arXiv subjects

Mehedi Hasan

Publications and source records attributed to Mehedi Hasan.

At least 19 recordsLinked to original sources

arXiv2026

TAP-Path: Task-Adaptive Structural and Token Pruning for Efficient and Trustworthy Pathology Foundation Models

Pathology foundation models improve transferable representation learning for histopathology, but recent gains often rely on encoders with hundreds of millions of parameters and high inference cost. We propose TAP-Path, a task-adaptive compression framework that directly restructures a pretrained Virchow2 encoder rather than distilling it into a separate student. TAP-Path combines validation-driven transformer-block selection, physical removal of redundant blocks, input-adaptive patch-token pruning, multi-depth feature recovery, and a lightweight gated task head. The final model retains 24 of 32 transformer blocks and 70% of patch tokens after pruning, reducing encoder parameters by 24.96% (631.24M to 473.70M) and analytical encoder compute by 35.20% (340.13G to 220.40G FLOPs). Across three task-head optimization seeds, TAP-Path achieved $87.98 \pm 0.067%$ test accuracy, $81.26 \pm 0.49%$ balanced accuracy, and $82.38 \pm 0.48%$ macro-F1 on a 32-class histopathology benchmark, compared with 86.89% for full Virchow2 and 87.67% for UNI2-h. TAP-Path achieved a Brier score of $0.1800 \pm 0.0005$ and failure-detection AUROC of $0.9047 \pm 0.0060$. A validation-only rare-aware objective improved rare-class balanced accuracy in a secondary operating analysis. Frozen external evaluation on 433 CPTAC samples yielded $91.22 \pm 0.83%$ accuracy and $91.10 \pm 0.81%$ balanced accuracy. These results show that task-adaptive structural and token sparsification can improve the accuracy-efficiency trade-off of large pathology foundation models while preserving reliability under internal and external evaluation.

cs.CV↗

arXiv2026

Linear Multi-Timescale Retention as a Memory-Efficient Vision-Language Bridge

Vision-Language Models (VLMs) face a critical computational bottleneck when processing high-resolution imagery due to the $O(N^2)$ memory complexity of Softmax Multi-Head Attention (MHA). While substituting MHA with independent Multi-Layer Perceptrons (MLPs) achieves $O(N)$ scaling, it strips the architecture of spatial sequence routing, severely degrading global scene understanding and object permanence. In this paper, we propose the Linear Multi-Timescale Retention (LIA-MTR) module, a memory-efficient cross-modal bridge. By integrating an ELU-based positive feature mapping with adaptive write-gating and log-linearly distributed recurrent decays, LIA-MTR mathematically compresses continuous visual sequences into bounded memory states. Theoretical analysis proves the architecture operates with strict $O(N)$ sequence-interaction complexity. Empirically, synthetic retrieval evaluations demonstrate that LIA-MTR flawlessly routes context across 16,000 tokens, eliminating the "Lost in the Middle" degradation typical of naive linear attention. Hardware benchmarking reveals infinite-context scaling capabilities, natively processing 262,144 visual patches within an 11.2 GB VRAM footprint, whereas standard MHA suffers out-of-memory failure at 16,384 patches. Furthermore, following instruction tuning on 665K conversational samples, LIA-MTR significantly outperforms an industry-standard MLP baseline on the MME benchmark (71.00% vs. 68.11%), driven by a 10% absolute improvement in object permanence and superior global semantic extraction. This work establishes a mathematically rigorous, computationally flat foundation for infinite-context Vision-Language integration.

cs.CV↗

arXiv2025

Physics-Guided Surrogate Modeling for Machine Learning-Driven DLD Design Optimization

Sorting cells based on their mechanical properties is essential for applications in disease diagnostics, cell therapy, and biomedical research. Deterministic Lateral Displacement (DLD) devices provide a label-free method for achieving such sorting, but their performance is highly sensitive to cell size and deformability. Designing effective DLD geometries often demands extensive trial-and-error experimentation, as even small variations in cellular mechanical traits can cause significant changes in migration behavior. To address this challenge, we propose a simulation-driven machine learning (ML) framework that predicts suitable DLD design candidates for a given cell type. Our approach integrates high-fidelity particle-based simulations to model cell deformation and migration through microfluidic pillar arrays with supervised ML models trained to estimate optimal geometries. By mapping mechanical parameters such as bending rigidity and shear modulus to deformation index and migration angle, the framework enables rapid, data-informed design of DLD systems. We also demonstrate a deployable web interface to make this tool accessible for real-world device prototyping.

