arXiv subjects

Zongyuan Zhang

Publications and source records attributed to Zongyuan Zhang.

At least 19 recordsLinked to original sources

arXiv2026

Brace Yourself: Task-Conditioned Environmental Bracing for Forceful Humanoid Manipulation

Forceful manipulation is challenging for humanoid robots because interaction forces can disturb whole-body balance. We introduce the Supporting Hand Strategy (SHS), which enables a humanoid to brace against the environment with one hand while performing forceful manipulation with the other. SHS optimises a task-conditioned support configuration that guides two synchronous reinforcement-learning policies, without human motion data or online whole-body trajectory planning. On a Unitree G1, SHS achieved usable contact forces up to 60 N, compared with a maximum sustained force of 13.5 N without environmental bracing, while substantially improving force tracking over a task-independent support configuration. The same policies generalised to different task regions without retraining. SHS therefore provides a simple mechanism for substantially extending humanoid forceful-manipulation capability.

cs.RO↗

arXiv2026

FluxShard: Motion-Aware Feature Cache Reuse for Collaborative Video Analytics in Mobile Edge Computing

Caching and reusing intermediate features across consecutive frames is a common technique to reduce redundant computation and transmission for edge-cloud video analytics in mobile edge computation. Existing methods manage the cache in a fixed or globally shifted coordinate system, treating it as an indivisible whole. Under the non-uniform motion patterns of mobile scenes, this whole-scene granularity invalidates large portions of the cache even when most content has merely shifted spatially, wasting computation and bandwidth. The root cause is a granularity mismatch: the cache is managed per scene, yet motion varies per region. In this paper, we present FluxShard, a motion-aware edge-cloud video analytics system that uses codec-level block motion vectors (MVs) to manage feature cache reuse and recomputation at the granularity of individual motion regions. By re-indexing cached features along per-block MVs, FluxShard separates spatial displacement from content changes, recovering reusable content that whole-scene methods would otherwise discard. To ensure correct reuse under heterogeneous motion, the Receptive Field Alignment Principle (RFAP) identifies, from the input-level MV field alone, the positions that must be recomputed due to inconsistent spatial composition within receptive fields. To maintain cache coherence across frames, MV-guided cache remapping warps the entire feature cache to the current coordinate system each frame, sustaining a high reuse ratio over time. A profiling-driven dispatcher routes the remaining sparse workload between edge and cloud for lower latency. Evaluation across multiple vision tasks, dynamic video benchmarks, and network conditions shows that FluxShard reduces latency by 32.6-83.8% and energy by 14.9-64.0% over all baselines under the prescribed accuracy budget.

cs.NI↗

arXiv2026

Acoustic Feedback for Closed-Loop Force Control in Robotic Grinding

Acoustic feedback is a critical indicator for assessing the contact condition between the tool and the workpiece when humans perform grinding tasks with rotary tools. In contrast, robotic grinding systems typically rely on force sensing, with acoustic information largely ignored. This reliance on force sensors is costly and difficult to adapt to different grinding tools, whereas audio sensors (microphones) are low-cost and can be mounted on any medium that conducts grinding sound. This paper introduces a low-cost Acoustic Feedback Robotic Grinding System (AFRG) that captures audio signals with a contact microphone, estimates grinding force from the audio in real time, and enables closed-loop force control of the grinding process. Compared with conventional force-sensing approaches, AFRG achieves a 4-fold improvement in consistency across different grinding disc conditions. AFRG relies solely on a low-cost microphone, which is approximately 200-fold cheaper than conventional force sensors, as the sensing modality, providing an easily deployable, cost-effective robotic grinding solution.

cs.RO↗

arXiv2026

Transformer-Based Multipath Congestion Control: A Decoupled Approach for Wireless Uplinks

The proliferation of artificial intelligence applications on edge devices necessitates efficient transport protocols that leverage multi-homed connectivity across heterogeneous networks. While Multipath TCP enables bandwidth aggregation, its in-kernel congestion control mechanisms lack the programmability and flexibility needed for achieving efficient transmission. Additionally, inherent measurement noise renders network state partially observable, challenging data-driven approaches like deep reinforcement learning (DRL). To address these challenges, we propose a Transformer-based Congestion Control Optimization (TCCO) framework for multipath transport. TCCO employs a decoupled architecture that offloads control decisions to an external decision engine via a lightweight in-kernel client and user-space proxy, enabling edge devices to leverage external computational resources while maintaining TCP/IP compatibility. The Transformer-based DRL agent in the external decision engine uses self-attention to capture temporal dependencies, filter noise, and coordinate control across subflows through a unified policy. Extensive evaluation on both simulated and real dual-band Wi-Fi testbeds demonstrates that TCCO achieves superior adaptability and performance than state-of-the-art baselines, validating the feasibility and effectiveness of TCCO for wireless networks.

