arXiv subjects

Mohammad Hosseini

Publications and source records attributed to Mohammad Hosseini.

At least 19 recordsLinked to original sources

arXiv2026

A Low-Power Wearable Respiratory Sensor for Non-Invasive Stress Monitoring

Respiration provides a continuously available window into physiological state and behavior. However, monitoring it outside controlled settings remains challenging because a wearable system must capture small body deformations while remaining comfortable, low power, and robust to changes in posture and motion. We present a compact non-invasive respiratory sensing system based on a force-sensitive resistor (FSR) embedded in an abdominal belt and integrated with a custom Bluetooth Low Energy acquisition board. The system combines a simple piezoresistive readout with a mechanical holder designed to transfer abdominal expansion to the sensor without analog amplification. We evaluate the complete sensing pipeline across multiple breathing patterns and body positions. In stationary settings, the recorded signals exhibit consistent amplitude changes and recurring peak-to-peak timing across breathing maneuvers; under light movement, these variations remain visible despite motion-induced baseline shifts. We further design a five-phase stress-induction protocol and collect respiratory recordings from 12 participants. Using interpretable time-domain features and standard classifiers, we examine whether the acquired signals distinguish relaxation from stress-induction phases. In this preliminary experiment, the best-performing model achieves 88.0% test accuracy, indicating that the extracted respiratory features distinguish stress-induced phases from relaxation phases in this dataset. Overall, our results show that the proposed platform enables real-time respiratory monitoring across diverse daily-life scenarios and captures respiratory changes that distinguish stress-induction from relaxation phases, supporting its potential for affective-computing applications.

q-bio.QM↗

arXiv2026

Cross-Subject Modeling for Widefield Calcium Imaging via Atlas-Aligned Spatiotemporal Tokenization

Large-scale, multi-subject widefield calcium imaging provides unprecedented access to brain-wide cortical dynamics. However, the high dimensionality, complex spatiotemporal structure, and substantial task-irrelevant activity in widefield recordings have largely restricted modeling efforts to single-session analyses, limiting scalability and generalization. While multi-subject pretrained models have been explored for some neural modalities, multi-subject models for widefield calcium imaging have not yet been demonstrated; further, subject-invariant zero-shot behavior decoding remains elusive for multi-subject models across neural modalities more broadly. As a first step toward foundation modeling of widefield data, we introduce WiCAT, a multi-subject model that leverages self-supervised pretraining to both outperform single-session models and enable zero-shot behavior decoding on unseen subjects. WiCAT introduces an atlas-grounded tokenization scheme without session-specific components and learns globally shared spatiotemporal representations. Across multiple widefield datasets, the pretrained model supports lightweight downstream decoding, transfers across subjects, tasks, and datasets, and outperforms baseline models. Notably, the model also achieves robust zero-shot continuous behavior decoding and left-out brain region reconstruction on unseen subjects.

cs.CV↗

arXiv2026

Machine learning reconstruction of digit bone Raman spectra enables noninvasive transcutaneous detection of systemic osteoporosis

Osteoporosis, a major global epidemic, often goes undetected until a fracture occurs, largely due to poor access to screening using gold standard methods, such as dual-energy X-ray absorptiometry (DXA). As a potential nonionizing radiation alternative, we present a transcutaneous spatially offset Raman spectroscopy (SORS) approach combined with machine learning (ML) to recover bone spectra through overlying soft tissue and extract diagnostic information. In a human cadaveric study spanning normal, osteopenic, and osteoporotic donors, we acquired paired Raman measurements from transcutaneous fingers at multiple spatial offsets (0, 3, and 6 mm) and from the corresponding exposed finger bones. Using this paired dataset, supervised machine-learning models were trained to reconstruct exposed-bone Raman spectra from transcutaneous measurements, enabling direct recovery of bone biochemical signatures from transcutaneous tissue. The ML predicted bone spectra preserved physiologically meaningful Raman features and demonstrated statistically significant differences between normal and osteoporotic groups across four key Raman-derived metrics (p < 0.05), representing, to our knowledge, the first demonstration of transcutaneous Raman discrimination between clinically established bone-health categories in a human cadaveric study. The ML-predicted spectra further correlated with distal-radius DXA T-scores (r = 0.73, RMSECV = 1.4), approaching the performance achieved using exposed-bone measurements (r = 0.9, RMSECV = 0.8). Finally, preliminary in vivo measurements from two volunteers revealed clear bone-related transcutaneous spectral features consistent with cadaveric data, supporting translational feasibility. Together, these results establish a foundation for nonionizing radiation, transcutaneous Raman assessment of bone health using supervised spectral extraction from accessible measurement sites

physics.med-ph↗

arXiv2025

Predicting Wrist Osteoporosis from excised human finger bones using spatially offset Raman spectroscopy, A Cadaveric Study

