arXiv subjects

Xiangwei Zhu

Publications and source records attributed to Xiangwei Zhu.

10 recordsLinked to original sources

arXiv2026

CycleRL: Sim-to-Real Deep Reinforcement Learning for Robust Autonomous Bicycle Control

Autonomous bicycles offer a promising agile solution for urban mobility and last-mile logistics. However, conventional control strategies often struggle with underactuated nonlinear dynamics, suffering from sensitivity to model mismatches and limited adaptability to real-world uncertainties. To address this, we develop CycleRL, a comprehensive sim-to-real framework for robust autonomous bicycle control. Our approach establishes a direct perception-to-action mapping within the high-fidelity NVIDIA Isaac Sim environment, leveraging Proximal Policy Optimization (PPO) to optimize the control policy. The framework features a composite reward function tailored for concurrent balance maintenance, velocity tracking, and steering control. Crucially, systematic domain randomization is employed to reduce the reliance on precise system modeling, bridge the simulation-to-reality gap and facilitate direct transfer. In simulation, CycleRL achieves promising performance, including a 99.90% balance success rate, a heading tracking error of 1.15{\deg}, and a velocity tracking error of 0.18 m/s. These quantitative results, coupled with successful hardware deployment, validate DRL as an effective paradigm for autonomous bicycle control, offering superior adaptability over traditional methods. Video demonstrations are available at https://cpnt-lab.github.io/CycleRL/.

cs.RO↗

arXiv2026

A molecular-spin photovoltaic device

We fabricated a C60-based molecular spin photovoltaic device that integrated a photovoltaic response with the spin transport across the molecular layer. The photovoltaic response can be modified under the application of a small magnetic field, with a magnetophotovoltage of up to 5% at room temperature. Device functionalities include a magnetic current inverter and the presence of diverging magnetocurrent at certain illumination levels that could be useful for sensing. Completely spin-polarized currents could be created by balancing the external partially spin polarized injection with the photogenerated carriers.

cond-mat.mtrl-sci↗

arXiv2025

Fully analogue in-memory neural computing via quantum tunneling effect

Fully analogue neural computation requires hardware that can implement both linear and nonlinear transformations without digital assistance. While analogue in-memory computing efficiently realizes matrix-vector multiplication, the absence of learnable analogue nonlinearities remains a central bottleneck. Here we introduce KANalogue, a fully analogue realization of Kolmogorov-Arnold Networks (KANs) that instantiates univariate basis functions directly using negative-differential-resistance (NDR) devices. By mapping the intrinsic current-voltage characteristics of NDR devices to learnable coordinate-wise nonlinear functions, KANalogue embeds function approximation into device physics while preserving a fully analogue signal path. Using cold-metal tunnel diodes as a representative platform, we construct diverse nonlinear bases and combine them through crossbar-based analogue summation. Experiments on MNIST, FashionMNIST, and CIFAR-10 demonstrate that KANalogue achieves competitive accuracy with substantially fewer parameters and higher crossbar node efficiency than analogue MLPs, while approaching the performance of digital KANs under strict hardware constraints. The framework is not limited to a specific device technology and naturally generalizes to a broad class of NDR devices. These results establish a device-grounded route toward scalable, energy-efficient, fully analogue neural networks.

cs.ET↗

arXiv2025

FLYINGTRUST: A Benchmark for Quadrotor Navigation Across Scenarios and Vehicles

Visual navigation algorithms for quadrotors often exhibit a large variation in performance when transferred across different vehicle platforms and scene geometries, which increases the cost and risk of field deployment. To support systematic early-stage evaluation, we introduce FLYINGTRUST, a high-fidelity, configurable benchmarking framework that measures how platform kinodynamics and scenario structure jointly affect navigation robustness. FLYINGTRUST models vehicle capability with two compact, physically interpretable indicators: maximum thrust-to-weight ratio and axis-wise maximum angular acceleration. The benchmark pairs a diverse scenario library with a heterogeneous set of real and virtual platforms and prescribes a standardized evaluation protocol together with a composite scoring method that balances scenario importance, platform importance and performance stability. We use FLYINGTRUST to compare representative optimization-based and learning-based navigation approaches under identical conditions, performing repeated trials per platform-scenario combination and reporting uncertainty-aware metrics. The results reveal systematic patterns: navigation success depends predictably on platform capability and scene geometry, and different algorithms exhibit distinct preferences and failure modes across the evaluated conditions. These observations highlight the practical necessity of incorporating both platform capability and scenario structure into algorithm design, evaluation, and selection, and they motivate future work on methods that remain robust across diverse platforms and scenarios.

