la Investigación Científica de la Universidad Michoacana; Royal Society - Newton Advanced Fellowship 180385; Gobierno de España and European Union - NextGenerationEU, grant CNS2023- 144099; The Program Management Unit for Human Resources & Institutional Development, Research and Innovation, NXPO (grant number B16F630069); Coordinación General Académica e Innovación (CGAI-UdeG), PRODEP-SEP UDG-CA-499; Institute of Cosmic Ray Research (ICRR), University of Tokyo. H.F. acknowledges support by NASA under award number 80GSFC21M0002. C.R. acknowledges support from National Research Foundation of Korea (RS-2023-00280210). We also acknowledge the significant contributions over many years of Stefan Westerhoff, Gaurang Yodh and Arnulfo Zepeda Domínguez, all deceased members of the HAWC collaboration. Thanks to Scott Delay, Luciano Díaz and Eduardo Murrieta for technical support. The support of the Namibian authorities and of the University of Namibia in facilitating the construction and operation of H.E.S.S. is gratefully acknowledged, as is the support by the German Ministry for Education and Research (BMBF), the Max Planck Society, the German Research Foundation (DFG), the Helmholtz Association, the Alexander von Humboldt Foundation, the French Ministry of Higher Education, Research and Innovation, the Centre National de la Recherche Scientifique (CNRS/IN2P3 and CNRS/INSU), the Commissariat à l’énergie atomique et aux énergies alternatives (CEA), the U.K. Science and Technology Facilities Council (STFC), the Irish Research Council (IRC) and the Science Foundation Ireland (SFI), the Knut and Alice Wallenberg Foundation, the Polish Ministry of Education and Science, agreement no. 2021/WK/06, the South African Department of Science and Technology and National Research Foundation, the University of Namibia, the National Commission on Research, Science & Technology of Namibia (NCRST), the Austrian Federal Ministry of Education, Science and Research and the Austrian Science Fund (FWF), the Australian Research Council (ARC), the Japan Society for the Promotion of Science, the University of Amsterdam and the Science Committee of Armenia grant 21AG-1C085. We appreciate the excellent work of the technical support staff in Berlin, Zeuthen, Heidelberg, Palaiseau, Paris, Saclay, Tübingen and in Namibia in the construction and operation of the equipment. This work benefited from services provided by the H.E.S.S. Virtual Organisation, supported by the national resource providers of the EGI Federation. The MAGIC collaboration would like to thank the Instituto de Astrofísica de Canarias for the excellent working conditions at the Observatorio del Roque de los Muchachos in La Palma. The financial support of the German BMBF, MPG and HGF; the Italian INFN and INAF; the Swiss National Fund SNF; the grants PID2019-104114RB-C31, PID2019-104114RB-C32, PID2019-104114RB-C33, PID2019-105510GB-C31, PID2019- 107847RB-C41, PID2019-107847RB-C42, PID2019-107847RB-C44, PID2019-107988GBC22, PID2022-136828NB-C41, PID2022-137810NB-C22, PID2022-138172NB-C41, PID2022- 138172NB-C42, PID2022-138172NB-C43, PID2022-139117NB-C41, PID2022-139117NBC42, PID2022-139117NB-C43, PID2022-139117NB-C44 funded by the Spanish MCIN/AEI/ 10.13039/501100011033 and “ERDF A way of making Europe”; the Indian Department of Atomic Energy; the Japanese ICRR, the University of Tokyo, JSPS, and MEXT; the Bulgarian Ministry of Education and Science, National RI Roadmap Project DO1- 400/18.12.2020 and the Academy of Finland grant nr. 320045 is gratefully acknowledged. This work was also been supported by Centros de Excelencia “Severo Ochoa” y Unidades “María de Maeztu” program of the Spanish MCIN/AEI/ 10.13039/501100011033 (CEX2019- 000920-S, CEX2019-000918-M, CEX2021-001131-S) and by the CERCA institution and grants 2021SGR00426 and 2021SGR00773 of the Generalitat de Catalunya; by the Croatian Science Foundation (HrZZ) Project IP-2022-10-4595 and the University of Rijeka Project uniri-prirod-18-48; by the Deutsche Forschungsgemeinschaft (SFB1491) and by the Lamarr-Institute for Machine Learning and Artificial Intelligence; by the Polish Ministry Of Education and Science grant No. 2021/WK/08; and by the Brazilian MCTIC, CNPq and FAPERJ. This work was supported by the Grant RYC2021-032552-I