Osteoporosis and osteopenia remain vastly underdiagnosed. Current clinical screening relies almost exclusively on dual-energy X-ray absorptiometry (DXA), which measures bone mineral density (BMD) but fails to capture the compositional changes that lead to BMD loss. We investigated whether Spatially Offset Raman Spectroscopy (SORS) applied to excised finger bones can assess subsurface biochemical markers capable of diagnosing osteoporosis and osteopenia and predicting wrist DXA T-scores. Raman spectra were acquired ex vivo on the mid-shaft of the proximal phalanx of the second digit from 25 female cadavers spanning the three T-score categories (n=8 normal, n=6 osteopenic, and n=11 osteoporotic) at spatial offsets of 0, 3, and 6 mm from a laser irradiation spot. After normalizing spectra to the PO43- peak, group-averaged spectra of the three categories, measured at 3-mm offset, showed clear differences in the CO32-, Amide III, CH2, and Amide I bands. Quantitatively, four out of five mineral-to-matrix ratios differed significantly (p < 0.05) between normal and osteopenic bone, and between osteopenic and osteoporotic bone, and all five ratios showed significant differences between normal and osteoporotic bone. In contrast, the 0-mm offset suffered diminished contrast, and the 6-mm offset did not enhance discrimination between different groups, compared with the 3-mm offset. A leave-one-out, partial-least-squares regression model built from the 3-mm spectra predicted distal radius DXA T-score with a Pearson correlation of r = 0.85 and a root-mean-square error of cross-validation of 1 T-score units, correctly classifying 92% of specimens.

physics.med-ph↗

arXiv2025

PerHalluEval: Persian Hallucination Evaluation Benchmark for Large Language Models

Hallucination is a persistent issue affecting all large language Models (LLMs), particularly within low-resource languages such as Persian. PerHalluEval (Persian Hallucination Evaluation) is the first dynamic hallucination evaluation benchmark tailored for the Persian language. Our benchmark leverages a three-stage LLM-driven pipeline, augmented with human validation, to generate plausible answers and summaries regarding QA and summarization tasks, focusing on detecting extrinsic and intrinsic hallucinations. Moreover, we used the log probabilities of generated tokens to select the most believable hallucinated instances. In addition, we engaged human annotators to highlight Persian-specific contexts in the QA dataset in order to evaluate LLMs' performance on content specifically related to Persian culture. Our evaluation of 12 LLMs, including open- and closed-source models using PerHalluEval, revealed that the models generally struggle in detecting hallucinated Persian text. We showed that providing external knowledge, i.e., the original document for the summarization task, could mitigate hallucination partially. Furthermore, there was no significant difference in terms of hallucination when comparing LLMs specifically trained for Persian with others.

cs.CL↗

arXiv2025

Dynamical Modeling of Behaviorally Relevant Spatiotemporal Patterns in Neural Imaging Data

High-dimensional imaging of neural activity, such as widefield calcium and functional ultrasound imaging, provide a rich source of information for understanding the relationship between brain activity and behavior. Accurately modeling neural dynamics in these modalities is crucial for understanding this relationship but is hindered by the high-dimensionality, complex spatiotemporal dependencies, and prevalent behaviorally irrelevant dynamics in these modalities. Existing dynamical models often employ preprocessing steps to obtain low-dimensional representations from neural image modalities. However, this process can discard behaviorally relevant information and miss spatiotemporal structure. We propose SBIND, a novel data-driven deep learning framework to model spatiotemporal dependencies in neural images and disentangle their behaviorally relevant dynamics from other neural dynamics. We validate SBIND on widefield imaging datasets, and show its extension to functional ultrasound imaging, a recent modality whose dynamical modeling has largely remained unexplored. We find that our model effectively identifies both local and long-range spatial dependencies across the brain while also dissociating behaviorally relevant neural dynamics. Doing so, SBIND outperforms existing models in neural-behavioral prediction. Overall, SBIND provides a versatile tool for investigating the neural mechanisms underlying behavior using imaging modalities.

