arXiv subjects

Abhijit Sarkar

Publications and source records attributed to Abhijit Sarkar.

At least 19 recordsLinked to original sources

arXiv2026

D3VL: Understanding Driving Scenes from 3D Time Series Data and Video with Language Models

Recent advances in Multimodal Large Language Models (MLLMs) have triggered the development of end-to-end MLLMs for autonomous driving. However, the main emphasis to date has been for MLLMs using 2D images and videos. In contrast, this paper considers MLLM effectiveness using 3D sensors, particularly LiDAR and stereo cameras. LiDAR presents unique challenges to integration within an MLLM, largely because of data sparsity and lack of a grid structure for the data. For similar reasons, fusion of camera and LiDAR data within an MLLM pipeline is also uncommon. However, most autonomous systems rely on LiDAR-based sensing, and incorporating 3D data has been proven to improve performance in traditional 3D scene perception tasks. This paper presents D3VL, a novel MLLM framework that integrates 2D and 3D time-series data in a single but simple architecture. The model aims to answer questions involving traffic scene understanding and safety. D3VL shows an 11% improvement in the KITTI Question-Answering (QA) dataset compared to baseline methods in processing 2D and 3D time-series data. This paper further introduces the Waymo QA dataset extension, which assesses models' capabilities in processing 3D and time-series data under diverse driving conditions. D3VL implementation code and WaymoQA extension can be found on our supplemental website: https://automotivesafety-lvlm.github.io

cs.CV↗

arXiv2026

SinD 2.0: A Multi-City UAV Dataset with Semantic Risk Annotations for SOTIF-Oriented Safety Validation at Signalized Intersections

Safety validation at signalized intersections remains a critical bottleneck for the deployment of autonomous driving systems (ADS), as these scenarios involve dense heterogeneous traffic, contested right of way, and long-tail safety-critical interactions, posing significant challenges to the Safety of the Intended Functionality (SOTIF). Existing naturalistic driving datasets often suffer from geographical homogeneity, sparsity of safety-critical events, and lack of semantic risk annotations, which limit the evaluation of algorithmic generalizability and targeted SOTIF verification. To address these gaps, this paper introduces SinD 2.0, a large-scale drone-based intersection dataset dedicated to cross-domain ADS safety analysis. The main contributions of SinD 2.0 are: (1) Cross-domain diversity: It covers six signalized intersections across four Chinese cities, capturing distinct intersection topologies and regional driving behavior characteristics; (2) High-density risk interactions: A total of 32,682 safety-critical events are extracted via surrogate safety measures, significantly enriching the density of boundary test scenarios; (3) Hierarchical semantic annotations: Besides integration with high-definition (HD) maps and Signal Phase and Timing (SPaT) data, it provides multi-dimensional semantic labels including traffic violations, high-risk interactions, visual shielding, and narrow feasible areas; (4) Full-stack testing toolchain: It supports automated scenario extraction, prediction-only evaluation, open-loop replay, reactive closed-loop testing, and photorealistic rendering. Benchmark experiments demonstrate that SinD 2.0 exhibits significant domain shifts across cities, and the semantic risk subsets can effectively expose the performance limitations of ADS algorithms. The dataset, annotations, and testing toolchain are available at https://github.com/SOTIF-AVLab/SinD/tree/main.

cs.RO↗

arXiv2026

EyeCue: Driver Cognitive Distraction Detection via Gaze-Empowered Egocentric Video Understanding

Driver cognitive distraction is a major cause of road collisions and remains difficult to detect. Unlike manual or visual distraction, cognitive distraction is diverted by thoughts unrelated to driving, even when the driver appears visually attentive and exhibits no explicit physical movements. In this work, we propose EyeCue, a gaze-empowered egocentric video understanding framework, to detect driver cognitive distraction. A key insight is that cognitive distraction manifests in the interaction between eye gaze and visual context. To capture this interaction, EyeCue integrates eye gaze with egocentric video to enable context-aware modeling of the driver's attention over time. Furthermore, to tackle the limited scale and diversity of existing datasets, we introduce CogDrive, a comprehensive multi-scenario dataset that augments four existing driving datasets with cognitive distraction annotations. Through extensive evaluations on CogDrive, we show that EyeCue achieves the highest accuracy of 74.38%, outperforming 11 baselines from 6 model families by over 7%. Notably, EyeCue can achieve an accuracy of over 70% across various driving scenarios (different road types, times of day, and weather conditions) with strong generalizability. These results highlight the importance of modeling gaze-context interactions and the effectiveness of cross-modal interaction modeling for multimodal cognitive distraction detection. Our codes and CogDrive dataset resources are available at https://github.com/langzhang2000/EyeCue.