funded by MCIN/AEI/10.13039/501100011033 and by the European Union NextGenerationEU/PRTR. The research of VERITAS is supported by grants from the U.S. Department of Energy Office of Science, the U.S. National Science Foundation and the Smithsonian Institution, by NSERC in Canada, and by the Helmholtz Association (including the Young Investigators Program of the Helmholtz Association) in Germany. This research used resources provided by the Open Science Grid, which is supported by the National Science Foundation and the U.S. Department of Energy’s Office of Science, and resources of the National Energy Research Scientific Computing Center (NERSC), a U.S. Department of Energy Office of Science User Facility operated under Contract No. DE-AC02- 05CH11231. We acknowledge the excellent work of the technical support staff at the Fred Lawrence Whipple Observatory and at the collaborating institutions in the construction and operation of the instrument.http://arxiv.org/abs/2508.2022
A Comprehensive Machine Learning Approach for Email and URL Threat Detection Using Feature Importance Analysis
No full textPhishing is the most prevalent form of cybercrime, where individuals are convinced to disclose sensitive details like account IDs, passwords, and banking information. These cyberattacks are often initiated through emails, instant messaging, and phone calls. The primary concern today revolves around the security of devices, computers, and software. This study presents the development of a website designed to scan incoming emails and attachments for potential viruses and security threats. This website includes validation attachment scanning, URL scanning, and IP address scanning. Integration with the VirusTotal database will be carried out to assess the safety of websites. Furthermore, the study incorporates machine learning algorithms to enhance phishing detection, ultimately mitigating risks and occurrences. The dataset utilized comprises diverse sources containing both regular and phishing emails, along with numerous attributes for identifying malicious emails and harmful URL links, some of which are sourced from VirusTotal. The outcomes of the experiments reveal promising levels of accuracy in identifying phishing attacks, underscoring the efficiency of machine learning as a vital component in enhancing email security. The study also addresses the obstacles and constraints faced by the proposed models, highlighting the evolving nature of phishing strategies and the necessity for continual model adaptation.https://www.i-csrs.org/Volumes/ijasca/2025.2.16.pd
SPATIAL ANALYSIS OF STREAM TEMPERATURE RESPONSE TO PATTERNS OF LAND COVER AND STORMWATER INFRASTRUCTURE IN AN URBAN WATERSHED
No full textThis study examined effects of summer storm runoff on stream temperature in the 47% impervious Dead Run watershed, Baltimore County, Maryland. Five-minute temperature data spanning 2.5 years from 204 sensors spaced 50-100 m apart were analyzed to identify spatial patterns of stream thermal response. Significant differences in temperature change between paired sensors occurred predominantly in headwater areas with drainage areas smaller than 1.5 km², revealing a clear spatial sensitivity threshold. Warming effects were associated with large drainage increments added between sensors near roadways and pipes, while cooling effects were associated with lower percent impervious cover. Differences in temperature impacts associated with inflow from stormwater management facilities were not significant, and differences in impervious cover were too small to cause significant temperature variation. Drainage area size proved to be the strongest predictor of thermal response. These findings highlight the utility of spatially targeted analyses to better understand stormwater thermal impacts
VERITAS Follow-Up Observations of the Ultra-High-Energy Neutrino Event KM3-230213A