q-bio.QM↗

arXiv2025

Toward Unified Interphase Engineering: The Solid-Electrolyte Interphase in Batteries and Supercapacitors

The development of next-generation electrochemical energy storage requires devices that combine the high energy density of batteries with the power capability and long cycle life of supercapacitors. However, the interfacial phenomena governing performance in these systems remain poorly unified. The solid-electrolyte interphase (SEI), a nanoscale film formed by electrolyte decomposition, is well studied in batteries but its counterpart in supercapacitors has received limited systematic investigation despite growing experimental evidence. This review argues that SEI formation is a universal electrochemical process that occurs whenever electrode potentials drive electron transfer into electrolyte orbitals beyond their stability limits, independent of whether charge storage is Faradaic or non-Faradaic. Differences between battery SEIs and supercapacitor interphases arise mainly from operating conditions, not fundamental chemistry. Engineered interphases created through electrolyte additives, protective coatings, or surface functionalization suppress leakage currents, improve capacitance retention, and enable stable high-voltage operation. By identifying shared mechanisms and establishing transferable design rules, this unified framework provides a foundation for predictive interphase engineering that supports long-lived, high-performance energy-storage technologies.

physics.chem-ph↗

arXiv2025

Smart Traffic Systems: A Comprehensive Review of Recent Advancements, Technologies, and Challenges

With an ever-growing urban population, the need for transportation is increasing at an alarming rate. Thus, the massive increase in the number of vehicles is creating traffic congestion which creates various environmental, societal, and economic problems. To tackle traffic-related issues, several Smart Traffic Systems (STS) have been proposed and implemented. As a result, a comprehensive review of STS has become necessary. The main objective of this paper is to provide an overview and a thorough review of the existing STSs in terms of various technological approaches, traffic detection technologies using different sensors, various networking/communication tools, and their pros and cons. The paper also provides information on major STS services. In addition, challenges related to modern STS are identified. Therefore, the taxonomy of STSs provided in this paper will aid researchers, urban planners, and policymakers to recognize and install the best-suited STSs for their settings.

eess.SY↗

arXiv2025

Curvature-Dependent Polarity of Interfacial Energy Flow in Functionalized CNT Polymer Nanocomposites: A Reactive Molecular Dynamics Perspective

Carbon nanotube (CNT)-polymer composites are widely engineered using surface coatings and chemical treatments to improve interfacial bonding and load transfer. It has been suggested in the nanocomposite literature that nanotube curvature, in conjunction with surface functionalization such as polydopamine (PDA) coating, could serve as an additional control knob for tuning interfacial bonding and energy dissipation in polymer-CNT systems. While experimental and simulation studies have demonstrated the benefits of PDA functionalization, the fundamental mechanism by which nanotube curvature modulates interfacial energy flow and mechanical polarity remains unresolved. This gap is sharpened by a persistent paradox: identical PDA functionalization strengthens some CNT-polymer systems while weakening others, a curvature-dependent inconsistency that has remained unexplained. Here, we employ reactive molecular dynamics (ReaxFF) simulations to resolve how curvature and PDA functionalization jointly govern interfacial energy evolution in CNT-polyvinyl alcohol (PVA) nanocomposites. Our investigation reveals that curvature and PDA functionalization jointly produce opposite regimes of interfacial energy flow: high-curvature CNTs generate dissipative, frictional interphases, whereas low-curvature CNTs confine energy in rigid, cohesive shells. This polarity inversion originates from a curvature-induced transition in PDA adsorption geometry that transforms the interphase from an energy-releasing to an energy-storing configuration. These results establish curvature as a fundamental design parameter for engineering polymer-nanotube interfaces, offering a predictive route to tune interfacial energy flow, mechanical resilience, and transport properties beyond the limits of conventional chemical functionalization.

physics.atm-clus↗

arXiv2025

Dual-Bus Resonator for Multi-Port Spectral Engineering

Microresonators are essential in integrated photonics, enabling optical filters, modulators, sensors, and frequency converters. Their spectral response is governed by bus-to-resonator coupling, typically classified as under-, critical-, or over-coupling. Conventional single-bus designs inevitably link the conditions for critical coupling, a transmission zero, and maximum intra-cavity power, preventing independent control of these phenomena and restricting the ability to engineer coupling regimes and resonance lineshapes. Here we propose and experimentally demonstrate a dual-bus racetrack resonator that breaks this constraint. Our design demonstrates complementary channel-specific coupling regimes and enables wavelength-dependent Lorentzian-to-Fano lineshaping. We model the device using three-waveguide coupled-mode theory and pole-zero analysis, which reveals that transmission zeros are decoupled from cavity-defined critical coupling and maximum intra-cavity power. Furthermore, the dual-bus scheme operates broadband, spanning visible to mid-infrared across all four transmission channels, highlighting its spectral richness and platform independence. These results establish a general framework for multi-port spectral engineering in integrated photonics, with broad implications for tunable filters, modulators, sensors, and nonlinear optical systems.