cs.CV↗

arXiv2025

Noise-Driven AI Sensors: Secure Healthcare Monitoring with PUFs

Wearable and implantable healthcare sensors are pivotal for real-time patient monitoring but face critical challenges in power efficiency, data security, and signal noise. This paper introduces a novel platform that leverages hardware noise as a dual-purpose resource to enhance machine learning (ML) robustness and secure data via Physical Unclonable Functions (PUFs). By integrating noise-driven signal processing, PUFbased authentication, and ML-based anomaly detection, our system achieves secure, low-power monitoring for devices like ECG wearables. Simulations demonstrate that noise improves ML accuracy by 8% (92% for detecting premature ventricular contractions (PVCs) and atrial fibrillation (AF)), while PUFs provide 98% uniqueness for tamper-resistant security, all within a 50 uW power budget. This unified approach not only addresses power, security, and noise challenges but also enables scalable, intelligent sensing for telemedicine and IoT applications.

eess.SP↗

arXiv2024

Re-evaluation of Face Anti-spoofing Algorithm in Post COVID-19 Era Using Mask Based Occlusion Attack

Face anti-spoofing algorithms play a pivotal role in the robust deployment of face recognition systems against presentation attacks. Conventionally, full facial images are required by such systems to correctly authenticate individuals, but the widespread requirement of masks due to the current COVID-19 pandemic has introduced new challenges for these biometric authentication systems. Hence, in this work, we investigate the performance of presentation attack detection (PAD) algorithms under synthetic facial occlusions using masks and glasses. We have used five variants of masks to cover the lower part of the face with varying coverage areas (low-coverage, medium-coverage, high-coverage, round coverage), and 3D cues. We have also used different variants of glasses that cover the upper part of the face. We systematically tested the performance of four PAD algorithms under these occlusion attacks using a benchmark dataset. We have specifically looked at four different baseline PAD algorithms that focus on, texture, image quality, frame difference/motion, and abstract features through a convolutional neural network (CNN). Additionally we have introduced a new hybrid model that uses CNN and local binary pattern textures. Our experiment shows that adding the occlusions significantly degrades the performance of all of the PAD algorithms. Our results show the vulnerability of face anti-spoofing algorithms with occlusions, which could be in the usage of such algorithms in the post-pandemic era.

cs.CV↗

arXiv2023

A Bayesian model calibration framework for stochastic compartmental models with both time-varying and time-invariant parameters

We consider state and parameter estimation for compartmental models having both time-varying and time-invariant parameters. Though the described Bayesian computational framework is general, we look at a specific application to the susceptible-infectious-removed (SIR) model which describes a basic mechanism for the spread of infectious diseases through a system of coupled nonlinear differential equations. The SIR model consists of three states, namely, the three compartments, and two parameters which control the coupling among the states. The deterministic SIR model with time-invariant parameters has shown to be overly simplistic for modelling the complex long-term dynamics of diseases transmission. Recognizing that certain model parameters will naturally vary in time due to seasonal trends, non-pharmaceutical interventions, and other random effects, the estimation procedure must systematically permit these time-varying effects to be captured, without unduly introducing artificial dynamics into the system. To this end, we leverage the robustness of the Markov Chain Monte Carlo (MCMC) algorithm for the estimation of time-invariant parameters alongside nonlinear filters for the joint estimation of the system state and time-varying parameters. We demonstrate performance of the framework by first considering a series of examples using synthetic data, followed by an exposition on public health data collected in the province of Ontario.

cs.CE↗

arXiv2023

Sparse Bayesian neural networks for regression: Tackling overfitting and computational challenges in uncertainty quantification

Neural networks (NNs) are primarily developed within the frequentist statistical framework. Nevertheless, frequentist NNs lack the capability to provide uncertainties in the predictions, and hence their robustness can not be adequately assessed. Conversely, the Bayesian neural networks (BNNs) naturally offer predictive uncertainty by applying Bayes' theorem. However, their computational requirements pose significant challenges. Moreover, both frequentist NNs and BNNs suffer from overfitting issues when dealing with noisy and sparse data, which render their predictions unwieldy away from the available data space. To address both these problems simultaneously, we leverage insights from a hierarchical setting in which the parameter priors are conditional on hyperparameters to construct a BNN by applying a semi-analytical framework known as nonlinear sparse Bayesian learning (NSBL). We call our network sparse Bayesian neural network (SBNN) which aims to address the practical and computational issues associated with BNNs. Simultaneously, imposing a sparsity-inducing prior encourages the automatic pruning of redundant parameters based on the automatic relevance determination (ARD) concept. This process involves removing redundant parameters by optimally selecting the precision of the parameters prior probability density functions (pdfs), resulting in a tractable treatment for overfitting. To demonstrate the benefits of the SBNN algorithm, the study presents an illustrative regression problem and compares the results of a BNN using standard Bayesian inference, hierarchical Bayesian inference, and a BNN equipped with the proposed algorithm. Subsequently, we demonstrate the importance of considering the full parameter posterior by comparing the results with those obtained using the Laplace approximation with and without NSBL.