cs.NI↗

arXiv2026

Origin of anomalous p-type conductivity in monolayer Fe-doped MoS2

Substitutional doping effectively modulates carrier polarity of semiconducting two-dimensional (2D) transition metal dichalcogenides (TMDs) like MoS2. Although Fe doping typically induces n-type conductivity in monolayer MoS2, anomalous p-type behavior has also been experimentally reported, the origin of which remains unresolved. Here, we prove that this anomalous p-type conductivity originates from defect associates formed through interactions between Fe dopants and S atoms, which consists of three Fe substituting Mo (FeMo) point defects arranged into an equilateral triangle with a central S atom, denoted as 3FeMo-S associate. Its p-type effect is directly verified through scanning tunneling microscopy/scanning tunneling spectroscopy (STM/STS) measurement, in sharp contrast to the n-type behavior induced by isolated FeMo point defects, and the conclusion is further supported by electrical transport measurements and first-principles calculations. Similar 3FeW-S associates and their p-type doping effect are also identified in monolayer Fe-doped WS2. This work resolves a longstanding controversy and highlights the critical role of defect associates in modulating properties of 2D TMDs.

cond-mat.mtrl-sci↗

arXiv2026

NSC-SL: A Bandwidth-Aware Neural Subspace Compression for Communication-Efficient Split Learning

The expanding scale of neural networks poses a major challenge for distributed machine learning, particularly under limited communication resources. While split learning (SL) alleviates client computational burden by distributing model layers between clients and server, it incurs substantial communication overhead from frequent transmission of intermediate activations and gradients. To tackle this issue, we propose NSC-SL, a bandwidth-aware adaptive compression algorithm for communication-efficient SL. NSC-SL first dynamically determines the optimal rank of low-rank approximation based on the singular value distribution for adapting real-time bandwidth constraints. Then, NSC-SL performs error-compensated tensor factorization using alternating orthogonal iteration with residual feedback, effectively minimizing truncation loss. The collaborative mechanisms enable NSC-SL to achieve high compression ratios while preserving semantic-rich information essential for convergence. Extensive experiments demonstrate the superb performance of NSC-SL.

cs.NI↗

arXiv2025

LLM-Driven Stationarity-Aware Expert Demonstrations for Multi-Agent Reinforcement Learning in Mobile Systems

Multi-agent reinforcement learning (MARL) has been increasingly adopted in many real-world applications. While MARL enables decentralized deployment on resource-constrained edge devices, it suffers from severe non-stationarity due to the synchronous updates of agent policies. This non stationarity results in unstable training and poor policy con vergence, especially as the number of agents increases. In this paper, we propose RELED, a scalable MARL framework that integrates large language model (LLM)-driven expert demonstrations with autonomous agent exploration. RELED incorporates a Stationarity-Aware Expert Demonstration module, which leverages theoretical non-stationarity bounds to enhance the quality of LLM-generated expert trajectories, thus providing high reward and training-stable samples for each agent. Moreover, a Hybrid Expert-Agent Policy Optimization module adaptively balances each agent's learning from both expert-generated and agent-generated trajectories, accelerating policy convergence and improving generalization. Extensive experiments with real city networks based on OpenStreetMap demonstrate that RELED achieves superior performance compared to state-of-the-art MARL methods.

