arXiv subjects

Feilong Wang

Publications and source records attributed to Feilong Wang.

At least 19 recordsLinked to original sources

arXiv2026

Field-amplified readouts of weak altermagnetic exchange in MnF$_2$

MnF$_2$, the textbook two-sublattice antiferromagnet, has reemerged as a prototypical altermagnet, yet the sublattice-odd exchange that defines this identity remains under active debate: it enters the magnon splitting only in quadrature with the dipole--dipole interaction, its magnitude suppressed and its sign erased. An overdetermined first-principles total-energy mapping resolves this scale as a seventh-neighbor imbalance $\delta{J_7}\simeq+8~\mu$eV. The resulting Hamiltonian, with the dipole--dipole interaction included explicitly, reproduces the low-energy gap and the visible finite-momentum splitting. A longitudinal field $B\parallel c$ then acts as a linear amplifier of the hidden scale, opening two signed, field-linear readouts. The first is the compensation field $B^\ast(\mathbf Q)$, the position of minimum splitting, which is equal and opposite at the rotation-related partner momenta: a shift from zero field is itself evidence of a finite imbalance, its side gives the sign, and its magnitude, $|B^\ast|\simeq0.34$~T here, gives the scale. The second is the fixed-field contrast of the partner splittings, $\simeq0.14$~meV at $1$~T, six times the zero-field excess: a sign check from just two spectra. Both readouts survive a $0.12$~meV energy resolution, and the construction carries over to any easy-axis collinear altermagnet, bringing $\mu$eV altermagnetic exchange within present instrumental reach.

cond-mat.str-el↗

arXiv2025

Robustness of LLM-enabled vehicle trajectory prediction under data security threats

The integration of large language models (LLMs) into automated driving systems has opened new possibilities for reasoning and decision-making by transforming complex driving contexts into language-understandable representations. Recent studies demonstrate that fine-tuned LLMs can accurately predict vehicle trajectories and lane-change intentions by gathering and transforming data from surrounding vehicles. However, the robustness of such LLM-based prediction models for safety-critical driving systems remains unexplored, despite the increasing concerns about the trustworthiness of LLMs. This study addresses this gap by conducting a systematic vulnerability analysis of LLM-enabled vehicle trajectory prediction. We propose a one-feature differential evolution attack that perturbs a single kinematic feature of surrounding vehicles within the LLM's input prompts under a black-box setting. Experiments on the highD dataset reveal that even minor, physically plausible perturbations can significantly disrupt model outputs, underscoring the susceptibility of LLM-based predictors to adversarial manipulation. Further analyses reveal a trade-off between accuracy and robustness, examine the failure mechanism, and explore potential mitigation solutions. The findings provide the very first insights into adversarial vulnerabilities of LLM-driven automated vehicle models in the context of vehicular interactions and highlight the need for robustness-oriented design in future LLM-based intelligent transportation systems.

cs.LG↗

arXiv2025

Phase transitions and spectral singularities in a class of one-dimensional parity-time-symmetric complex potentials

We investigate a two-parametric family of one-dimensional non-Hermitian complex potentials with parity-time ($\mathcal{PT}$) symmetry. We find that there exist two distinct types of phase transitions, from an unbroken phase (characterized by a real spectrum) to a broken phase (where the spectrum becomes complex). The first type involves the emergence of a pair of complex eigenvalues bifurcating from the continuous spectrum. The second type is associated with the collision of such pairs at the bottom of the continuous spectrum. The first transition type is closely related to spectral singularities (SSs), at which point the transmission and reflection coefficients are divergent simultaneously. The second is associated with the emergence of bound states. In particular, under specific parameter conditions, we construct an exact bound state solution. By systematically exploring the parameter space, we establish a universal relationship governing the number of SSs in these potentials. These findings provide a fundamental theoretical framework for manipulating wave scattering in non-Hermitian systems, offering promising implications for designing advanced optical and quantum devices.

quant-ph↗

arXiv2025

Bistability and Exact Reflectionless States in Nonlinear Scattering of a Bose--Einstein Condensate

We investigate the mean-field scattering dynamics of a quasi-one-dimensional Bose--Einstein condensate interacting with a Rosen--Morse potential. For specific potential and nonlinearity parameters, we derive analytically exact, degenerate scattering states (doubly or triply degenerate) exhibiting perfect transmission. Using the Bogoliubov--de Gennes approach, we analyze the stability of these reflectionless degenerate states, demonstrating that only one solution within each degenerate manifold is dynamically stable. Furthermore, we study a configuration with spatially localized nonlinearity, identifying an exact reflectionless state under specific conditions. Numerical analysis shows that this state marks the system's transition from monostability to bistability as the incident wave amplitude increases. Our work establishes an analytic framework for these multistable transmission phenomena, directly relevant to coherent matter-wave transport in ultracold atomic systems and optical propagation in engineered photonic lattices.