cs.CE↗

arXiv2023

Multilevel Scalable Solvers for Stochastic Linear and Nonlinear Problems

This article discusses the uncertainty quantification (UQ) for time-independent linear and nonlinear partial differential equation (PDE)-based systems with random model parameters carried out using sampling-free intrusive stochastic Galerkin method leveraging multilevel scalable solvers constructed combining two-grid Schwarz method and AMG. High-resolution spatial meshes along with a large number of stochastic expansion terms increase the system size leading to significant memory consumption and computational costs. Domain decomposition (DD)-based parallel scalable solvers are developed to this end for linear and nonlinear stochastic PDEs. A generalized minimum residual (GMRES) iterative solver equipped with a multilevel preconditioner consisting of restricted additive Schwarz (RAS) for the fine grid and algebraic multigrid (AMG) for the coarse grid is constructed to improve scalability. Numerical experiments illustrate the scalabilities of the proposed solver for stochastic linear and nonlinear Poisson problems.

cs.CE↗

arXiv2023

Exploring hierarchical framework of nonlinear sparse Bayesian learning algorithm through numerical investigations

Sparse Bayesian learning (SBL) has been extensively utilized in data-driven modeling to combat the issue of overfitting. While SBL excels in linear-in-parameter models, its direct applicability is limited in models where observations possess nonlinear relationships with unknown parameters. Recently, a semi-analytical Bayesian framework known as nonlinear sparse Bayesian learning (NSBL) was introduced by the authors to induce sparsity among model parameters during the Bayesian inversion of nonlinear-in-parameter models. NSBL relies on optimally selecting the hyperparameters of sparsity-inducing Gaussian priors. It is inherently an approximate method since the uncertainty in the hyperparameter posterior is disregarded as we instead seek the maximum a posteriori (MAP) estimate of the hyperparameters (type-II MAP estimate). This paper aims to investigate the hierarchical structure that forms the basis of NSBL and validate its accuracy through a comparison with a one-level hierarchical Bayesian inference as a benchmark in the context of three numerical experiments: (i) a benchmark linear regression example with Gaussian prior and Gaussian likelihood, (ii) the same regression problem with a highly non-Gaussian prior, and (iii) an example of a dynamical system with a non-Gaussian prior and a highly non-Gaussian likelihood function, to explore the performance of the algorithm in these new settings. Through these numerical examples, it can be shown that NSBL is well-suited for physics-based models as it can be readily applied to models with non-Gaussian prior distributions and non-Gaussian likelihood functions. Moreover, we illustrate the accuracy of the NSBL algorithm as an approximation to the one-level hierarchical Bayesian inference and its ability to reduce the computational cost while adequately exploring the parameter posteriors.

cs.CE↗

arXiv2023

GAN-based Deidentification of Drivers' Face Videos: An Assessment of Human Factors Implications in NDS Data

This paper addresses the problem of sharing drivers' face videos for transportation research while adhering to proper ethical guidelines. The paper first gives an overview of the multitude of problems associated with sharing such data and then proposes a framework on how artificial intelligence-based techniques, specifically face swapping, can be used for de-identifying drivers' faces. Through extensive experimentation with an Oak Ridge National Laboratory (ORNL) dataset, we demonstrate the effectiveness of face-swapping algorithms in preserving essential attributes related to human factors research, including eye movements, head movements, and mouth movements. The efficacy of the framework was also tested on various naturalistic driving study data collected at the Virginia Tech Transportation Institute. The results achieved through the proposed techniques were evaluated qualitatively and quantitatively using various metrics. Finally, we discuss possible measures for sharing the de-identified videos with the greater research community.

cs.CV↗

arXiv2022

Probabilistic Prediction of Coalescence Flutter Using Measurements: Application to the Flutter Margin Method