physics.optics↗

arXiv2025

Superfluid-Mott transition in a frustrated triangular optical lattice

Geometric frustration can significantly increase the complexity and richness of many-body physics and, for instance, suppress antiferromagnetic order in quantum magnets. Here, we employ ultracold bosonic $^{39}$K atoms in a triangular optical lattice to study geometric frustration by stabilizing the gas at the frustrated upper band edge using negative absolute temperatures. We find that geometric frustration suppresses the critical interaction strength for the (chiral-)superfluid to Mott insulator ($χ$-SF-MI) quantum phase transition by a factor of 2.7(3) and furthermore changes the critical dynamics of the transition. Although the emergence of coherence during fast ramps from MI to the ($χ$-)SF regime is continuous in both cases, for ramps longer than a few tunnelling times, significant differences emerge. In the \frs case, no long-range order emerges on the studied timescales, highlighting a significantly reduced rate or even saturation of the emerging coherence. This work opens the door to quantum simulations of frustrated systems that are often intractable by classical simulations.

cond-mat.quant-gas↗

arXiv2024

Enhancing EmoBot: An In-Depth Analysis of User Satisfaction and Faults in an Emotion-Aware Chatbot

The research community has traditionally shown a keen interest in emotion modeling, with a notable emphasis on the detection aspect. In contrast, the exploration of emotion generation has received less attention.This study delves into an existing state-of-the-art emotional chatbot, EmoBot, designed for generating emotions in general-purpose conversations. This research involves a comprehensive examination, including a survey to evaluate EmoBot's proficiency in key dimensions like usability, accuracy, and overall user satisfaction, with a specific focus on fault tolerance. By closely examining the chatbot's operations, we identified some noteworthy shortcomings in the existing model. We propose some solutions designed to address and overcome the identified issues.

cs.HC↗

arXiv2023

Twisted van der Waals Quantum Materials: Fundamentals, Tunability and Applications

Twisted vdW quantum materials have emerged as a rapidly developing field of 2D semiconductors. These materials establish a new central research area and provide a promising platform for studying quantum phenomena and investigating the engineering of novel optoelectronic properties such as single-photon emission, non-linear optical response, magnon physics, and topological superconductivity. These captivating electronic and optical properties result from, and can be tailored by, the interlayer coupling using moiré patterns formed by vertically stacking atomic layers with controlled angle misorientation or lattice mismatch. Their outstanding properties and the high degree of tunability position them as compelling building blocks for both compact quantum-enabled devices and classical optoelectronics. This article offers a comprehensive review of recent advancements in the understanding and manipulation of twisted van der Waals structures and presents a survey of the state-of-the-art research on moiré superlattices, encompassing interdisciplinary interests. It delves into fundamental theories, synthesis and fabrication, and visualization techniques, and the wide range of novel physical phenomena exhibited by these structures, with a focus on their potential for practical device integration in applications ranging from quantum information to biosensors, and including classical optoelectronics such as modulators, light emitting diodes (LEDs), lasers, and photodetectors. It highlights the unique ability of moiré superlattices to connect multiple disciplines, covering chemistry, electronics, optics, photonics, magnetism, topological and quantum physics. This comprehensive review provides a valuable resource for researchers interested in moiré superlattices, shedding light on their fundamental characteristics and their potential for transformative applications in various fields.

cond-mat.mtrl-sci↗

arXiv2023

Bengali Intent Classification with Generative Adversarial BERT

Intent classification is a fundamental task in natural language understanding, aiming to categorize user queries or sentences into predefined classes to understand user intent. The most challenging aspect of this particular task lies in effectively incorporating all possible classes of intent into a dataset while ensuring adequate linguistic variation. Plenty of research has been conducted in the related domains in rich-resource languages like English. In this study, we introduce BNIntent30, a comprehensive Bengali intent classification dataset containing 30 intent classes. The dataset is excerpted and translated from the CLINIC150 dataset containing a diverse range of user intents categorized over 150 classes. Furthermore, we propose a novel approach for Bengali intent classification using Generative Adversarial BERT to evaluate the proposed dataset, which we call GAN-BnBERT. Our approach leverages the power of BERT-based contextual embeddings to capture salient linguistic features and contextual information from the text data, while the generative adversarial network (GAN) component complements the model's ability to learn diverse representations of existing intent classes through generative modeling. Our experimental results demonstrate that the GAN-BnBERT model achieves superior performance on the newly introduced BNIntent30 dataset, surpassing the existing Bi-LSTM and the stand-alone BERT-based classification model.