q-bio.NC↗

arXiv2025

Revolutionizing Traffic Management with AI-Powered Machine Vision: A Step Toward Smart Cities

The rapid urbanization of cities and increasing vehicular congestion have posed significant challenges to traffic management and safety. This study explores the transformative potential of artificial intelligence (AI) and machine vision technologies in revolutionizing traffic systems. By leveraging advanced surveillance cameras and deep learning algorithms, this research proposes a system for real-time detection of vehicles, traffic anomalies, and driver behaviors. The system integrates geospatial and weather data to adapt dynamically to environmental conditions, ensuring robust performance in diverse scenarios. Using YOLOv8 and YOLOv11 models, the study achieves high accuracy in vehicle detection and anomaly recognition, optimizing traffic flow and enhancing road safety. These findings contribute to the development of intelligent traffic management solutions and align with the vision of creating smart cities with sustainable and efficient urban infrastructure.

cs.CV↗

arXiv2025

A Survey of Data Compression Algorithms and their Applications

Today, with the growing demands of information storage and data transfer, data compression is becoming increasingly important. Data Compression is a technique which is used to decrease the size of data. This is very useful when some huge files have to be transferred over networks or being stored on a data storage device and the size is more than the capacity of the data storage or would consume so much bandwidth for transmission in a network. With the advent of the Internet and mobile devices with limited resources, data compression has gained even more importance. It can be effectively used to save both storage and bandwidth, thus to decrease download duration. Data compression can be achieved by a host of techniques. During this survey, I'm going to thoroughly discuss some of important data compression algorithms, their performance evaluation, and their major applications along with today's issues and recent research approaches.

cs.IT↗

arXiv2024

GREI Data Repository AI Taxonomy

The Generalist Repository Ecosystem Initiative (GREI), funded by the NIH, developed an AI taxonomy tailored to data repository roles to guide AI integration across repository management. It categorizes the roles into stages, including acquisition, validation, organization, enhancement, analysis, sharing, and user support, providing a structured framework for implementing AI in repository workflows.

cs.DL↗

arXiv2023

Faster, Lighter, More Accurate: A Deep Learning Ensemble for Content Moderation

To address the increasing need for efficient and accurate content moderation, we propose an efficient and lightweight deep classification ensemble structure. Our approach is based on a combination of simple visual features, designed for high-accuracy classification of violent content with low false positives. Our ensemble architecture utilizes a set of lightweight models with narrowed-down color features, and we apply it to both images and videos. We evaluated our approach using a large dataset of explosion and blast contents and compared its performance to popular deep learning models such as ResNet-50. Our evaluation results demonstrate significant improvements in prediction accuracy, while benefiting from 7.64x faster inference and lower computation cost. While our approach is tailored to explosion detection, it can be applied to other similar content moderation and violence detection use cases as well. Based on our experiments, we propose a "think small, think many" philosophy in classification scenarios. We argue that transforming a single, large, monolithic deep model into a verification-based step model ensemble of multiple small, simple, and lightweight models with narrowed-down visual features can possibly lead to predictions with higher accuracy.

cs.CV↗

arXiv2023

Ultra-fast yttrium hydride chemistry at high pressures via non-equilibrium states induced by x-ray free electron laser

Controlling the formation and stoichiometric content of desired phases of materials has become a central interest for the study of a variety of fields, notably high temperature superconductivity under extreme pressures. The further possibility of accessing metastable states by initiating reactions by x-ray triggered mechanisms over ultra-short timescales is enabled with the development of x-ray free electron lasers (XFEL). Utilizing the exceptionally high brilliance x-ray pulses from the EuXFEL, we report the synthesis of a previously unobserved yttrium hydride under high pressure, along with non-stoichiometric changes in hydrogen content as probed at a repetition rate of 4.5\,MHz using time-resolved x-ray diffraction. Exploiting non-equilibrium pathways we synthesize and characterize a hydride with yttrium cations in an \textit{A}15 structure type at 125\,GPa, predicted using crystal structure searches, with a hydrogen content between 4.0--5.75 hydrogens per cation, that is enthalpically metastable on the convex hull. We demonstrate a tailored approach to changing hydrogen content using changes in x-ray fluence that is not accessible using conventional synthesis methods, and reveals a new paradigm in metastable chemical physics.