cond-mat.quant-gas↗

arXiv2025

Model-Targeted Data Poisoning Attacks against ITS Applications with Provable Convergence

The growing reliance of intelligent systems on data makes the systems vulnerable to data poisoning attacks. Such attacks could compromise machine learning or deep learning models by disrupting the input data. Previous studies on data poisoning attacks are subject to specific assumptions, and limited attention is given to learning models with general (equality and inequality) constraints or lacking differentiability. Such learning models are common in practice, especially in Intelligent Transportation Systems (ITS) that involve physical or domain knowledge as specific model constraints. Motivated by ITS applications, this paper formulates a model-target data poisoning attack as a bi-level optimization problem with a constrained lower-level problem, aiming to induce the model solution toward a target solution specified by the adversary by modifying the training data incrementally. As the gradient-based methods fail to solve this optimization problem, we propose to study the Lipschitz continuity property of the model solution, enabling us to calculate the semi-derivative, a one-sided directional derivative, of the solution over data. We leverage semi-derivative descent to solve the bi-level optimization problem, and establish the convergence conditions of the method to any attainable target model. The model and solution method are illustrated with a simulation of a poisoning attack on the lane change detection using SVM.

math.OC↗

arXiv2024

Set-Valued Sensitivity Analysis of Deep Neural Networks

This paper proposes a sensitivity analysis framework based on set valued mapping for deep neural networks (DNN) to understand and compute how the solutions (model weights) of DNN respond to perturbations in the training data. As a DNN may not exhibit a unique solution (minima) and the algorithm of solving a DNN may lead to different solutions with minor perturbations to input data, we focus on the sensitivity of the solution set of DNN, instead of studying a single solution. In particular, we are interested in the expansion and contraction of the set in response to data perturbations. If the change of solution set can be bounded by the extent of the data perturbation, the model is said to exhibit the Lipschitz like property. This "set-to-set" analysis approach provides a deeper understanding of the robustness and reliability of DNNs during training. Our framework incorporates both isolated and non-isolated minima, and critically, does not require the assumption that the Hessian of loss function is non-singular. By developing set-level metrics such as distance between sets, convergence of sets, derivatives of set-valued mapping, and stability across the solution set, we prove that the solution set of the Fully Connected Neural Network holds Lipschitz-like properties. For general neural networks (e.g., Resnet), we introduce a graphical-derivative-based method to estimate the new solution set following data perturbation without retraining.

cs.LG↗

arXiv2024

Mitigating biases in big mobility data: a case study of monitoring large-scale transit systems

Big mobility datasets (BMD) have shown many advantages in studying human mobility and evaluating the performance of transportation systems. However, the quality of BMD remains poorly understood. This study evaluates biases in BMD and develops mitigation methods. Using Google and Apple mobility data as examples, this study compares them with benchmark data from governmental agencies. Spatio-temporal discrepancies between BMD and benchmark are observed and their impacts on transportation applications are investigated, emphasizing the urgent need to address these biases to prevent misguided policymaking. This study further proposes and tests a bias mitigation method. It is shown that the mitigated BMD could generate valuable insights into large-scale public transit systems across 100+ US counties, revealing regional disparities of the recovery of transit systems from the COVID-19. This study underscores the importance of caution when using BMD in transportation research and presents effective mitigation strategies that would benefit practitioners.

physics.soc-ph↗

arXiv2024

Data Poisoning Attacks in Intelligent Transportation Systems: A Survey

Emerging technologies drive the ongoing transformation of Intelligent Transportation Systems (ITS). This transformation has given rise to cybersecurity concerns, among which data poisoning attack emerges as a new threat as ITS increasingly relies on data. In data poisoning attacks, attackers inject malicious perturbations into datasets, potentially leading to inaccurate results in offline learning and real-time decision-making processes. This paper concentrates on data poisoning attack models against ITS. We identify the main ITS data sources vulnerable to poisoning attacks and application scenarios that enable staging such attacks. A general framework is developed following rigorous study process from cybersecurity but also considering specific ITS application needs. Data poisoning attacks against ITS are reviewed and categorized following the framework. We then discuss the current limitations of these attack models and the future research directions. Our work can serve as a guideline to better understand the threat of data poisoning attacks against ITS applications, while also giving a perspective on the future development of trustworthy ITS.

cs.CR↗

arXiv2024

Localized State-Induced Enhanced Emission in Perovskite Nanocrystals

The luminescent mechanism of nano-sized materials with indirect bandgap and parity-forbidden transition has always been a critical issue in breaking through bottlenecks in traditional luminescent materials. The lack of understanding has resulted in great disputes regarding the origin of fluorescence in weak-transition nanocrystals (NCs). This paper proposes a new physical luminescence model, named as localized state-induced enhanced emission (LIEE), i.e., localized processes induced by size-limitation or partial doping, to explain the anomalous luminescence from non-luminescent state to luminescent state in various perovskite nanocrystals with indirect/direct bandgap or parity-forbidden transition. These findings provide a theoretical viewpoint to design efficient lead-free perovskite NCs and promote the development of fast optical transitions in luminescent materials.