cs.CL↗

arXiv2023

BanglaBait: Semi-Supervised Adversarial Approach for Clickbait Detection on Bangla Clickbait Dataset

Intentionally luring readers to click on a particular content by exploiting their curiosity defines a title as clickbait. Although several studies focused on detecting clickbait titles in English articles, low resource language like Bangla has not been given adequate attention. To tackle clickbait titles in Bangla, we have constructed the first Bangla clickbait detection dataset containing 15,056 labeled news articles and 65,406 unlabelled news articles extracted from clickbait dense news sites. Each article has been labeled by three expert linguists and includes an article's title, body, and other metadata. By incorporating labeled and unlabelled data, we finetune a pretrained Bangla transformer model in an adversarial fashion using Semi Supervised Generative Adversarial Networks (SS GANs). The proposed model acts as a good baseline for this dataset, outperforming traditional neural network models (LSTM, GRU, CNN) and linguistic feature based models. We expect that this dataset and the detailed analysis and comparison of these clickbait detection models will provide a fundamental basis for future research into detecting clickbait titles in Bengali articles. We have released the corresponding code and dataset.

cs.CL↗

arXiv2023

PT-symmetric cross injection dual optoelectronic oscillator

An optoelectronic oscillator (OEO) is a time delay oscillator (TDO) that uses photonics technology to provide the long delay required to generate pristine microwave carriers. Parity-time (PT) symmetry concepts applied to an OEO offer the potential to achieve combined low phase noise and high sidemode suppression. A TDO composed of a pair of identical ring resonators coupled by a 2x2 coupler is modelled, and the coupler transmission matrix required for the oscillator to be PT- symmetric is derived. In a first configuration, the coupler is interpreted as the composition of a gain/loss block and a Mach-Zehnder interferometer (MZI) block. In practice, there are excess losses that must be compensated by a special dual amplifier with saturation characteristics compatible with PT- symmetry. The PT- symmetry phase transition determined by the gain/loss and the MZI differential phase parameters is found to be global and not local in its effect on modes. This is resolved by placing a short delay line within one arm of the MZI resulting in a frequency dependent and hence local mode-selective PT- symmetry phase transition. In addition, it is demonstrated that the first configuration may be transformed into a second but equivalent configuration as a cross-injection dual TDO with imbalanced delays. The local PT- symmetry phase transition may then be understood in terms of the Vernier effect. Advantageously, the second configuration enables the special dual amplifier to be replaced by a pair of matched but otherwise independent amplifiers. Thereby, the second configuration is amenable to practical implementation as a dual OEO using standard RF-photonic and RF-electronic components. The theory is validated by complex envelope model simulations using Simulink and phase model analytic results evaluated using MATLAB. There is excellent agreement between the theoretical and simulation results.

physics.optics↗

arXiv2023

Investigating self-supervised, weakly supervised and fully supervised training approaches for multi-domain automatic speech recognition: a study on Bangladeshi Bangla

Despite huge improvements in automatic speech recognition (ASR) employing neural networks, ASR systems still suffer from a lack of robustness and generalizability issues due to domain shifting. This is mainly because principal corpus design criteria are often not identified and examined adequately while compiling ASR datasets. In this study, we investigate the robustness of the state-of-the-art transfer learning approaches such as self-supervised wav2vec 2.0 and weakly supervised Whisper as well as fully supervised convolutional neural networks (CNNs) for multi-domain ASR. We also demonstrate the significance of domain selection while building a corpus by assessing these models on a novel multi-domain Bangladeshi Bangla ASR evaluation benchmark - BanSpeech, which contains approximately 6.52 hours of human-annotated speech and 8085 utterances from 13 distinct domains. SUBAK.KO, a mostly read speech corpus for the morphologically rich language Bangla, has been used to train the ASR systems. Experimental evaluation reveals that self-supervised cross-lingual pre-training is the best strategy compared to weak supervision and full supervision to tackle the multi-domain ASR task. Moreover, the ASR models trained on SUBAK.KO face difficulty recognizing speech from domains with mostly spontaneous speech. The BanSpeech will be publicly available to meet the need for a challenging evaluation benchmark for Bangla ASR.

cs.CL↗

arXiv2023

Survey on Leveraging Uncertainty Estimation Towards Trustworthy Deep Neural Networks: The Case of Reject Option and Post-training Processing