cs.LG↗

arXiv2025

Two Heads are Better than One: Robust Learning Meets Multi-branch Models

Deep neural networks (DNNs) are vulnerable to adversarial examples, in which DNNs are misled to false outputs due to inputs containing imperceptible perturbations. Adversarial training, a reliable and effective method of defense, may significantly reduce the vulnerability of neural networks and becomes the de facto standard for robust learning. While many recent works practice the data-centric philosophy, such as how to generate better adversarial examples or use generative models to produce additional training data, we look back to the models themselves and revisit the adversarial robustness from the perspective of deep feature distribution as an insightful complementarity. In this paper, we propose \textit{Branch Orthogonality adveRsarial Training} (BORT) to obtain state-of-the-art performance with solely the original dataset for adversarial training. To practice our design idea of integrating multiple orthogonal solution spaces, we leverage a simple multi-branch neural network and propose a corresponding loss function, branch-orthogonal loss, to make each solution space of the multi-branch model orthogonal. We evaluate our approach on CIFAR-10, CIFAR-100 and SVHN against $\ell_{\infty}$ norm-bounded perturbations of size $ε= 8/255$, respectively. Exhaustive experiments are conducted to show that our method goes beyond all state-of-the-art methods without any tricks. Compared to all methods that do not use additional data for training, our models achieve 67.3\% and 41.5\% robust accuracy on CIFAR-10 and CIFAR-100 (improving upon the state-of-the-art by +7.23\% and +9.07\%).

cs.CV↗

arXiv2025

LEED: A Highly Efficient and Scalable LLM-Empowered Expert Demonstrations Framework for Multi-Agent Reinforcement Learning

Multi-agent reinforcement learning (MARL) holds substantial promise for intelligent decision-making in complex environments. However, it suffers from a coordination and scalability bottleneck as the number of agents increases. To address these issues, we propose the LLM-empowered expert demonstrations framework for multi-agent reinforcement learning (LEED). LEED consists of two components: a demonstration generation (DG) module and a policy optimization (PO) module. Specifically, the DG module leverages large language models to generate instructions for interacting with the environment, thereby producing high-quality demonstrations. The PO module adopts a decentralized training paradigm, where each agent utilizes the generated demonstrations to construct an expert policy loss, which is then integrated with its own policy loss. This enables each agent to effectively personalize and optimize its local policy based on both expert knowledge and individual experience. Experimental results show that LEED achieves superior sample efficiency, time efficiency, and robust scalability compared to state-of-the-art baselines.

cs.MA↗

arXiv2025

Sample Efficient Experience Replay in Non-stationary Environments

Reinforcement learning (RL) in non-stationary environments is challenging, as changing dynamics and rewards quickly make past experiences outdated. Traditional experience replay (ER) methods, especially those using TD-error prioritization, struggle to distinguish between changes caused by the agent's policy and those from the environment, resulting in inefficient learning under dynamic conditions. To address this challenge, we propose the Discrepancy of Environment Dynamics (DoE), a metric that isolates the effects of environment shifts on value functions. Building on this, we introduce Discrepancy of Environment Prioritized Experience Replay (DEER), an adaptive ER framework that prioritizes transitions based on both policy updates and environmental changes. DEER uses a binary classifier to detect environment changes and applies distinct prioritization strategies before and after each shift, enabling more sample-efficient learning. Experiments on four non-stationary benchmarks demonstrate that DEER further improves the performance of off-policy algorithms by 11.54 percent compared to the best-performing state-of-the-art ER methods.

cs.LG↗

arXiv2025

State-Aware Perturbation Optimization for Robust Deep Reinforcement Learning

Recently, deep reinforcement learning (DRL) has emerged as a promising approach for robotic control. However, the deployment of DRL in real-world robots is hindered by its sensitivity to environmental perturbations. While existing whitebox adversarial attacks rely on local gradient information and apply uniform perturbations across all states to evaluate DRL robustness, they fail to account for temporal dynamics and state-specific vulnerabilities. To combat the above challenge, we first conduct a theoretical analysis of white-box attacks in DRL by establishing the adversarial victim-dynamics Markov decision process (AVD-MDP), to derive the necessary and sufficient conditions for a successful attack. Based on this, we propose a selective state-aware reinforcement adversarial attack method, named STAR, to optimize perturbation stealthiness and state visitation dispersion. STAR first employs a soft mask-based state-targeting mechanism to minimize redundant perturbations, enhancing stealthiness and attack effectiveness. Then, it incorporates an information-theoretic optimization objective to maximize mutual information between perturbations, environmental states, and victim actions, ensuring a dispersed state-visitation distribution that steers the victim agent into vulnerable states for maximum return reduction. Extensive experiments demonstrate that STAR outperforms state-of-the-art benchmarks.

