Searcharxiv⌕ Search

arXiv subjects

Xinyu Wang

Publications and source records attributed to Xinyu Wang.

At least 127 records · Page 7Linked to original sources
arXiv2026

Edge-Efficient Two-Stream Multimodal Architecture for Non-Intrusive Bathroom Fall Detection

Falls in wet bathroom environments are a major safety risk for seniors living alone. Recent work has shown that mmWave-only, vibration-only, and existing multimodal schemes, such as vibration-triggered radar activation, early feature concatenation, and decision-level score fusion, can support privacy-preserving, non-intrusive fall detection. However, these designs still treat motion and impact as loosely coupled streams, depending on coarse temporal alignment and amplitude thresholds, and do not explicitly encode the causal link between radar-observed collapse and floor impact or address timing drift, object drop confounders, and latency and energy constraints on low-power edge devices. To this end, we propose a two-stream architecture that encodes radar signals with a Motion--Mamba branch for long-range motion patterns and processes floor vibration with an Impact--Griffin branch that emphasizes impact transients and cross-axis coupling. Cross-conditioned fusion uses low-rank bilinear interaction and a Switch--MoE head to align motion and impact tokens and suppress object-drop confounders. The model keeps inference cost suitable for real-time execution on a Raspberry Pi 4B gateway. We construct a bathroom fall detection benchmark dataset with frame-level annotations, comprising more than 3~h of synchronized mmWave radar and triaxial vibration recordings across eight scenarios under running water, together with subject-independent training, validation, and test splits. On the test split, our model attains 96.1% accuracy, 94.8% precision, 88.0% recall, a 91.1% macro F1 score, and an AUC of 0.968. Compared with the strongest baseline, it improves accuracy by 2.0 percentage points and fall recall by 1.3 percentage points, while reducing latency from 35.9 ms to 15.8 ms and lowering energy per 2.56 s window from 14200 mJ to 10750 mJ on the Raspberry Pi 4B gateway.

cs.CV↗
arXiv2026

Co-Evolution of Policy and Internal Reward for Language Agents

Large language model (LLM) agents learn by interacting with environments, but long-horizon training remains fundamentally bottlenecked by sparse and delayed rewards. Existing methods typically address this challenge through post-hoc credit assignment or external reward models, which provide limited guidance at inference time and often separate reward improvement from policy improvement. We propose Self-Guide, a self-generated internal reward for language agents that supports both inference-time guidance and training-time supervision. Specifically, the agent uses Self-Guide as a short self-guidance signal to steer the next action during inference, and converts the same signal into step-level internal reward for denser policy optimization during training. This creates a co-evolving loop: better policy produces better guidance, and better guidance further improves policy as internal reward. Across three agent benchmarks, inference-time self-guidance already yields clear gains, while jointly evolving policy and internal reward with GRPO brings further improvements (8\%) over baselines trained solely with environment reward. Overall, our results suggest that language agents can improve not only by collecting more experience, but also by learning to generate and refine their own internal reward during acting and learning.

cs.LG↗
arXiv2026

TABQAWORLD: Optimizing Multimodal Reasoning for Multi-Turn Table Question Answering

Multimodal reasoning has emerged as a powerful framework for enhancing reasoning capabilities of reasoning models. While multi-turn table reasoning methods have improved reasoning accuracy through tool use and reward modeling, they rely on fixed text serialization for table state readouts. This introduces representation errors in table encoding that significantly accumulate over multiple turns. Such accumulation is alleviated by tabular grounding methods in the expense of inference compute and cost, rendering real world deployment impractical. To address this, we introduce TABQAWORLD, a table reasoning framework that jointly optimizes tabular action through representation and estimation. For representation, TABQAWORLD employs an action-conditioned multimodal selection policy, which dynamically switches between visual and textual representations to maximize table state readout reliability. For estimation, TABQAWORLD optimizes stepwise reasoning trajectory through table metadata including dimension, data types and key values, safely planning trajectory and compressing low-complexity actions to reduce conversation turns and latency. Designed as a training-free framework, empirical evaluations show that TABQAWORLD achieves state-of-the-art performance with 4.87% accuracy improvements over baselines, with 5.42% accuracy gain and 33.35% inference latency reduction over static settings, establishing a new standard for reliable and efficient table reasoning.

cs.AI↗
arXiv2026

Ultrasound-CLIP: Semantic-Aware Contrastive Pre-training for Ultrasound Image-Text Understanding

Ultrasound imaging is widely used in clinical diagnostics due to its real-time capability and radiation-free nature. However, existing vision-language pre-training models, such as CLIP, are primarily designed for other modalities, and are difficult to directly apply to ultrasound data, which exhibit heterogeneous anatomical structures and diverse diagnostic attributes. To bridge this gap, we construct US-365K, a large-scale ultrasound image-text dataset containing 365k paired samples across 52 anatomical categories. We establish Ultrasonographic Diagnostic Taxonomy (UDT) containing two hierarchical knowledge frameworks. Ultrasonographic Hierarchical Anatomical Taxonomy standardizes anatomical organization, and Ultrasonographic Diagnostic Attribute Framework formalizes nine diagnostic dimensions, including body system, organ, diagnosis, shape, margins, echogenicity, internal characteristics, posterior acoustic phenomena, and vascularity. Building upon these foundations, we propose Ultrasound-CLIP, a semantic-aware contrastive learning framework that introduces semantic soft labels and semantic loss to refine sample discrimination. Moreover, we construct a heterogeneous graph modality derived from UDAF's textual representations, enabling structured reasoning over lesion-attribute relations. Extensive experiments with patient-level data splitting demonstrate that our approach achieves state-of-the-art performance on classification and retrieval benchmarks, while also delivering strong generalization to zero-shot, linear probing, and fine-tuning tasks.