cond-mat.mtrl-sci↗

arXiv2023

Morpheus: Extending the Last Level Cache Capacity in GPU Systems Using Idle GPU Core Resources

Graphics Processing Units (GPUs) are widely-used accelerators for data-parallel applications. In many GPU applications, GPU memory bandwidth bottlenecks performance, causing underutilization of GPU cores. Hence, disabling many cores does not affect the performance of memory-bound workloads. While simply power-gating unused GPU cores would save energy, prior works attempt to better utilize GPU cores for other applications (ideally compute-bound), which increases the GPU's total throughput. In this paper, we introduce Morpheus, a new hardware/software co-designed technique to boost the performance of memory-bound applications. The key idea of Morpheus is to exploit unused core resources to extend the GPU last level cache (LLC) capacity. In Morpheus, each GPU core has two execution modes: compute mode and cache mode. Cores in compute mode operate conventionally and run application threads. However, for the cores in cache mode, Morpheus invokes a software helper kernel that uses the cores' on-chip memories (i.e., register file, shared memory, and L1) in a way that extends the LLC capacity for a running memory-bound workload. Morpheus adds a controller to the GPU hardware to forward LLC requests to either the conventional LLC (managed by hardware) or the extended LLC (managed by the helper kernel). Our experimental results show that Morpheus improves the performance and energy efficiency of a baseline GPU architecture by an average of 39% and 58%, respectively, across several memory-bound workloads. Morpheus' performance is within 3% of a GPU design that has a quadruple-sized conventional LLC. Morpheus can thus contribute to reducing the hardware dedicated to a conventional LLC by exploiting idle cores' on-chip memory resources as additional cache capacity.

cs.AR↗

arXiv2022

Deep Live Video Ad Placement on the 5G Edge

The video broadcasting industry has been growing significantly in the recent years, specially on delivering personalized contents to the end users. While video broadcasting has continued to grow beyond TV, video adverting has become a key marketing tool to deliver targeted messages directly to the audience. However, unfortunately for broadband TV, a key problem is that the TV commercials target the broad audience, therefore lacking user-specific and personalized ad contents. In this paper, we propose a deep edge-cloud ad-placement system, and briefly describe our methodologies and the architecture of our designed ad placement system for delivering both the Video on Demand (VoD) and live broadcast TV contents over MMT streaming protocol. The aim of our paper is to showcase how to enable targeted, personalized, and user-specific advertising services deployed on the future 5G MEC platforms, which in turn can have high potentials to increase ad revenues for the mobile operator industry.

cs.MM↗

arXiv2021

The Case for High-Accuracy Classification: Think Small, Think Many!

To facilitate implementation of high-accuracy deep neural networks especially on resource-constrained devices, maintaining low computation requirements is crucial. Using very deep models for classification purposes not only decreases the neural network training speed and increases the inference time, but also need more data for higher prediction accuracy and to mitigate false positives. In this paper, we propose an efficient and lightweight deep classification ensemble structure based on a combination of simple color features, which is particularly designed for "high-accuracy" image classifications with low false positives. We designed, implemented, and evaluated our approach for explosion detection use-case applied to images and videos. Our evaluation results based on a large test test show considerable improvements on the prediction accuracy compared to the popular ResNet-50 model, while benefiting from 7.64x faster inference and lower computation cost. While we applied our approach to explosion detection, our approach is general and can be applied to other similar classification use cases as well. Given the insight gained from our experiments, we hence propose a "think small, think many" philosophy in classification scenarios: that transforming a single, large, monolithic deep model into a verification-based step model ensemble of multiple small, simple, lightweight models with narrowed-down color spaces can possibly lead to predictions with higher accuracy.

cs.CV↗

arXiv2020

Large signal analysis of multiple quantum well transistor lasers: Investigation of imbalanced carrier and photon density distribution