cs.LG↗

arXiv2025

Robust Deep Reinforcement Learning in Robotics via Adaptive Gradient-Masked Adversarial Attacks

Deep reinforcement learning (DRL) has emerged as a promising approach for robotic control, but its realworld deployment remains challenging due to its vulnerability to environmental perturbations. Existing white-box adversarial attack methods, adapted from supervised learning, fail to effectively target DRL agents as they overlook temporal dynamics and indiscriminately perturb all state dimensions, limiting their impact on long-term rewards. To address these challenges, we propose the Adaptive Gradient-Masked Reinforcement (AGMR) Attack, a white-box attack method that combines DRL with a gradient-based soft masking mechanism to dynamically identify critical state dimensions and optimize adversarial policies. AGMR selectively allocates perturbations to the most impactful state features and incorporates a dynamic adjustment mechanism to balance exploration and exploitation during training. Extensive experiments demonstrate that AGMR outperforms state-of-the-art adversarial attack methods in degrading the performance of the victim agent and enhances the victim agent's robustness through adversarial defense mechanisms.

cs.LG↗

arXiv2025

Symmetry-Broken Kondo Screening and Zero-Energy Mode in the Kagome Superconductor CsV3Sb5

The quantum states of matter reorganize themselves in response to defects, giving rise to emergent local excitations that imprint unique characteristics of the host states. While magnetic impurities are known to generate Kondo screening in a Fermi liquid and Yu-Shiba-Rusinov (YSR) states in a conventional superconductor, it remains unclear whether they can evoke distinct phenomena in the kagome superconductor AV3Sb5 (where A is K, Rb or Cs), which may host an orbital-antiferromagnetic charge density wave (CDW) state and an unconventional superconducting state driven by the convergence of topology, geometric frustration and electron correlations. In this work, we visualize the local density of states induced near various types of impurities in both the CDW and superconducting phases of CsV3-xMxSb5 (M = Ta, Cr) using scanning tunneling microscopy. We observe Kondo resonance states near magnetic Cr dopants. Notably, unlike in any known metal or CDW compound, the spatial pattern of Kondo screening breaks all in-plane mirror symmetries of the kagome lattice, suggesting an electronic chirality due to putative orbital loop currents. While Cooper pairs show relative insensitivity to nonmagnetic impurities, native V vacancies with weak magnetic moments induce a pronounced zero-bias conductance peak (ZBCP). This ZBCP coexists with trivial YSR states within the superconducting gap and does not split in energy with increasing tunneling transmission, tending instead to saturate. This behavior is reminiscent of signature of Majorana zero modes, which could be trapped by a sign-change boundary in the superconducting order parameter near a V vacancy, consistent with a surface topological superconducting state. Our findings provide a new approach to exploring novel quantum states on kagome lattices.

cond-mat.supr-con↗

arXiv2025

Rethinking Adversarial Attacks in Reinforcement Learning from Policy Distribution Perspective

Deep Reinforcement Learning (DRL) suffers from uncertainties and inaccuracies in the observation signal in realworld applications. Adversarial attack is an effective method for evaluating the robustness of DRL agents. However, existing attack methods targeting individual sampled actions have limited impacts on the overall policy distribution, particularly in continuous action spaces. To address these limitations, we propose the Distribution-Aware Projected Gradient Descent attack (DAPGD). DAPGD uses distribution similarity as the gradient perturbation input to attack the policy network, which leverages the entire policy distribution rather than relying on individual samples. We utilize the Bhattacharyya distance in DAPGD to measure policy similarity, enabling sensitive detection of subtle but critical differences between probability distributions. Our experiment results demonstrate that DAPGD achieves SOTA results compared to the baselines in three robot navigation tasks, achieving an average 22.03% higher reward drop compared to the best baseline.

cs.LG↗

arXiv2024

HE-Nav: A High-Performance and Efficient Navigation System for Aerial-Ground Robots in Cluttered Environments