Zimmerman and Weissenburger's flutter margin method is widely used to estimate the aeroelastic coalescence flutter speed. In contrast to aeroelastic decay rates, the flutter margin exhibits monotonic decay with respect to airspeed redering it effective in extrapolating the flutter speed using flight test data conducted at pre-flutter airspeeds. This paper reports the generalization of the Bayesian formulation of the flutter margin method by Khalil et al. developed to tackle measurement and modeling uncertainties. This paper improves the predictive performance of the previous algorithm by incorporating the joint prior of aeroelastic modal frequencies and decay rates among airspeeds in order to better estimate the joint posterior of modal parameters using observational data. The modal parameter prior is constructed using the classical two-degree-of-freedom pitch-plunge aeroelastic model whose system matrices (e.g. structural stiffness and damping matrices) vary randomly. Such joint modal parameter prior enforces statistical dependence among posteriors of modal parameters and the associated flutter margins across airspeeds. Numerical studies demonstrate a considerable reduction of uncertainties on the predicted flutter speed obtained from the generalized Bayesian flutter margin method. This improved algorithm can cut cost by reducing the number of flight tests and better assess the uncertainty against aeroelastic flutter.

cs.CE↗

arXiv2022

Encoding nonlinear and unsteady aerodynamics of limit cycle oscillations using nonlinear sparse Bayesian learning

This paper investigates the applicability of a recently-proposed nonlinear sparse Bayesian learning (NSBL) algorithm to identify and estimate the complex aerodynamics of limit cycle oscillations. NSBL provides a semi-analytical framework for determining the data-optimal sparse model nested within a (potentially) over-parameterized model. This is particularly relevant to nonlinear dynamical systems where modelling approaches involve the use of physics-based and data-driven components. In such cases, the data-driven components, where analytical descriptions of the physical processes are not readily available, are often prone to overfitting, meaning that the empirical aspects of these models will often involve the calibration of an unnecessarily large number of parameters. While it may be possible to fit the data well, this can become an issue when using these models for predictions in regimes that are different from those where the data was recorded. In view of this, it is desirable to not only calibrate the model parameters, but also to identify the optimal compromise between data-fit and model complexity. In this paper, this is achieved for an aeroelastic system where the structural dynamics are well-known and described by a differential equation model, coupled with a semi-empirical aerodynamic model for laminar separation flutter resulting in low-amplitude limit cycle oscillations. For the purpose of illustrating the benefit of the algorithm, in this paper, we use synthetic data to demonstrate the ability of the algorithm to correctly identify the optimal model and model parameters, given a known data-generating model. The synthetic data are generated from a forward simulation of a known differential equation model with parameters selected so as to mimic the dynamics observed in wind-tunnel experiments.

cs.CE↗

arXiv2022

Robust Bayesian state and parameter estimation framework for stochastic dynamical systems with combined time-varying and time-invariant parameters

We consider state and parameter estimation for a dynamical system having both time-varying and time-invariant parameters. It has been shown that the robustness of the Markov Chain Monte Carlo (MCMC) algorithm for estimating time-invariant parameters alongside nonlinear filters for state estimation provided more reliable estimates than the estimates obtained solely using nonlinear filters for combined state and parameter estimation. In a similar fashion, we adopt the extended Kalman filter (EKF) for state estimation and the estimation of the time-varying system parameters, but reserve the task of estimating time-invariant parameters to the MCMC algorithm. In a standard method, we augment the state vector to include the original states of the system and the subset of the parameters that are time-varying. Each time-varying parameter is perturbed by a white noise process, and we treat the strength of this artificial noise as an additional time-invariant parameter to be estimated by MCMC, circumventing the need for manual tuning. Conventionally, both time-varying and time-invariant parameters are appended in the state vector, and thus for the purpose of estimation, both are free to vary in time. However, allowing time-invariant system parameters to vary in time introduces artificial dynamics into the system, which we avoid by treating these time-invariant parameters as static and estimating them using MCMC. Furthermore, by estimating the time-invariant parameters by MCMC, the augmented state is smaller and the nonlinearity in the ensuing state space model will tend to be weaker than in the conventional approach. We illustrate the above-described approach for a simple dynamical system in which some model parameters are time-varying, while the remaining parameters are time-invariant.

cs.CE↗

arXiv2022

Domain Decomposition of Stochastic PDEs: Development of Probabilistic Wirebasket-based Two-level Preconditioners

