arXiv subjects

Ziyu Li

Publications and source records attributed to Ziyu Li.

At least 19 recordsLinked to original sources

arXiv2026

BIFTA: Brain-Inspired Few-Shot Tactile Adaptation for Unknown Sensors

Advances in tactile sensing have made contact-rich perception possible, accelerating progress in robotic manipulation, material understanding, and embodied interaction. However, because optical design, elastomer mechanics, and imaging geometry differ substantially across tactile sensors, models trained on known sensor types can suffer an abrupt performance collapse on unknown sensors. To address this problem, we propose the Brain-Inspired Few-Shot Tactile Adaptation (BIFTA) framework; it draws on the brain's rapid sensory adaptation mechanism to adapt a frozen encoder to an unknown tactile sensor from a small labeled support set. BIFTA preserves pretrained representations through dual-view statistical memory, constructs support-conditioned spectral graphs to repair sensor-dependent feature neighborhoods, and applies uncertainty-gated recurrent propagation to strengthen reliable cross-query evidence. Extensive benchmarks across three tactile datasets show that BIFTA substantially improves adaptation to unknown sensors: with only 10\% labeled target data on SITR, it raises mean Sparsh accuracy from 6.86\% for the frozen source classifier to 87.09\%, exceeding the strongest implemented prior comparison by 47.22 percentage points, and these gains generalize across datasets, pretrained backbones, and tactile tasks. These results validate BIFTA for data-efficient adaptation to unknown tactile sensors and offer a promising route toward tactile models that transfer across heterogeneous hardware.

cs.RO↗

arXiv2026

Not All EEG Moments Are Equal: Position-Adaptive Time Scheduling for EEG Generation

Electroencephalography (EEG) generation is essential for alleviating data scarcity and enabling large scale neural modeling in brain computer interface applications. However, existing flow based approaches assume that every channel and every time segment within a sample shares a single global time progression, overlooking the fact that not all EEG moments are equal. To address this overlooked heterogeneity, we propose an adaptive EEG generation framework built on conditional flow matching. The framework introduces Position-Adaptive Time Scheduling, which tracks per position reconstruction error to modulate a position specific time progress within the flow matching trajectory. It further incorporates Factorized Spatio-Temporal Attention and a frequency aligned multi resolution spectral consistency loss to model inter channel dependencies induced by volume conduction and compensate for the power law spectral bias of EEG, thereby improving the quality of generated signals. Extensive experiments on three EEG datasets with distinct acquisition protocols and task semantics show that our framework consistently outperforms the strongest baseline, reducing TS-FID by up to 62.2\% and improving downstream classification accuracy gain by up to 6.77 percentage points. These results suggest that the proposed method represents a promising step toward scalable, high fidelity data augmentation for real world brain computer interface applications.

eess.SP↗

arXiv2026

A Hadamard Formula for Equilibrium Envelopes under Parallel Deformation

Let $(X,\omega_0)$ be a compact K\"ahler manifold, and let $U\Subset X$ have $C^{3,1}$ uniformly strongly pseudoconvex boundary. For the equilibrium envelope $u_0$ associated with $X\setminus U$, the normalized Monge--Amp\`ere measure vanishes on $U$ and equals the background measure $V^{-1}\omega_0^n$ on $X\setminus\overline U$; its remaining component is a singular measure supported on $\partial U$, where $V=\int_X\omega_0^n$. We identify this boundary component as the negative outward Anzellotti trace of a divergence-measure flux current. We then prove a one-sided Hadamard formula for the normalized Monge--Amp\`ere energy along the outward parallel family $U_\varepsilon=\{\rho<\varepsilon\}$. The nonlinear telescoping identity gives the mixed Bedford--Taylor boundary traces that sum up to the trace of the current \[ \mathcal J_{\mathrm{tot}} =\frac1V\mathrm{d}^c u_0\wedge\sum_{p=0}^{n-1}(p+1)\omega_{u_0}^p\wedge\omega_0^{n-1-p}, \qquad \omega_{u_0}=\omega_0+\mathrm{d}\mathrm{d}^c u_0. \] These results provide a local weak formulation of boundary flux and normal variation for regular interface problems related to Darcy/Hele--Shaw type problems and Monge--Amp\`ere growth.

