arXiv subjects

Xi Zheng

Publications and source records attributed to Xi Zheng.

At least 19 recordsLinked to original sources

arXiv2026

Health Inquiry with AI: How Empathetic Expression and Conversational Contexts Shape Users' Communicative Acts

As online health information-seeking shifts to conversational AI, high-quality information retrieval increasingly relies on users' ``communicative acts''(proactively sharing and seeking information)---similar to how effective diagnosis and personalized guidance are elicited in patient-clinician communication. Drawing on health communication research, this study examines how a chatbot's modality of empathetic expression (Verbal, Visual, Multimodal) and the conversational context (General, Sensitive, Mental Health) influence these acts through a 2 x 2 x 3 within-subjects experiment (N = 48). The results revealed that while verbal and multimodal empathy significantly increased reply length, communicative acts were largely shaped by conversational context, with Sensitive context triggering more question-asking and Mental Health context leading to heightened concerns, assertive responses, and unprompted information disclosure. Combined with qualitative findings, we discuss design implications for building context-sensitive AI health inquiry systems that can encourage active user participation.

cs.HC↗

arXiv2026

Making Embodied AI Reliable: A Community Agenda from Testing to Formal Verification

Embodied AI systems are increasingly deployed in open-world environments, yet ensuring their reliability remains a fundamental challenge. Drawing on discussions from the AAAI'26 Bridge Program on "Making Embodied AI Reliable with Testing and Formal Verification", this article argues that reliability in embodied AI is inherently a lifecycle assurance problem arising from uncertainty, human interaction, and emergent behaviors across tightly coupled system components. We identify three complementary directions toward reliable embodied AI: (1) trustworthy scenario-based testing supported by validated specifications and meaningful coverage metrics, (2) compositional verification enabled by structured symbolic representations of system behavior and environmental context, and (3) runtime assurance mechanisms capable of adapting to uncertainty and distribution shifts during deployment. Rather than treating these approaches independently, we advocate integrated assurance workflows that connect testing, verification, and runtime adaptation through shared neuro-symbolic representations and continuous feedback across the system lifecycle. Such integration provides a foundation for building trustworthy embodied AI systems that can operate safely and reliably in complex real-world environments.

cs.SE↗

arXiv2026

From Particles to Perils: SVGD-Based Hazardous Scenario Generation for Autonomous Driving Systems Testing

Simulation-based testing of autonomous driving systems (ADS) must uncover realistic and diverse failures in dense, heterogeneous traffic. However, existing search-based seeding methods (e.g., genetic algorithms) struggle in high-dimensional spaces, often collapsing to limited modes and missing many failure scenarios. We present PtoP, a framework that combines adaptive random seed generation with Stein Variational Gradient Descent (SVGD) to produce diverse, failure-inducing initial conditions. SVGD balances attraction toward high-risk regions and repulsion among particles, yielding risk-seeking yet well-distributed seeds across multiple failure modes. PtoP is plug-and-play and enhances existing online testing methods (e.g., reinforcement learning--based testers) by providing principled seeds. Evaluation in CARLA on two industry-grade ADS (Apollo, Autoware) and a native end-to-end system shows that PtoP improves safety violation rate (up to 27.68%), scenario diversity (9.6%), and map coverage (16.78%) over baselines.

cs.SE↗

arXiv2026

An LLM-driven Scenario Generation Pipeline Using an Extended Scenic DSL for Autonomous Driving Safety Validation

Real-world crash reports, which combine textual summaries and sketches, are valuable for scenario-based testing of autonomous driving systems (ADS). However, current methods cannot effectively translate this multimodal data into precise, executable simulation scenarios, hindering the scalability of ADS safety validation. In this work, we propose a scalable and verifiable pipeline that uses a large language model (GPT-4o mini) and a probabilistic intermediate representation (an Extended Scenic domain-specific language) to automatically extract semantic scenario configurations from crash reports and generate corresponding simulation-ready scenarios. Unlike earlier approaches such as ScenicNL and LCTGen (which generate scenarios directly from text) or TARGET (which uses deterministic mappings from traffic rules), our method introduces an intermediate Scenic DSL layer to separate high-level semantic understanding from low-level scenario rendering, reducing errors and capturing real-world variability. We evaluated the pipeline on cases from the NHTSA CIREN database. The results show high accuracy in knowledge extraction: 100% correctness for environmental and road network attributes, and 97% and 98% for oracle and actor trajectories, respectively, compared to human-derived ground truth. We executed the generated scenarios in the CARLA simulator using the Autoware driving stack, and they consistently triggered the intended traffic-rule violations (such as opposite-lane crossing and red-light running) across 2,000 scenario variations. These findings demonstrate that the proposed pipeline provides a legally grounded, scalable, and verifiable approach to ADS safety validation.