Existing AGR navigation systems have advanced in lightly occluded scenarios (e.g., buildings) by employing 3D semantic scene completion networks for voxel occupancy prediction and constructing Euclidean Signed Distance Field (ESDF) maps for collision-free path planning. However, these systems exhibit suboptimal performance and efficiency in cluttered environments with severe occlusions (e.g., dense forests or tall walls), due to limitations arising from perception networks' low prediction accuracy and path planners' high computational overhead. In this paper, we present HE-Nav, the first high-performance and efficient navigation system tailored for AGRs operating in cluttered environments. The perception module utilizes a lightweight semantic scene completion network (LBSCNet), guided by a bird's eye view (BEV) feature fusion and enhanced by an exquisitely designed SCB-Fusion module and attention mechanism. This enables real-time and efficient obstacle prediction in cluttered areas, generating a complete local map. Building upon this completed map, our novel AG-Planner employs the energy-efficient kinodynamic A* search algorithm to guarantee planning is energy-saving. Subsequent trajectory optimization processes yield safe, smooth, dynamically feasible and ESDF-free aerial-ground hybrid paths. Extensive experiments demonstrate that HE-Nav achieved 7x energy savings in real-world situations while maintaining planning success rates of 98% in simulation scenarios. Code and video are available on our project page: https://jmwang0117.github.io/HE-Nav/.

cs.RO↗

arXiv2024

AGRNav: Efficient and Energy-Saving Autonomous Navigation for Air-Ground Robots in Occlusion-Prone Environments

The exceptional mobility and long endurance of air-ground robots are raising interest in their usage to navigate complex environments (e.g., forests and large buildings). However, such environments often contain occluded and unknown regions, and without accurate prediction of unobserved obstacles, the movement of the air-ground robot often suffers a suboptimal trajectory under existing mapping-based and learning-based navigation methods. In this work, we present AGRNav, a novel framework designed to search for safe and energy-saving air-ground hybrid paths. AGRNav contains a lightweight semantic scene completion network (SCONet) with self-attention to enable accurate obstacle predictions by capturing contextual information and occlusion area features. The framework subsequently employs a query-based method for low-latency updates of prediction results to the grid map. Finally, based on the updated map, the hierarchical path planner efficiently searches for energy-saving paths for navigation. We validate AGRNav's performance through benchmarks in both simulated and real-world environments, demonstrating its superiority over classical and state-of-the-art methods. The open-source code is available at https://github.com/jmwang0117/AGRNav.

cs.RO↗

arXiv2021

Three-dimensional charge density wave and robust zero-bias conductance peak inside the superconducting vortex core of a kagome superconductor CsV$_3$Sb$_5$

The transition-metal-based kagome metals provide a versatile platform for correlated topological phases hosting various electronic instabilities. While superconductivity is rare in layered kagome compounds, its interplay with nontrivial topology could offer an engaging space to realize exotic excitations of quasiparticles. Here, we use scanning tunneling microscopy (STM) to study a newly discovered Z$_2$ topological kagome metal CsV$_3$Sb$_5$ with a superconducting ground state. We observe charge modulation associated with the opening of an energy gap near the Fermi level. When across single-unit-cell surface step edges, the intensity of this charge modulation exhibits a π-phase shift, suggesting a three-dimensional 2$\times$2$\times$2 charge density wave ordering. Interestingly, a robust zero-bias conductance peak is observed inside the superconducting vortex core on the Cs 2$\times$2 surfaces that does not split in a large distance when moving away from the vortex center, resembling the Majorana bound states arising from the superconducting Dirac surface states in Bi$_2$Te$_3$/NbSe$_2$ heterostructures. Our findings establish CsV$_3$Sb$_5$ as a promising candidate for realizing exotic excitations at the confluence of nontrivial lattice geometry, topology and multiple electronic orders.

cond-mat.supr-con↗

arXiv2020

Eightfold Fermionic Excitation in a Charge Density Wave Compound

Unconventional quasiparticle excitations in condensed matter systems have become one of the most important research frontiers. Beyond two- and fourfold degenerate Weyl and Dirac fermions, three-, six- and eightfold symmetry protected degeneracies have been predicted however remain challenging to realize in solid state materials. Here, charge density wave compound TaTe4 is proposed to hold eightfold fermionic excitation and Dirac point in energy bands. High quality TaTe4 single crystals are prepared, where the charge density wave is revealed by directly imaging the atomic structure and a pseudogap of about 45 meV on the surface. Shubnikov de-Haas oscillations of TaTe4 are consistent with band structure calculation. Scanning tunneling microscopy reveals atomic step edge states on the surface of TaTe4. This work uncovers that charge density wave is able to induce new topological phases and sheds new light on the novel excitations in condensed matter materials.

cond-mat.mtrl-sci↗