math.CV↗

arXiv2026

Cuprate d-wave superconductivity based on Non perturbative many body theory for spin fluctuation

We employ the nonperturbative many-body spin GW method to compute the single-particle Green's function, and the covariance method to evaluate the two-body correlation functions, for the Hubbard model of superconductors. The method, benchmarked at strong coupling and low temperatures, allows simulations on lattices up to $64\times64$, thereby enabling access to the thermodynamic limit. We find the key transition temperatures---namely the N\'eel temperature and the superconducting critical temperature $T_c$---to be in quantitative agreement with experimental data. Above the $d$-wave superconducting dome, there exists a broad pseudogap region in which the normal-state excitations are suppressed. These results indicate our approach may offer a new route toward investigating strongly correlated systems, especially cuprate superconductors.

cond-mat.str-el↗

arXiv2026

A physics-informed foundation model for quantitative diffusion MRI

Understanding the human brain requires access to its microscopic tissue architecture. Diffusion magnetic resonance imaging (MRI) provides the only noninvasive window into whole-brain microstructure in vivo, yet reliable quantitative mapping remains confined to specialized research settings requiring dense sampling and optimized acquisition protocols. To address this gap, we present a physics-informed generative microstructure network (PIGMENT) that learns a universal generative prior of human brain microstructure and adapts it zero-shot to each participant's measured data to recover subject-specific maps. Trained on 11375 scans spanning multiple sites, vendors, and field strengths, PIGMENT enabled reliable quantitative mapping for tensor, kurtosis, and NODDI models across external datasets from five independent centers. It remains effective where conventional fitting becomes unreliable, recovering meaningful maps from extremely sparse acquisitions while supporting downstream tractography and structural connectivity mapping. PIGMENT estimates demonstrated strong biological validity, preserving submillimeter cortical microarchitectural patterns and early-childhood white matter developmental trajectories from 10-fold accelerated scans. Furthermore, PIGMENT enables reliable quantitative tensor mapping on cost-efficient low-field systems and the extraction of tumor-related biomarkers using ultra-fast clinical protocols. Together, these results establish PIGMENT as a physics-informed foundation model that extends quantitative diffusion MRI into regimes traditionally too sparse, heterogeneous, or clinically constrained for reliable analysis.

eess.IV↗

arXiv2026

Rapid Approximation Prediction for Kriging

Exact Kriging and conditional simulation (CS) for uncertainty quantification are computationally infeasible for modern spatial analyses with large numbers of observations and dense prediction grids. We present a rapid approximation to the Kriging prediction step for stationary Gaussian processes for a regular prediction grid by approximating each off-grid covariance vector by a sparse linear combination of on-grid covariances within a local $L$-order neighborhood of $M = (2L)^2$ neighboring grid points. This reformulation reduces complexity from $O(N n^3)$ to $O(N \log N + nM + M^3)$ while preserving accuracy. A factorial study shows that approximation error decreases systematically with increased Mat\'{e}rn smoothness, neighbor order $L$, and grid resolution, aligning with bounds from kernel approximation theory. In a North American summer-rainfall application ($n=1368$), our method produces predictions visually indistinguishable from exact Kriging with point-wise errors on the order of $10^{-5}$ inches and achieves more than $150$ times speedups at a $350\times350$ grid, also outperforming Vecchia and LatticeKrig predictions. Embedded in a fast CS scheme, the approach reproduces Kriging standard errors and scales favorably with both $n$ and $N$. We recommend a practical workflow that uses a fast method for parameter estimation followed by our rapid predictor for fine-grid mapping and uncertainty quantification.

stat.ME↗

arXiv2026

CoilDrop-MRI: Self-supervised physics-guided MRI reconstruction with coil dropout