In this paper, we present a large signal and switching analysis for the Heterojunction Bipolar Transistor Laser (TL) to reveal its optical and electrical behavior under high current injection conditions. Utilizing appropriate models for carrier transport, nonlinear optical gain and optical confinement factor (OCF), we have simulated the large signal response of the HBTL in relatively low and high modulation frequencies. Our results predict that for multiple quantum well structures at low frequencies there should not be a difference in carrier density either the photon density. However, carrier concentration can be differently distributed between subsequent wells in case of a high speed yet large signal input. This leads to increased linewidth instead as it depends on ΔNqw. We show the effect of different structural parameters on the switching behavior by performing a switching analysis of the Single Quantum Well (SQW) and Multiple Quantum Well (MQW) structures using computationally efficient numerical methods. A set of coupled rate equations is solved to investigate the large-signal and switching behavior of MQW-HBTL. Finally, to have a comprehensive judgment about this optoelectronic device, we introduce a relative performance factor to taking into account all the optoelectronic characteristics such as output power, ac current gain, modulation bandwidth, and base threshold current as well as turn-on time in order to design a suitable TL for Opto-Electronic Integrated Circuits (OEICs).

physics.app-ph↗

arXiv2019

Adaptive Rate Allocation for View-Aware Point-Cloud Streaming

In the context of view-dependent point-cloud streaming in a scene, our rate allocation is "adaptive" in the sense that it priorities the point-cloud models depending on the camera view and the visibility of the objects and their distance as described. The algorithm delivers higher bitrate to the point-cloud models which are inside user's viewport, more likely for the user to look at, or are closer to the view camera or, while delivers lower quality level to the point-cloud models outside of a user's immediate viewport or farther away from the camera. For that purpose, we hereby explain the rate allocation problem within the context of multi-point-cloud streaming where multiple point-cloud models are aimed to be streamed to the target device, and propose a rate allocation heuristic algorithm to enable the adaptations within this context. To the best of our knowledge, this is the first work to mathematically model, and propose a rate allocation heuristic algorithm within the context of point-cloud streaming.

cs.MM↗

arXiv2019

R-Storm: Resource-Aware Scheduling in Storm

The era of big data has led to the emergence of new systems for real-time distributed stream processing, e.g., Apache Storm is one of the most popular stream processing systems in industry today. However, Storm, like many other stream processing systems lacks an intelligent scheduling mechanism. The default round-robin scheduling currently deployed in Storm disregards resource demands and availability, and can therefore be inefficient at times. We present R-Storm (Resource-Aware Storm), a system that implements resource-aware scheduling within Storm. R-Storm is designed to increase overall throughput by maximizing resource utilization while minimizing network latency. When scheduling tasks, R-Storm can satisfy both soft and hard resource constraints as well as minimizing network distance between components that communicate with each other. We evaluate R-Storm on set of micro-benchmark Storm applications as well as Storm applications used in production at Yahoo! Inc. From our experimental results we conclude that R-Storm achieves 30-47% higher throughput and 69-350% better CPU utilization than default Storm for the micro-benchmarks. For the Yahoo! Storm applications, R-Storm outperforms default Storm by around 50% based on overall throughput. We also demonstrate that R-Storm performs much better when scheduling multiple Storm applications than default Storm.

cs.DC↗

arXiv2019

Dynamic Adaptive Point Cloud Streaming

High-quality point clouds have recently gained interest as an emerging form of representing immersive 3D graphics. Unfortunately, these 3D media are bulky and severely bandwidth intensive, which makes it difficult for streaming to resource-limited and mobile devices. This has called researchers to propose efficient and adaptive approaches for streaming of high-quality point clouds. In this paper, we run a pilot study towards dynamic adaptive point cloud streaming, and extend the concept of dynamic adaptive streaming over HTTP (DASH) towards DASH-PC, a dynamic adaptive bandwidth-efficient and view-aware point cloud streaming system. DASH-PC can tackle the huge bandwidth demands of dense point cloud streaming while at the same time can semantically link to human visual acuity to maintain high visual quality when needed. In order to describe the various quality representations, we propose multiple thinning approaches to spatially sub-sample point clouds in the 3D space, and design a DASH Media Presentation Description manifest specific for point cloud streaming. Our initial evaluations show that we can achieve significant bandwidth and performance improvement on dense point cloud streaming with minor negative quality impacts compared to the baseline scenario when no adaptations is applied.

cs.MM↗