No full text39th International Cosmic Ray Conference (ICRC2025), Geneva, Switzerland, July 2025Authors: VERITAS Collaboration, A. Archer , P. Bangale , J. T. Bartkoske , W. Benbow , Y. Chen , J. L. Christiansen , A. J. Chromey , A. Duerr , M. Errando , M. Escobar Godoy , J. Escudero Pedrosa , Q. Feng , S. Filbert , L. Fortson , A. Furniss , W. Hanlon , O. Hervet , C. E. Hinrichs, , J. Holder, T. B. Humensky,, M. Iskakova , W. Jin , M. N. Johnson , E. Joshi, M. Kertzman , M. Kherlakian, D. Kieda , T. K. Kleiner, N. Korzoun, S. Kumar, M. J. Lang, M. Lundy, G. Maier, C. E McGrath, P. Moriarty, R. Mukherjee , W. Ning , R. A. Ong , A. Pandey , M. Pohl,, E. Pueschel, J. Quinn, P. L. Rabinowitz , K. Ragan, P. T. Reynolds, D. Ribeiro , E. Roache , I. Sadeh, L. Saha , H. Salzmann , M. Santander, G. H. Sembroski, B. Shen, M. Splettstoesser , A. K. Talluri , S. Tandon, J. V. Tucci, J. Valverde,, V. V. Vassiliev , D. A. Williams , S. L. Wong, T. YoshikoshiThe recent announcement of the detection of the ultra-high-energy (UHE) neutrino event KM3-230213A by the KM3NeT telescope represents a critical opportunity to explore the origins of cosmic neutrinos and their potential gamma-ray counterparts. With an inferred neutrino energy exceeding 100 PeV, this event stands as the most energetic neutrino observed to date. The large offset from the galactic plane (11 degrees) and the presence of several blazars with temporally correlated multiwavelength counterparts within the 3 degrees localization region raise the possibility of an extragalactic origin. Additionally, the event's apparent tension with IceCube constraints suggests that it could be transient in nature rather than cosmogenic. VERITAS conducted a targeted follow-up campaign to search for very-high-energy (VHE, greater than 100 GeV) gamma-ray emission associated with KM3-230213A. Observations were performed in February and March 2025, using a four-point wobble strategy centered on the best-fit neutrino position, covering nearly the entire 90 percent confidence region. These observations probe potential hadronic gamma-ray emission from a common origin with the neutrino, placing constraints on particle-acceleration scenarios. We present the results of this search, including upper limits on very-high-energy gamma-ray flux and their implications for possible source models of KM3-230213A.This research is supported by grants from the U.S. Department of Energy Office of Science, the U.S. National Science Foundation and the Smithsonian Institution, by NSERC in Canada, and by the Helmholtz Association in Germany. This research used resources provided by the Open Science Grid, which is supported by the National Science Foundation and the U.S. Department of Energy’s Office of Science, and resources of the National Energy Research Scientific Computing Center (NERSC), a U.S. Department of Energy Office of Science User Facility operated under Contract No. DE-AC02-05CH11231. We acknowledge the excellent work of the technical support staff at the Fred Lawrence Whipple Observatory and at the collaborating institutions in the construction and operation of the instrument.http://arxiv.org/abs/2509.2542
Visual Bias and Interpretability in Deep Learning for Dermatological Image Analysis
No full text4th IEEE International Conference on Image Processing and Media Computing (ICIPMC) 2025, Xi'an, China, June 27-29, 2025.Accurate skin disease classification is a critical yet challenging task due to high inter-class similarity, intra-class variability, and complex lesion textures. While deep learning-based computer-aided diagnosis (CAD) systems have shown promise in automating dermatological assessments, their performance is highly dependent on image pre-processing and model architecture. This study proposes a deep learning framework for multi-class skin disease classification, systematically evaluating three image pre-processing techniques: standard RGB, CMY color space transformation, and Contrast Limited Adaptive Histogram Equalization (CLAHE). We benchmark the performance of pre-trained convolutional neural networks (DenseNet201, Efficient-NetB5) and transformer-based models (ViT, Swin Transformer, DinoV2 Large) using accuracy and F1-score as evaluation metrics. Results show that DinoV2 with RGB pre-processing achieves the highest accuracy (up to 93%) and F1-scores across all variants. Grad-CAM visualizations applied to RGB inputs further reveal precise lesion localization, enhancing interpretability. These findings underscore the importance of effective pre-processing and model choice in building robust and explainable CAD systems for dermatology.http://arxiv.org/abs/2508.0457