Self-supervised deep learning-based methods have shown great promise for accelerated magnetic resonance imaging (MRI) reconstruction, achieving high image quality without requiring fully sampled data for training. These methods typically partition the acquired data into two disjoint subsets to construct input-target pairs for optimizing the reconstruction network. However, existing approaches perform this partition exclusively within the spatial frequency (k-space) domain, leaving the coil dimension unexplored. To enforce full exploitation of signal correlation across receiver coils, we propose CoilDrop-MRI, which applies coil-wise dropout to the input and uses the dropped data as training targets in a self-supervised framework. This method is integrated into unrolled architectures in both image-domain (SENSE) and k-space (SPIRiT) formulations. We further demonstrate its versatility by extending CoilDrop-MRI to multi-shot, phase-corrected diffusion MRI (dMRI) reconstruction. CoilDrop-MRI is extensively validated on multi-site, multi-field-strength (0.3T, 0.55T, and 3T), and multi-modality (T1-weighted, T2-weighted, T2-FLAIR, and dMRI) datasets and consistently outperforms state-of-the-art self-supervised methods, achieving quality comparable to supervised reconstruction methods without requiring fully sampled reference training data. Moreover, CoilDrop-MRI exhibits strong data efficiency and robust generalization across imaging conditions, establishing it as a practical and versatile framework for self-supervised parallel MRI reconstruction.

cs.CV↗

arXiv2026

iGSP:Implicit Gradient Subspace Projection for Efficient Continual Learning of Vision-Language Models

Vision-Language Models require efficient adaptation to continually emerging downstream tasks. While Parameter-Efficient Fine-Tuning mitigates catastrophic forgetting, assigning isolated modules per task leads to parameter explosion. Conversely, recent similarity-driven sharing mechanisms falsely equate superficial visual similarity with underlying alignment consistency. This fundamental mismatch triggers severe negative transfer between visually similar but logically distinct tasks and fails to exploit alignment reuse across visually diverse ones. We argue thatalignment sharing is fundamentally a geometric problem of overlapping optimization trajectories within shared low-rank subspaces. Grounded in this insight, we propose iGSP, a novel framework that achieves efficient adaptation via implicit gradient subspace projection. Leveraging the early convergence of MoE routers to establish the subspace basis, iGSP bifurcates the adaptation process into two phases. First, the Subspace Identification phase introduces candidate experts via basis pre-expansion, applies a novel subspace-constrained regularization to implicitly project new task gradients onto the historical subspace, and precisely prunes redundant dimensions by treating routing probabilities as gradient flow indicators, ultimately to maximize knowledge reuse. Second, the Orthogonal Subspace Fine-Tuning phase fixes this structural basis and removes the regularization to rapidly fit the task-specific residual loss. Extensive experiments on the MTIL benchmark demonstrate that iGSP achieves state-of-the-art accuracy while significantly improving training efficiency, reducing the average trainable parameters by 42.7\% compared to current SOTA methods, and decreasing the final total parameters by 86.9\% relative to counterparts. The source code is available at https://github.com/GeoX-Lab/iGSP.

cs.CV↗

arXiv2026

RS-Claw: Progressive Active Tool Exploration via Hierarchical Skill Trees for Remote Sensing Agents

The rise of multi-modal large language models (MLLMs) is shifting remote sensing (RS) intelligence from "see" to "action", as OpenClaw-style frameworks enable agents to autonomously operate massive RS image-processing tools for complex tasks. Existing RS agents adopt a passive selection paradigm for tool invocation, relying on either full tool registration (Flat) or retrieval-augmented generation (RAG). However, in the massive and multi-source heterogeneous RS tool ecosystem, such passive mechanisms struggle to dynamically balance "context load" and "toolset completeness" throughout task reasoning, thus exhibiting inherent limitations: full tool registration triggers context space deficits during long-horizon tasks, whereas RAG retrieval may omit critical tools in essential steps. To overcome these bottlenecks, this paper redefines tool selection by arguing that the agent should act as an active explorer within the tool space. Based on this perspective, we propose RS-Claw, a novel RS agent architecture. By leveraging Skill encapsulation technology at the tool end, this architecture hierarchically structures tool descriptions, enabling the agent to execute on-demand sequential decision-making: initially selecting relevant skill branches by reading only tool summaries, then dynamically loading detailed descriptions, and ultimately achieving precise invocation. This active paradigm not only significantly liberates the agent's context space but also effectively ensures the accurate hit rate of critical tools during long-horizon reasoning. Systematic experiments on the Earth-Bench benchmark demonstrate that RS-Claw's active exploration mechanism effectively filters semantic noise and substantially frees up reasoning space, achieving an input token compression ratio of up to 86%, and comprehensively outperforming existing Flat and RAG baselines across complex reasoning evaluations.