cs.RO↗

arXiv2025

CrossEarth: Geospatial Vision Foundation Model for Domain Generalizable Remote Sensing Semantic Segmentation

The field of Remote Sensing Domain Generalization (RSDG) has emerged as a critical and valuable research frontier, focusing on developing models that generalize effectively across diverse scenarios. Despite the substantial domain gaps in RS images that are characterized by variabilities such as location, wavelength, and sensor type, research in this area remains underexplored: (1) Current cross-domain methods primarily focus on Domain Adaptation (DA), which adapts models to predefined domains rather than to unseen ones; (2) Few studies targeting the RSDG issue, especially for semantic segmentation tasks, where existing models are developed for specific unknown domains, struggling with issues of underfitting on other unknown scenarios; (3) Existing RS foundation models tend to prioritize in-domain performance over cross-domain generalization. To this end, we introduce the first vision foundation model for RSDG semantic segmentation, CrossEarth. CrossEarth demonstrates strong cross-domain generalization through a specially designed data-level Earth-Style Injection pipeline and a model-level Multi-Task Training pipeline. In addition, for the semantic segmentation task, we have curated an RSDG benchmark comprising 32 cross-domain settings across various regions, spectral bands, platforms, and climates, providing a comprehensive framework for testing the generalizability of future RSDG models. Extensive experiments on this benchmark demonstrate the superiority of CrossEarth over existing state-of-the-art methods.

cs.CV↗

arXiv2025

Clock Synchronization for Drone-Based Entanglement Quantum Key Distribution

Drone-based entanglement distribution provides full spatiotemporal coverage for quantum networks, enabling quantum key distribution (QKD) in dynamic environments. The security of QKD fundamentally depends on high-fidelity quantum state measurements, for which high-precision clock synchronization is indispensable, as timing jitter is inversely correlated with quantum state fidelity. However, drone-based clock synchronization is constrained by SWaP (Size, Weight, and Power) limitations and dynamic mobility effects. Here, we propose a synchronization protocol for drone-based entanglement distribution, leveraging nanosecond-accurate Global Navigation Satellite System (GNSS) timing and entanglement-based timing correction to overcome SWaP constraints. Experimental results demonstrate 24 ps RMS synchronization in simulated free-space quantum channels with distance dynamics, without requiring precision reference clock. Our protocol enables drone-based entanglement distribution, paving the way for seamless wide-area and local-area quantum internet.

quant-ph↗

arXiv2025

Multi-Grained Query-Guided Set Prediction Network for Grounded Multimodal Named Entity Recognition

Grounded Multimodal Named Entity Recognition (GMNER) is an emerging information extraction (IE) task, aiming to simultaneously extract entity spans, types, and corresponding visual regions of entities from given sentence-image pairs data. Recent unified methods employing machine reading comprehension or sequence generation-based frameworks show limitations in this difficult task. The former, utilizing human-designed type queries, struggles to differentiate ambiguous entities, such as Jordan (Person) and off-White x Jordan (Shoes). The latter, following the one-by-one decoding order, suffers from exposure bias issues. We maintain that these works misunderstand the relationships of multimodal entities. To tackle these, we propose a novel unified framework named Multi-grained Query-guided Set Prediction Network (MQSPN) to learn appropriate relationships at intra-entity and inter-entity levels. Specifically, MQSPN explicitly aligns textual entities with visual regions by employing a set of learnable queries to strengthen intra-entity connections. Based on distinct intra-entity modeling, MQSPN reformulates GMNER as a set prediction, guiding models to establish appropriate inter-entity relationships from a optimal global matching perspective. Additionally, we incorporate a query-guided Fusion Net (QFNet) as a glue network to boost better alignment of two-level relationships. Extensive experiments demonstrate that our approach achieves state-of-the-art performances in widely used benchmarks.