cs.SE↗

arXiv2026

Visual Marker Search for Autonomous Drone Landing in Diverse Urban Environments

Marker-based landing is widely used in drone delivery and return-to-base systems for its simplicity and reliability. However, most approaches assume idealized landing site visibility and sensor performance, limiting robustness in complex urban settings. We present a simulation-based evaluation suite on the AirSim platform with systematically varied urban layouts, lighting, and weather to replicate realistic operational diversity. Using onboard camera sensors (RGB for marker detection and depth for obstacle avoidance), we benchmark two heuristic coverage patterns and a reinforcement learning-based agent, analyzing how exploration strategy and scene complexity affect success rate, path efficiency, and robustness. Results underscore the need to evaluate marker-based autonomous landing under diverse, sensor-relevant conditions to guide the development of reliable aerial navigation systems.

cs.RO↗

arXiv2025

Towards autonomous normative multi-agent systems for Human-AI software engineering teams

This paper envisions a transformative paradigm in software engineering, where Artificial Intelligence, embodied in fully autonomous agents, becomes the primary driver of the core software development activities. We introduce a new class of software engineering agents, empowered by Large Language Models and equipped with beliefs, desires, intentions, and memory to enable human-like reasoning. These agents collaborate with humans and other agents to design, implement, test, and deploy software systems with a level of speed, reliability, and adaptability far beyond the current software development processes. Their coordination and collaboration are governed by norms expressed as deontic modalities - commitments, obligations, prohibitions and permissions - that regulate interactions and ensure regulatory compliance. These innovations establish a scalable, transparent and trustworthy framework for future Human-AI software engineering teams.

cs.SE↗

arXiv2025

Rethinking Data Value: Asymmetric Data Shapley for Structure-Aware Valuation in Data Markets and Machine Learning Pipelines

Rigorous valuation of individual data sources is critical for fair compensation in data markets, informed data acquisition, and transparent development of ML/AI models. Classical Data Shapley (DS) provides a essential axiomatic framework for data valuation but is constrained by its symmetry axiom that assumes interchangeability of data sources. This assumption fails to capture the directional and temporal dependencies prevalent in modern ML/AI workflows, including the reliance of duplicated or augmented data on original sources and the order-specific contributions in sequential pipelines such as federated learning and multi-stage LLM fine tuning. To address these limitations, we introduce Asymmetric Data Shapley (ADS), a structure-aware data valuation framework for modern ML/AI pipelines. ADS relaxes symmetry by averaging marginal contributions only over permutations consistent with an application-specific ordering of data groups. It preserves efficiency and linearity, maintains within group symmetry and directional precedence across groups, and reduces to DS when the ordering collapses to a single group. We develop two complementary computational procedures for ADS: (i) a Monte Carlo estimator (MC-ADS) with finite-sample accuracy guarantees, and (ii) a k-nearest neighbor surrogate (KNN-ADS) that is exact and efficient for KNN predictors. Across representative settings with directional and temporal dependence, ADS consistently outperforms benchmark methods by distinguishing novel from redundant contributions and respecting the sequential nature of training. These results establish ADS as a principled and practical approach to equitable data valuation in data markets and complex ML/AI pipelines.

cs.GT↗

arXiv2025

Experiences from Benchmarking Vision-Language-Action Models for Robotic Manipulation