cs.AI↗

arXiv2026

Towards Order Fairness: Mitigating LLMs Order Sensitivity through Dual Group Advantage Optimization

Large Language Models (LLMs) suffer from order bias, where their performance is affected by the arrangement order of input elements. This unfairness limits the model's applications in scenarios such as in-context learning and Retrieval-Augmented Generation (RAG). Recent studies attempt to obtain optimal or suboptimal arrangements based on statistical results or using dataset-based search, but these methods increase inference overhead while leaving the model's inherent order bias unresolved. Other studies mitigate order sensitivity through supervised fine-tuning using augmented training sets with multiple order variants, but often at the cost of accuracy, trapping the model in consistent yet incorrect hallucinations. In this paper, we propose \textbf{D}ual \textbf{G}roup \textbf{A}dvantage \textbf{O}ptimization (\textbf{DGAO}), which aims to improve model accuracy and order stability simultaneously. DGAO calculates and balances intra-group relative accuracy advantage and inter-group relative stability advantage, rewarding the policy model for generating order-stable and correct outputs while penalizing order-sensitive or incorrect responses. This marks the first time reinforcement learning has been used to mitigate LLMs' order sensitivity. We also propose two new metrics, Consistency Rate and Overconfidence Rate, to reveal the pseudo-stability of previous methods and guide more comprehensive evaluation. Extensive experiments demonstrate that DGAO achieves superior order fairness while improving performance on RAG, mathematical reasoning, and classification tasks. Our code is available at: https://github.com/Hyalinesky/DGAO.

cs.LG↗

arXiv2026

INFANiTE: Implicit Neural representation for high-resolution Fetal brain spatio-temporal Atlas learNing from clinical Thick-slicE MRI

Spatio-temporal fetal brain atlases are important for characterizing normative neurodevelopment and identifying congenital anomalies. However, existing atlas construction pipelines necessitate days for slice-to-volume reconstruction (SVR) to generate high-resolution 3D brain volumes and several additional days for iterative volume registration, thereby rendering atlas construction from large-scale cohorts prohibitively impractical. We address these limitations with INFANiTE, an Implicit Neural Representation (INR) framework for high-resolution Fetal brain spatio-temporal Atlas learNing from clinical Thick-slicE MRI scans, bypassing both the costly SVR and the iterative non-rigid registration steps entirely, thereby substantially accelerating atlas construction. Extensive experiments demonstrate that INFANiTE outperforms existing baselines in subject consistency, reference fidelity, intrinsic quality and biological plausibility, even under challenging sparse-data settings. Additionally, INFANiTE reduces the end-to-end processing time (i.e., from raw scans to the final atlas) from days to hours compared to the traditional 3D volume-based pipeline (e.g., SyGN), facilitating large-scale population-level fetal brain analysis. Code: https://github.com/hu2274898/INFANiTE

cs.CV↗

arXiv2026

Bidirectional Semantic Complementary Tool Retrieval for Remote Sensing Agents

Large language model (LLM)-based agents provide a novel paradigm for the automated processing of remote sensing(RS) data. Their success in complex RS tasks rely on extensive specialized tool libraries. However, tool documentation often exceeds the context window limits of LLMs, making precise tool retrieval essential for agentic workflows. Existing tool retrieval methods face "semantic asymmetry" bottleneck: natural language queries typically express macro-level intentions lacking tool-specific semantics, while tool documentation provides fine-grained technical descriptions lacking operational context for workflows. To bridge this semantic gap, this paper proposes a bidirectional semantic complementary tool retrieval method. First, on the query side, we introduce a planning-based query enhancement mechanism that leverages the reasoning capabilities of agents to decompose abstract intentions into logical subtasks, thereby actively supplementing the query with missing functional semantics. Second, on the tool side, addressing the strong coupling characteristics of RS tool chains, we construct a dynamic tool dependency graph with continual learning capabilities. By employing a neighborhood information aggregation mechanism, contextual information from precursor tools is explicitly injected into the current node representation, enriching tool descriptions with contextual semantics. Experimental results on the RS dataset GeoPlan-bench and the general-purpose dataset API- Bank demonstrate that the proposed method not only significantly improves tool retrieval accuracy for complex RS tasks but also exhibits robust extensibility for transfer to general-domain tasks. The source code and dataset are available at https://github.com/geox-lab/BSCTR.