cs.IR↗

arXiv2024

CoDA: Instructive Chain-of-Domain Adaptation with Severity-Aware Visual Prompt Tuning

Unsupervised Domain Adaptation (UDA) aims to adapt models from labeled source domains to unlabeled target domains. When adapting to adverse scenes, existing UDA methods fail to perform well due to the lack of instructions, leading their models to overlook discrepancies within all adverse scenes. To tackle this, we propose CoDA which instructs models to distinguish, focus, and learn from these discrepancies at scene and image levels. Specifically, CoDA consists of a Chain-of-Domain (CoD) strategy and a Severity-Aware Visual Prompt Tuning (SAVPT) mechanism. CoD focuses on scene-level instructions to divide all adverse scenes into easy and hard scenes, guiding models to adapt from source to easy domains with easy scene images, and then to hard domains with hard scene images, thereby laying a solid foundation for whole adaptations. Building upon this foundation, we employ SAVPT to dive into more detailed image-level instructions to boost performance. SAVPT features a novel metric Severity that divides all adverse scene images into low-severity and high-severity images. Then Severity directs visual prompts and adapters, instructing models to concentrate on unified severity features instead of scene-specific features, without adding complexity to the model architecture. CoDA achieves SOTA performances on widely-used benchmarks under all adverse scenes. Notably, CoDA outperforms the existing ones by 4.6%, and 10.3% mIoU on the Foggy Driving, and Foggy Zurich benchmarks, respectively. Our code is available at https://github.com/Cuzyoung/CoDA

cs.CV↗

arXiv2024

Quantum Random Number Generation Based on Phase Reconstruction

Quantum random number generator (QRNG) utilizes the intrinsic randomness of quantum systems to generate completely unpredictable and genuine random numbers, finding wide applications across many fields. QRNGs relying on the phase noise of a laser have attracted considerable attention due to their straightforward system architecture and high random number generation rates. However, traditional phase noise QRNGs suffer from a 50\% loss of quantum entropy during the randomness extraction process. In this paper, we propose a phase-reconstruction quantum random number generation scheme, in which the phase noise of a laser is reconstructed by simultaneously measuring the orthogonal quadratures of the light field using balanced detectors. This enables direct discretization of uniform phase noise, and the min-entropy can achieve a value of 1. Furthermore, our approach exhibits inherent robustness against the classical phase fluctuations of the unbalanced interferometer, eliminating the need for active compensation. Finally, we conducted experimental validation using commercial optical hybrid and balanced detectors, achieving a random number generation rate of 1.96 Gbps at a sampling rate of 200 MSa/s.

quant-ph↗

arXiv2023

FGO-ILNS: Tightly Coupled Multi-Sensor Integrated Navigation System Based on Factor Graph Optimization for Autonomous Underwater Vehicle

Multi-sensor fusion is an effective way to enhance the positioning performance of autonomous underwater vehicles (AUVs). However, underwater multi-sensor fusion faces challenges such as heterogeneous frequency and dynamic availability of sensors. Traditional filter-based algorithms suffer from low accuracy and robustness when sensors become unavailable. The factor graph optimization (FGO) can enable multi-sensor plug-and-play despite data frequency. Therefore, we present an FGO-based strapdown inertial navigation system (SINS) and long baseline location (LBL) system tightly coupled navigation system (FGO-ILNS). Sensors such as Doppler velocity log (DVL), magnetic compass pilot (MCP), pressure sensor (PS), and global navigation satellite system (GNSS) can be tightly coupled with FGO-ILNS to satisfy different navigation scenarios. In this system, we propose a floating LBL slant range difference factor model tightly coupled with IMU preintegration factor to achieve unification of global position above and below water. Furthermore, to address the issue of sensor measurements not being synchronized with the LBL during fusion, we employ forward-backward IMU preintegration to construct sensor factors such as GNSS and DVL. Moreover, we utilize the marginalization method to reduce the computational load of factor graph optimization. Simulation and public KAIST dataset experiments have verified that, compared to filter-based algorithms like the extended Kalman filter and federal Kalman filter, as well as the state-of-the-art optimization-based algorithm ORB-SLAM3, our proposed FGO-ILNS leads in accuracy and robustness.

cs.RO↗