Foundation models applied in robotics, particularly \textbf{Vision--Language--Action (VLA)} models, hold great promise for achieving general-purpose manipulation. Yet, systematic real-world evaluations and cross-model comparisons remain scarce. This paper reports our \textbf{empirical experiences} from benchmarking four representative VLAs -- \textbf{ACT}, \textbf{OpenVLA--OFT}, \textbf{RDT-1B}, and \boldmath{$\pi_0$} -- across four manipulation tasks conducted in both simulation and on the \textbf{ALOHA Mobile} platform. We establish a \textbf{standardized evaluation framework} that measures performance along three key dimensions: (1) \textit{accuracy and efficiency} (success rate and time-to-success), (2) \textit{adaptability} across in-distribution, spatial out-of-distribution, and instance-plus-spatial out-of-distribution settings, and (3) \textit{language instruction-following accuracy}. Through this process, we observe that \boldmath{$\pi_0$} demonstrates superior adaptability in out-of-distribution scenarios, while \textbf{ACT} provides the highest stability in-distribution. Further analysis highlights differences in computational demands, data-scaling behavior, and recurring failure modes such as near-miss grasps, premature releases, and long-horizon state drift. These findings reveal practical trade-offs among VLA model architectures in balancing precision, generalization, and deployment cost, offering actionable insights for selecting and deploying VLAs in real-world robotic manipulation tasks.

cs.RO↗

arXiv2025

Human-Inspired Neuro-Symbolic World Modeling and Logic Reasoning for Interpretable Safe UAV Landing Site Assessment

Reliable assessment of safe landing sites in unstructured environments is essential for deploying Unmanned Aerial Vehicles (UAVs) in real-world applications such as delivery, inspection, and surveillance. Existing learning-based approaches often degrade under covariate shift and offer limited transparency, making their decisions difficult to interpret and validate on resource-constrained platforms. We present NeuroSymLand, a neuro-symbolic framework for marker-free UAV landing site safety assessment that explicitly separates perception-driven world modeling from logic-based safety reasoning. A lightweight segmentation model incrementally constructs a probabilistic semantic scene graph encoding objects, attributes, and spatial relations. Symbolic safety rules, synthesized offline via large language models with human-in-the-loop refinement, are executed directly over this world model at runtime to perform white-box reasoning, producing ranked landing candidates with human-readable explanations of the underlying safety constraints. Across 72 simulated and hardware-in-the-loop landing scenarios, NeuroSymLand achieves 61 successful assessments, outperforming four competitive baselines, which achieve between 37 and 57 successes. Qualitative analysis highlights its superior interpretability and transparent reasoning, while deployment incurs negligible edge overhead. Our results suggest that combining explicit world modeling with symbolic reasoning can support accurate, interpretable, and edge-deployable safety assessment in mobile systems, as demonstrated through UAV landing site assessment.

cs.RO↗

arXiv2025

AutoMT: A Multi-Agent LLM Framework for Automated Metamorphic Testing of Autonomous Driving Systems

Autonomous Driving Systems (ADS) are safety-critical, where failures can be severe. While Metamorphic Testing (MT) is effective for fault detection in ADS, existing methods rely heavily on manual effort and lack automation. We present AutoMT, a multi-agent MT framework powered by Large Language Models (LLMs) that automates the extraction of Metamorphic Relations (MRs) from local traffic rules and the generation of valid follow-up test cases. AutoMT leverages LLMs to extract MRs from traffic rules in Gherkin syntax using a predefined ontology. A vision-language agent analyzes scenarios, and a search agent retrieves suitable MRs from a RAG-based database to generate follow-up cases via computer vision. Experiments show that AutoMT achieves up to 5 x higher test diversity in follow-up case generation compared to the best baseline (manual expert-defined MRs) in terms of validation rate, and detects up to 20.55% more behavioral violations. While manual MT relies on a fixed set of predefined rules, AutoMT automatically extracts diverse metamorphic relations that augment real-world datasets and help uncover corner cases often missed during in-field testing and data collection. Its modular architecture separating MR extraction, filtering, and test generation supports integration into industrial pipelines and potentially enables simulation-based testing to systematically cover underrepresented or safety-critical scenarios.

cs.SE↗

arXiv2025

Diffusion Models for Reinforcement Learning: Foundations, Taxonomy, and Development

Diffusion Models (DMs), as a leading class of generative models, offer key advantages for reinforcement learning (RL), including multi-modal expressiveness, stable training, and trajectory-level planning. This survey delivers a comprehensive and up-to-date synthesis of diffusion-based RL. We first provide an overview of RL, highlighting its challenges, and then introduce the fundamental concepts of DMs, investigating how they are integrated into RL frameworks to address key challenges in this research field. We establish a dual-axis taxonomy that organizes the field along two orthogonal dimensions: a function-oriented taxonomy that clarifies the roles DMs play within the RL pipeline, and a technique-oriented taxonomy that situates implementations across online versus offline learning regimes. We also provide a comprehensive examination of this progression from single-agent to multi-agent domains, thereby forming several frameworks for DM-RL integration and highlighting their practical utility. Furthermore, we outline several categories of successful applications of diffusion-based RL across diverse domains, discuss open research issues of current methodologies, and highlight key directions for future research to advance the field. Finally, we summarize the survey to identify promising future development directions. We are actively maintaining a GitHub repository (https://github.com/ChangfuXu/D4RL-FTD) for papers and other related resources to apply DMs for RL.