cs.IR↗

arXiv2026

RS-HyRe-R1: A Hybrid Reward Mechanism to Overcome Perceptual Inertia for Remote Sensing Images Understanding

Reinforcement learning (RL) post-training substantially improves remote sensing vision-language models (RS-VLMs). However, when handling complex remote sensing imagery (RSI) requiring exhaustive visual scanning, models tend to rely on localized salient cues for rapid inference. We term this RL-induced bias "perceptual inertia". Driven by reward maximization, models favor quick outcome fitting, leading to two limitations: cognitively, overreliance on specific features impedes complete evidence construction; operationally, models struggle to flexibly shift visual focus across tasks. To address this bias and encourage comprehensive visual evidence mining, we propose RS-HyRe-R1, a hybrid reward framework for RSI understanding. It introduces: (1) a spatial reasoning activation reward that enforces structured visual reasoning; (2) a perception correctness reward that provides adaptive quality anchors across RS tasks, ensuring accurate geometric and semantic alignment; and (3) a visual-semantic path evolution reward that penalizes repetitive reasoning and promotes exploration of complementary cues to build richer evidence chains. Experiments show RS-HyRe-R1 effectively mitigates "perceptual inertia", encouraging deeper, more diverse reasoning. With only 3B parameters, it achieves state-of-the-art performance on REC, OVD, and VQA tasks, outperforming models up to 7B parameters. It also demonstrates strong zero-shot generalization, surpassing the second-best model by 3.16%, 3.97%, and 2.72% on VQA, OVD, and REC, respectively. Code and datasets are available at https://github.com/geox-lab/RS-HyRe-R1.

cs.CV↗

arXiv2026

High-fidelity 3D multi-slab diffusion MRI using Slab-shifting for Harmonized 3D Acquisition and Reconstruction with Profile Encoding Networks (SHARPEN)

Three-dimensional (3D) multi-slab imaging is a promising approach for high-resolution in vivo diffusion MRI (dMRI) due to its compatibility with short TR (1-2 s), providing optimal signal-to-noise ratio (SNR) efficiency. A major challenge, however, is slab boundary artifacts arising from non-ideal slab-selective RF excitation. Non-rectangular slab profiles reduce signal intensity at slab boundaries, while profile overlap across adjacent slabs introduces inter-slab crosstalk, where repeated excitation shortens the local TR and limits T1 recovery. To mitigate slab boundary artifacts without increasing scan time, we build on slab profile encoding and propose Slab-shifting for Harmonized 3D Acquisition and Reconstruction with Profile Encoding Networks (SHARPEN). For different diffusion directions, SHARPEN applies inter-volume field-of-view shifts along the slice direction to provide complementary slab profile encoding without prolonging acquisition. Slab profiles are estimated using a lightweight self-supervised neural network that exploits consistency across shifted acquisitions and known physical properties of slab profiles and diffusion images, and corrected images are reconstructed accordingly. SHARPEN was validated using simulated and prospectively acquired high-resolution in vivo data and demonstrates accurate slab profile estimation and robust boundary artifact correction, even in the presence of inter-volume motion. SHARPEN does not require high-quality reference training data and supports subject-specific training. Its efficient GPU-based implementation delivers faster and more accurate correction than NPEN, yielding slice-wise quantitative profiles that closely match those from reference 2D acquisitions. SHARPEN enables high-quality dMRI at 0.7 mm isotropic resolution on a 3T clinical scanner, highlighting its potential to advance submillimeter dMRI for neuroscience research.

physics.med-ph↗

arXiv2026

From Prompt to Graph: Comparing LLM-Based Information Extraction Strategies in Domain-Specific Ontology Development

Ontologies are essential for structuring domain knowledge, improving accessibility, sharing, and reuse. However, traditional ontology construction relies on manual annotation and conventional natural language processing (NLP) techniques, making the process labour-intensive and costly, especially in specialised fields like casting manufacturing. The rise of Large Language Models (LLMs) offers new possibilities for automating knowledge extraction. This study investigates three LLM-based approaches, including pre-trained LLM-driven method, in-context learning (ICL) method and fine-tuning method to extract terms and relations from domain-specific texts using limited data. We compare their performances and use the best-performing method to build a casting ontology that validated by domian expert.