Although neural networks (especially deep neural networks) have achieved \textit{better-than-human} performance in many fields, their real-world deployment is still questionable due to the lack of awareness about the limitation in their knowledge. To incorporate such awareness in the machine learning model, prediction with reject option (also known as selective classification or classification with abstention) has been proposed in literature. In this paper, we present a systematic review of the prediction with the reject option in the context of various neural networks. To the best of our knowledge, this is the first study focusing on this aspect of neural networks. Moreover, we discuss different novel loss functions related to the reject option and post-training processing (if any) of network output for generating suitable measurements for knowledge awareness of the model. Finally, we address the application of the rejection option in reducing the prediction time for the real-time problems and present a comprehensive summary of the techniques related to the reject option in the context of extensive variety of neural networks. Our code is available on GitHub: \url{https://github.com/MehediHasanTutul/Reject_option}

cs.LG↗

arXiv2023

Broadband high-resolution integrated spectrometer architecture & data processing method

Up-to-date network telemetry is the key enabler for resource optimization by capacity scaling, fault recovery, and network reconfiguration among other means. Reliable optical performance monitoring in general and, specifically, the monitoring of the spectral profile of WDM signals in fixed- and flex- grid architectures across the entire C-band, remains challenging. This article describes a two-stage spectrometer architecture amenable to integration on a single chip that can measure quantitatively the spectrum across the entire C-band with a resolution of 1 GHz approximately. The first stage consists of a ring resonator with intra-ring phase shifter to provide a tuneable fine filter. The second stage makes use of an AWG subsystem and novel processing algorithm to synthesize a tuneable coarse filter with a flat passband which isolates individual resonances of a multiplicity of ring resonances. Due to its maturity and low loss, CMOS compatible Si$_3$N$_4$ is chosen for integration. A fabricated ring resonator functioning over the entire C-band with 1.3 GHz FWHM bandwidth resonances tunable over a complete free spectral range of 50 GHz is experimentally demonstrated. The complete system operation is demonstrated using an industry standard simulation tool and AWG constructor data. The operation of the circuit is invariant to the optical path length between individual components substantially improving robustness to fabrication process variations.

physics.optics↗

arXiv2023

Continuous tuning & thermally induced frequency drift stabilisation of time delay oscillators such as the optoelectronic oscillator

Delay line oscillators based on photonic components, such as the optoelectronic oscillator (OEO), offer the potential for realization of phase noise levels orders of magnitude lower than achievable by conventional microwave sources. Fibre optic-based delay lines can realize the large delay required for low phase noise systems whilst simultaneously achieving insertion loss levels that can be compensated by available microwave and photonic amplification technologies. However, the long fibre is vulnerable to environmental perturbations such as mechanical vibrations and variations in ambient temperature, which result in short term fluctuations and thermally induced drift of the oscillation frequency. The phase shifter used conventionally to adjust the frequency of an OEO to enable phase lock has a finite range that is insufficient to compensate the delay change resulting from operational temperature ranges. A solution to continuous tuning without mode-hopping and to compensation of thermally induced frequency drift without loss of lock of a time delay oscillator is proposed. The basic concept is to introduce a tuning mechanism that works with Cartesian co-ordinates on the complex plane and to avoid explicit use of polar co-ordinates. Consequently, the transmission of the tuning component may traverse the unit circle in either direction multiple times without range limitation to the phase. Thereby tuning by mode-hopping is avoided and expedients to stabilisation, such as the use of tunable lasers within a control loop or precision temperature stabilisation measures are not required. The concept is verified by Simulink simulations. The method has been experimentally tested successfully using a prototype OEO phase locked to a system reference. Solid lock was maintained even when the OEO was placed in an oven and cycled over a temperature range from ambient to 80 °C.

eess.SP↗

arXiv2022

Datta-Das transistor for atomtronic circuits using artificial gauge fields

Spin-dependent electrical injection has found useful applications in storage devices, but fully operational spin-dependent semiconductor electronics remain a challenging task because of weak spin-orbit couplings and/or strong spin relaxations. These limitations are lifted considering atoms instead of electrons or holes as spin carriers. In this emerging field of atomtronics, we demonstrate the equivalent of a Datta-Das transistor using a degenerate Fermi gas of strontium atoms as spin carriers in interaction with a tripod laser-beams scheme. We explore the dependence of spin rotation, and we identify two key control parameters which we interpret as equivalent to the gate-source and drain-source voltages of a field effect transistor. Our finding broadens the spectrum of atomtronics devices for implementation of operational spin-sensitive circuits.

cond-mat.mes-hall↗