cs.LG↗

arXiv2025

RF-LSCM: Pushing Radiance Fields to Multi-Domain Localized Statistical Channel Modeling for Cellular Network Optimization

Accurate localized wireless channel modeling is a cornerstone of cellular network optimization, enabling reliable prediction of network performance during parameter tuning. Localized statistical channel modeling (LSCM) is the state-of-the-art channel modeling framework tailored for cellular network optimization. However, traditional LSCM methods, which infer the channel's Angular Power Spectrum (APS) from Reference Signal Received Power (RSRP) measurements, suffer from critical limitations: they are typically confined to single-cell, single-grid and single-carrier frequency analysis and fail to capture complex cross-domain interactions. To overcome these challenges, we propose RF-LSCM, a novel framework that models the channel APS by jointly representing large-scale signal attenuation and multipath components within a radiance field. RF-LSCM introduces a multi-domain LSCM formulation with a physics-informed frequency-dependent Attenuation Model (FDAM) to facilitate the cross frequency generalization as well as a point-cloud-aided environment enhanced method to enable multi-cell and multi-grid channel modeling. Furthermore, to address the computational inefficiency of typical neural radiance fields, RF-LSCM leverages a low-rank tensor representation, complemented by a novel Hierarchical Tensor Angular Modeling (HiTAM) algorithm. This efficient design significantly reduces GPU memory requirements and training time while preserving fine-grained accuracy. Extensive experiments on real-world multi-cell datasets demonstrate that RF-LSCM significantly outperforms state-of-the-art methods, achieving up to a 30% reduction in mean absolute error (MAE) for coverage prediction and a 22% MAE improvement by effectively fusing multi-frequency data.

cs.LG↗

arXiv2025

Neural Field-Based 3D Surface Reconstruction of Microstructures from Multi-Detector Signals in Scanning Electron Microscopy

The 3D characterization of microstructures is crucial for understanding and designing functional materials. However, the scanning electron microscope (SEM), widely used in scientific research, captures only 2D electron intensity distributions. Existing SEM 3D reconstruction methods struggle with textureless regions, shadowing artifacts, and calibration dependencies, whereas advanced learning-based approaches fail to generalize to microscopic SEM domains due to the lack of physical priors and domain-specific data. We introduce NFH-SEM, a neural field-based hybrid framework that reconstructs high-fidelity 3D surfaces from multi-view, multi-detector SEM images. NFH-SEM integrates coarse multi-view geometry with photometric stereo cues from detector signals through a continuous neural field, incorporating a learnable forward model that embeds SEM imaging physics for self-calibrated, shadow-robust reconstruction. NFH-SEM achieves precise recovery across diverse specimens, revealing 478 nm layered features in two-photon lithography samples, 782 nm surface textures on pollen grains, and 1.559 $\mu$m fracture steps on silicon carbide particles, demonstrating its accuracy and broad applicability. Our code and real-world dataset are available at https://github.com/zju3dv/NFH-SEM.

eess.IV↗

arXiv2025

Policy Learning under Unobserved Confounding: A Robust and Efficient Approach

This paper develops a robust and efficient method for policy learning from observational data in the presence of unobserved confounding, complementing existing instrumental variable (IV) based approaches. We employ the marginal sensitivity model (MSM) to relax the commonly used yet restrictive unconfoundedness assumption by introducing a sensitivity parameter that captures the extent of selection bias induced by unobserved confounders. Building on this framework, we consider two distributionally robust welfare criteria, defined as the worst-case welfare and policy improvement functions, evaluated over an uncertainty set of counterfactual distributions characterized by the MSM. Closed-form expressions for both welfare criteria are derived. Leveraging these identification results, we construct doubly robust scores and estimate the robust policies by maximizing the proposed criteria. Our approach accommodates flexible machine learning methods for estimating nuisance components, even when these converge at moderately slow rates. We establish asymptotic regret bounds for the resulting policies, providing a robust guarantee against the most adversarial confounding scenario. The proposed method is evaluated through extensive simulation studies and empirical applications to the JTPA study and Head Start program.