cond-mat.mtrl-sci↗

arXiv2023

Infrastructure-enabled GPS Spoofing Detection and Correction

Accurate and robust localization is crucial for supporting high-level driving automation and safety. Modern localization solutions rely on various sensors, among which GPS has been and will continue to be essential. However, GPS can be vulnerable to malicious attacks and GPS spoofing has been identified as a high threat. With transportation infrastructure becoming increasingly important in supporting emerging vehicle technologies and systems, this study explores the potential of applying infrastructure data for defending against GPS spoofing. We propose an infrastructure-enabled framework using roadside units as an independent, secured data source. A real-time detector, based on the Isolation Forest, is constructed to detect GPS spoofing. Once spoofing is detected, GPS measurements are isolated, and the potentially compromised location estimator is corrected using secure infrastructure data. We test the proposed method using both simulation and real-world data and show its effectiveness in defending against various GPS spoofing attacks, including stealthy attacks that are proposed to fail the production-grade autonomous driving systems.

eess.SP↗

arXiv2022

Localized State-Induced Enhanced Intrinsic Phonon-Free Optical Transition in Silicon Nanocrystals

Silicon photoluminescence and lasing have been critical issues to breakthrough bottlenecks in the understanding of luminescence mechanisms. Unfortunately, long-standing disputes about the exciton recombination mechanism and fluorescence lifetime remain unresolved, especially about whether silicon nanocrystals (Si NCs) can realize fast direct-bandgap-like optical transitions. Here, using ground-state and excited-state density functional theory (DFT), we obtained intrinsic phonon-free optical transitions at sizes from Si22 to Si705, showing that very small Si NCs can realize a strong direct optical transition. Orbital labeling results show that this rapid transition does not come from the Γ-Γ -like transition, contrary to the conclusions from the effective mass approximation (EMA) and that Γ-X mixing leads to a quasi-direct bandgap. This anomalous transition is particularly intense with decreasing size (or enhancement of quantum confinement). By investigating electron and hole distributions generated in the optical transition, localized state-induced enhanced emission (LIEE) in Si NCs was proposed. Quantum confinement (QC) distorts the excited-state electron spatial distribution by localizing Bloch waves into the NC core, resulting in increased hole and electron overlap, thus inducing a fast optical process. This work resolves important debates and proposes LIEE to explain the anomalous luminescence--a phase transition from weak (or none) luminescent state to strong optical transition, which will aid attempts at realizing high-radiative-rate NCs materials and application-level Si lasers.

cond-mat.mes-hall↗

arXiv2020

Extracting Trips from Multi-Sourced Data for Mobility Pattern Analysis: An App-Based Data Example

Passively-generated data, such as GPS data and cellular data, bring tremendous opportunities for human mobility analysis and transportation applications. Since their primary purposes are often non-transportation related, the passively-generated data need to be processed to extract trips. Most existing trip extraction methods rely on data that are generated via a single positioning technology such as GPS or triangulation through cellular towers (thereby called single-sourced data), and methods to extract trips from data generated via multiple positioning technologies (or, multi-sourced data) are absent. And yet, multi-sourced data are now increasingly common. Generated using multiple technologies (e.g., GPS, cellular network- and WiFi-based), multi-sourced data contain high variances in their temporal and spatial properties. In this study, we propose a 'Divide, Conquer and Integrate' (DCI) framework to extract trips from multi-sourced data. We evaluate the proposed framework by applying it to an app-based data, which is multi-sourced and has high variances in both location accuracy and observation interval (i.e. time interval between two consecutive observations). On a manually labeled sample of the app-based data, the framework outperforms the state-of-the-art SVM model that is designed for GPS data. The effectiveness of the framework is also illustrated by consistent mobility patterns obtained from the app-based data and an externally collected household travel survey data for the same region and the same period.

stat.AP↗

arXiv2020

Assessing Impacts of Abnormal Events on Travel Patterns Leveraging Passively Collected Trajectory Data