Realistic physical phenomena exhibit random fluctuations across many scales in the input and output processes. Models of these phenomena require stochastic PDEs. For three-dimensional coupled (vector-valued) stochastic PDEs (SPDEs), for instance, arising in linear elasticity, the existing two-level domain decomposition solvers with the vertex-based coarse grid show poor numerical and parallel scalabilities. Therefore, new algorithms with a better resolved coarse grid are needed. The probabilistic wirebasket-based coarse grid for a two-level solver is devised in three dimensions. This enriched coarse grid provides an efficient mechanism for global error propagation and thus improves the convergence. This development enhances the scalability of the two-level solver in handling stochastic PDEs in three dimensions. Numerical and parallel scalabilities of this algorithm are studied using MPI and PETSc libraries on high-performance computing (HPC) systems. Implementational challenges of the intrusive spectral stochastic finite element methods (SSFEM) are addressed by coupling domain decomposition solvers with FEniCS general purpose finite element package. This work generalizes the applications of intrusive SSFEM to tackle a variety of stochastic PDEs and emphasize the usefulness of the domain decomposition-based solvers and HPC for uncertainty quantification.

cs.CE↗

arXiv2022

Scalable Computational Algorithms for Geo-spatial Covid-19 Spread in High Performance Computing

A nonlinear partial differential equation (PDE) based compartmental model of COVID-19 provides a continuous trace of infection over space and time. Finer resolutions in the spatial discretization, the inclusion of additional model compartments and model stratifications based on clinically relevant categories contribute to an increase in the number of unknowns to the order of millions. We adopt a parallel scalable solver allowing faster solutions for these high fidelity models. The solver combines domain decomposition and algebraic multigrid preconditioners at multiple levels to achieve the desired strong and weak scalability. As a numerical illustration of this general methodology, a five-compartment susceptible-exposed-infected-recovered-deceased (SEIRD) model of COVID-19 is used to demonstrate the scalability and effectiveness of the proposed solver for a large geographical domain (Southern Ontario). It is possible to predict the infections up to three months for a system size of 92 million (using 1780 processes) within 7 hours saving months of computational effort needed for the conventional solvers.

cs.CE↗

arXiv2022

Color Invariant Skin Segmentation

This paper addresses the problem of automatically detecting human skin in images without reliance on color information. A primary motivation of the work has been to achieve results that are consistent across the full range of skin tones, even while using a training dataset that is significantly biased toward lighter skin tones. Previous skin-detection methods have used color cues almost exclusively, and we present a new approach that performs well in the absence of such information. A key aspect of the work is dataset repair through augmentation that is applied strategically during training, with the goal of color invariant feature learning to enhance generalization. We have demonstrated the concept using two architectures, and experimental results show improvements in both precision and recall for most Fitzpatrick skin tones in the benchmark ECU dataset. We further tested the system with the RFW dataset to show that the proposed method performs much more consistently across different ethnicities, thereby reducing the chance of bias based on skin color. To demonstrate the effectiveness of our work, extensive experiments were performed on grayscale images as well as images obtained under unconstrained illumination and with artificial filters. Source code: https://github.com/HanXuMartin/Color-Invariant-Skin-Segmentation

cs.CV↗

arXiv2021

Behavior of anisotropic fluids with Chaplygin equation of state in Buchdahl spacetime

In the present study we have proposed a new model of an anisotropic compact star which admits the Chaplygin equation of state. For this purpose, we consider Buchdahl ansatz. We obtain the solution of proposed model in closed form which is non-singular, regular and well-behaved. In addition to this, we show that the model satisfies all the energy conditions and maintains the hydrostatic equilibrium equation. This model represents compact stars like PSR B0943+10, Her X-1 and SAX J1808.4-3658 to a very good approximate.

gr-qc↗

arXiv2021

GCNNMatch: Graph Convolutional Neural Networks for Multi-Object Tracking via Sinkhorn Normalization

This paper proposes a novel method for online Multi-Object Tracking (MOT) using Graph Convolutional Neural Network (GCNN) based feature extraction and end-to-end feature matching for object association. The Graph based approach incorporates both appearance and geometry of objects at past frames as well as the current frame into the task of feature learning. This new paradigm enables the network to leverage the "context" information of the geometry of objects and allows us to model the interactions among the features of multiple objects. Another central innovation of our proposed framework is the use of the Sinkhorn algorithm for end-to-end learning of the associations among objects during model training. The network is trained to predict object associations by taking into account constraints specific to the MOT task. Experimental results demonstrate the efficacy of the proposed approach in achieving top performance on the MOT '15, '16, '17 and '20 Challenges among state-of-the-art online approaches. The code is available at https://github.com/IPapakis/GCNNMatch.

cs.CV↗