cs.AI↗

arXiv2025

Parallel Diffusion Solver via Residual Dirichlet Policy Optimization

Diffusion models (DMs) have achieved state-of-the-art generative performance but suffer from high sampling latency due to their sequential denoising nature. Existing solver-based acceleration methods often face significant image quality degradation under a low-latency budget, primarily due to accumulated truncation errors arising from the inability to capture high-curvature trajectory segments. In this paper, we propose the Ensemble Parallel Direction solver (dubbed as EPD-Solver), a novel ODE solver that mitigates these errors by incorporating multiple parallel gradient evaluations in each step. Motivated by the geometric insight that sampling trajectories are largely confined to a low-dimensional manifold, EPD-Solver leverages the Mean Value Theorem for vector-valued functions to approximate the integral solution more accurately. Importantly, since the additional gradient computations are independent, they can be fully parallelized, preserving low-latency sampling nature. We introduce a two-stage optimization framework. Initially, EPD-Solver optimizes a small set of learnable parameters via a distillation-based approach. We further propose a parameter-efficient Reinforcement Learning (RL) fine-tuning scheme that reformulates the solver as a stochastic Dirichlet policy. Unlike traditional methods that fine-tune the massive backbone, our RL approach operates strictly within the low-dimensional solver space, effectively mitigating reward hacking while enhancing performance in complex text-to-image (T2I) generation tasks. In addition, our method is flexible and can serve as a plugin (EPD-Plugin) to improve existing ODE samplers.

cs.CV↗

arXiv2025

Automated Penetration Testing with LLM Agents and Classical Planning

While penetration testing plays a vital role in cybersecurity, achieving fully automated, hands-off-the-keyboard execution remains a significant research challenge. In this paper, we introduce the "Planner-Executor-Perceptor (PEP)" design paradigm and use it to systematically review existing work and identify the key challenges in this area. We also evaluate existing penetration testing systems, with a particular focus on the use of Large Language Model (LLM) agents for this task. The results show that the out-of-the-box Claude Code and Sonnet 4.5 exhibit superior penetration capabilities observed to date, substantially outperforming all prior systems. However, a detailed analysis of their testing processes reveals specific strengths and limitations; notably, LLM agents struggle with maintaining coherent long-horizon plans, performing complex reasoning, and effectively utilizing specialized tools. These limitations significantly constrain its overall capability, efficiency, and stability. To address these limitations, we propose CHECKMATE, a framework that integrates enhanced classical planning with LLM agents, providing an external, structured "brain" that mitigates the inherent weaknesses of LLM agents. Our evaluation shows that CHECKMATE outperforms the state-of-the-art system (Claude Code) in penetration capability, improving benchmark success rates by over 20%. In addition, it delivers substantially greater stability, cutting both time and monetary costs by more than 50%.

cs.CR↗

arXiv2025

Asking like Socrates: Socrates helps VLMs understand remote sensing images

Recent multimodal reasoning models, inspired by DeepSeek-R1, have significantly advanced vision-language systems. However, in remote sensing (RS) tasks, we observe widespread pseudo reasoning: models narrate the process of reasoning rather than genuinely reason toward the correct answer based on visual evidence. We attribute this to the Glance Effect, where a single, coarse perception of large-scale RS imagery results in incomplete understanding and reasoning based on linguistic self-consistency instead of visual evidence. To address this, we propose RS-EoT (Remote Sensing Evidence-of-Thought), a language-driven, iterative visual evidence-seeking paradigm. To instill this paradigm, we propose SocraticAgent, a self-play multi-agent system that synthesizes reasoning traces via alternating cycles of reasoning and visual inspection. To enhance and generalize these patterns, we propose a two-stage progressive RL strategy: first, RL on fine-grained Grounding tasks to enhance RS-EoT capabilities, followed by RL on RS VQA to generalize to broader understanding scenarios. Experiments show RS-EoT achieves state-of-the-art performance on multiple RS VQA and grounding benchmarks. Analyses reveal clear iterative cycles of reasoning and evidence seeking, confirming RS-EoT mitigates the Glance Effect and enables genuine evidence-grounded reasoning. Our code, data, and models are available at https://geox-lab.github.io/Asking_like_Socrates

cs.CV↗