Travel patterns can be impacted by abnormal events. Assessing the impacts has important implications for relief operations and improving preparedness or planning for future events. Conventionally, the assessment is done followed by data collection from post-event surveys, which are economically costly, suffering low-response rate, time-consuming and usually delayed for months (or even years) after an event, leading to inefficient and unreliable assessment and creating obstacles for relief organizations to reach people in need. Penetration of smartphones and services enabled by them continuously generate large amount of trajectory data (e.g., Call Records Data, App-based data), containing trajectories of massive users. These trajectory data are passively and timely collected and without additional cost and contain information of travel patterns of the massive number of individuals in a region for a prolonged time period (e.g., months to years). We propose a framework to assessing the impacts on travel patterns using these data. Utilizing the passively collected trajectory data, the proposed framework seeks to capturing and understanding the full spectrum of travel pattern changes, which helps to assess who, when and how people in a certain area were impacted. The proposed framework is applied to a mobile phone trajectory dataset containing about half-year trajectories of a million anonymous users to assess the impacts of Hurricane Harvey (the second-costliest hurricane in US history). The results are validated and show that the proposed framework can provide a comprehensive assessment of impacts of Harvey on travel patterns, which could guide the response to and the recovery from the impacts.

physics.soc-ph↗

arXiv2019

Color Recognition for Rubik's Cube Robot

In this paper, we proposed three methods to solve color recognition of Rubik's cube, which includes one offline method and two online methods. Scatter balance \& extreme learning machine (SB-ELM), a offline method, is proposed to illustrate the efficiency of training based method. We also point out the conception of color drifting which indicates offline methods are always ineffectiveness and can not work well in continuous change circumstance. By contrast, dynamic weight label propagation is proposed for labeling blocks color by known center blocks color of Rubik's cube. Furthermore, weak label hierarchic propagation, another online method, is also proposed for unknown all color information but only utilizes weak label of center block in color recognition. We finally design a Rubik's cube robot and construct a dataset to illustrate the efficiency and effectiveness of our online methods and to indicate the ineffectiveness of offline method by color drifting in our dataset.

cs.CV↗

arXiv2018

Active User Detection of Uplink Grant-Free SCMA in Frequency Selective Channel

Massive machine type communication (mMTC) is one of the three fifth generation mobile networking (5G) key usage scenarios, which is characterized by a very large number of connected devices typically transmitting a relatively low volume of non-delay sensitive data. To support the mMTC communication, an uplink (UL) grant-free sparse code multiple access (SCMA) system has been proposed. In this system, the knowledge of user equipments' (UEs') status should be obtained before decoding the data by a message passing algorithm (MPA). An existing solution is to use the compressive sensing (CS) theory to detect active UEs under the assumed condition of flat fading channel. But the assumed condition is not suitable for the frequency selective channel and will decrease the accuracy of active UEs detection. This paper proposes a new simple module named refined active UE detector (RAUD), which is based on frequency selective channel gain analyzing. By making full use of the channel gain and analyzing the difference between characteristic values of the two status of UEs, RAUD module can enhance the active UEs detection accuracy. Meanwhile, the addition of the proposed module has a negligible effect on the complexity of UL grant-free SCMA receiver.

eess.SP↗

arXiv2018

Rough extreme learning machine: a new classification method based on uncertainty measure

Extreme learning machine (ELM) is a new single hidden layer feedback neural network. The weights of the input layer and the biases of neurons in hidden layer are randomly generated, the weights of the output layer can be analytically determined. ELM has been achieved good results for a large number of classification tasks. In this paper, a new extreme learning machine called rough extreme learning machine (RELM) was proposed. RELM uses rough set to divide data into upper approximation set and lower approximation set, and the two approximation sets are utilized to train upper approximation neurons and lower approximation neurons. In addition, an attribute reduction is executed in this algorithm to remove redundant attributes. The experimental results showed, comparing with the comparison algorithms, RELM can get a better accuracy and repeatability in most cases, RELM can not only maintain the advantages of fast speed, but also effectively cope with the classification task for high-dimensional data.

cs.LG↗

arXiv2017

Feature Fusion using Extended Jaccard Graph and Stochastic Gradient Descent for Robot

Robot vision is a fundamental device for human-robot interaction and robot complex tasks. In this paper, we use Kinect and propose a feature graph fusion (FGF) for robot recognition. Our feature fusion utilizes RGB and depth information to construct fused feature from Kinect. FGF involves multi-Jaccard similarity to compute a robust graph and utilize word embedding method to enhance the recognition results. We also collect DUT RGB-D face dataset and a benchmark datset to evaluate the effectiveness and efficiency of our method. The experimental results illustrate FGF is robust and effective to face and object datasets in robot applications.

cs.CV↗

arXiv2017

Neural method for Explicit Mapping of Quasi-curvature Locally Linear Embedding in image retrieval

This paper proposed a new explicit nonlinear dimensionality reduction using neural networks for image retrieval tasks. We first proposed a Quasi-curvature Locally Linear Embedding (QLLE) for training set. QLLE guarantees the linear criterion in neighborhood of each sample. Then, a neural method (NM) is proposed for out-of-sample problem. Combining QLLE and NM, we provide a explicit nonlinear dimensionality reduction approach for efficient image retrieval. The experimental results in three benchmark datasets illustrate that our method can get better performance than other state-of-the-art out-of-sample methods.

cs.CV↗