econ.EM↗

arXiv2025

Generating Adversarial Point Clouds Using Diffusion Model

Adversarial attack methods for 3D point cloud classification reveal the vulnerabilities of point cloud recognition models. This vulnerability could lead to safety risks in critical applications that use deep learning models, such as autonomous vehicles. To uncover the deficiencies of these models, researchers can evaluate their security through adversarial attacks. However, most existing adversarial attack methods are based on white-box attacks. While these methods achieve high attack success rates and imperceptibility, their applicability in real-world scenarios is limited. Black-box attacks, which are more meaningful in real-world scenarios, often yield poor results. This paper proposes a novel black-box adversarial example generation method that utilizes a diffusion model to improve the attack success rate and imperceptibility in the black-box setting, without relying on the internal information of the point cloud classification model to generate adversarial samples. We use a 3D diffusion model to use the compressed features of the point cloud as prior knowledge to guide the reverse diffusion process to add adversarial points to clean examples. Subsequently, its reverse process is employed to transform the distribution of other categories into adversarial points, which are then added to the point cloud.

cs.CR↗

arXiv2025

Regression Testing Optimization for ROS-based Autonomous Systems: A Comprehensive Review of Techniques

Regression testing plays a critical role in maintaining software reliability, particularly for ROS-based autonomous systems (ROSAS), which frequently undergo continuous integration and iterative development. However, conventional regression testing techniques face significant challenges when applied to autonomous systems due to their dynamic and non-deterministic behaviors, complex multi-modal sensor data, asynchronous distributed architectures, and stringent safety and real-time constraints. Although numerous studies have explored test optimization in traditional software contexts, regression testing optimization specifically for ROSAS remains largely unexplored. To address this gap, we present the first comprehensive survey systematically reviewing regression testing optimization techniques tailored for ROSAS. We analyze and categorize 122 representative studies into regression test case prioritization, minimization, and selection methods. A structured taxonomy is introduced to clearly illustrate their applicability and limitations within ROSAS contexts. Furthermore, we highlight major challenges specific to regression testing for ROSAS, including effectively prioritizing tests in response to frequent system modifications, efficiently minimizing redundant tests, and difficulty in accurately selecting impacted test cases. Finally, we propose research insights and identify promising future directions, such as leveraging frame-to-vector coverage metrics, multi-source foundation models, and neurosymbolic reasoning to enhance regression testing efficiency and effectiveness. This survey provides a foundational reference and practical roadmap for advancing the state-of-the-art in regression testing optimization for ROSAS.

cs.SE↗

arXiv2025

A Step-by-Step Guide to Creating a Robust Autonomous Drone Testing Pipeline

Autonomous drones are rapidly reshaping industries ranging from aerial delivery and infrastructure inspection to environmental monitoring and disaster response. Ensuring the safety, reliability, and efficiency of these systems is paramount as they transition from research prototypes to mission-critical platforms. This paper presents a step-by-step guide to establishing a robust autonomous drone testing pipeline, covering each critical stage: Software-in-the-Loop (SIL) Simulation Testing, Hardware-in-the-Loop (HIL) Testing, Controlled Real-World Testing, and In-Field Testing. Using practical examples, including the marker-based autonomous landing system, we demonstrate how to systematically verify drone system behaviors, identify integration issues, and optimize performance. Furthermore, we highlight emerging trends shaping the future of drone testing, including the integration of Neurosymbolic and LLMs, creating co-simulation environments, and Digital Twin-enabled simulation-based testing techniques. By following this pipeline, developers and researchers can achieve comprehensive validation, minimize deployment risks, and prepare autonomous drones for safe and reliable real-world operations.

cs.SE↗

arXiv2025

Towards Robust Autonomous Landing Systems: Iterative Solutions and Key Lessons Learned

Uncrewed Aerial Vehicles (UAVs) have become a focal point of research, with both established companies and startups investing heavily in their development. This paper presents our iterative process in developing a robust autonomous marker-based landing system, highlighting the key challenges encountered and the solutions implemented. It reviews existing systems for autonomous landing processes, and through this aims to contribute to the community by sharing insights and challenges faced during development and testing.

cs.RO↗