cs.CV↗
arXiv2026

MiroEval: Benchmarking Multimodal Deep Research Agents in Process and Outcome

Recent progress in deep research systems has been impressive, but evaluation still lags behind real user needs. Existing benchmarks predominantly assess final reports using fixed rubrics, failing to evaluate the underlying research process. Most also offer limited multimodal coverage, rely on synthetic tasks that do not reflect real-world query complexity, and cannot be refreshed as knowledge evolves. To address these gaps, we introduce MiroEval, a benchmark and evaluation framework for deep research systems. The benchmark comprises 100 tasks (70 text-only, 30 multimodal), all grounded in real user needs and constructed via a dual-path pipeline that supports periodic updates, enabling a live and evolving setting. The proposed evaluation suite assesses deep research systems along three complementary dimensions: adaptive synthesis quality evaluation with task-specific rubrics, agentic factuality verification via active retrieval and reasoning over both web sources and multimodal attachments, and process-centric evaluation audits how the system searches, reasons, and refines throughout its investigation. Evaluation across 13 systems yields three principal findings: the three evaluation dimensions capture complementary aspects of system capability, with each revealing distinct strengths and weaknesses across systems; process quality serves as a reliable predictor of overall outcome while revealing weaknesses invisible to output-level metrics; and multimodal tasks pose substantially greater challenges, with most systems declining by 3 to 10 points. The MiroThinker series achieves the most balanced performance, with MiroThinker-H1 ranking the highest overall in both settings. Human verification and robustness results confirm the reliability of the benchmark and evaluation framework. MiroEval provides a holistic diagnostic tool for the next generation of deep research agents.

cs.AI↗
arXiv2026

AgentSwing: Adaptive Parallel Context Management Routing for Long-Horizon Web Agents

As large language models (LLMs) evolve into autonomous agents for long-horizon information-seeking, managing finite context capacity has become a critical bottleneck. Existing context management methods typically commit to a single fixed strategy throughout the entire trajectory. Such static designs may work well in some states, but they cannot adapt as the usefulness and reliability of the accumulated context evolve during long-horizon search. To formalize this challenge, we introduce a probabilistic framework that characterizes long-horizon success through two complementary dimensions: search efficiency and terminal precision. Building on this perspective, we propose AgentSwing, a state-aware adaptive parallel context management routing framework. At each trigger point, AgentSwing expands multiple context-managed branches in parallel and uses lookahead routing to select the most promising continuation. Experiments across diverse benchmarks and agent backbones show that AgentSwing consistently outperforms strong static context management methods, often matching or exceeding their performance with up to $3\times$ fewer interaction turns while also improving the ultimate performance ceiling of long-horizon web agents. Beyond the empirical gains, the proposed probabilistic framework provides a principled lens for analyzing and designing future context management strategies for long-horizon agents.

cs.CL↗
arXiv2026

Group Editing: Edit Multiple Images in One Go

In this paper, we tackle the problem of performing consistent and unified modifications across a set of related images. This task is particularly challenging because these images may vary significantly in pose, viewpoint, and spatial layout. Achieving coherent edits requires establishing reliable correspondences across the images, so that modifications can be applied accurately to semantically aligned regions. To address this, we propose GroupEditing, a novel framework that builds both explicit and implicit relationships among images within a group. On the explicit side, we extract geometric correspondences using VGGT, which provides spatial alignment based on visual features. On the implicit side, we reformulate the image group as a pseudo-video and leverage the temporal coherence priors learned by pre-trained video models to capture latent relationships. To effectively fuse these two types of correspondences, we inject the explicit geometric cues from VGGT into the video model through a novel fusion mechanism. To support large-scale training, we construct GroupEditData, a new dataset containing high-quality masks and detailed captions for numerous image groups. Furthermore, to ensure identity preservation during editing, we introduce an alignment-enhanced RoPE module, which improves the model's ability to maintain consistent appearance across multiple images. Finally, we present GroupEditBench, a dedicated benchmark designed to evaluate the effectiveness of group-level image editing. Extensive experiments demonstrate that GroupEditing significantly outperforms existing methods in terms of visual quality, cross-view consistency, and semantic alignment.

cs.CV↗
arXiv2026

BAWSeg: A UAV Multispectral Benchmark for Barley Weed Segmentation

Accurate weed mapping in cereal fields requires pixel-level segmentation from UAV imagery that remains reliable across fields, seasons, and illumination. Existing multispectral pipelines often depend on thresholded vegetation indices, which are brittle under radiometric drift and mixed crop--weed pixels, or on single-stream CNN and Transformer backbones that ingest stacked bands and indices, where radiance cues and normalized index cues interfere and reduce sensitivity to small weed clusters embedded in crop canopy. We propose VISA, a two-stream segmentation network that decouples these cues and fuses them at native resolution. The radiance stream learns from calibrated five-band reflectance using local residual convolutions, channel recalibration, spatial gating, and skip-connected decoding, which preserve fine textures, row boundaries, and small weed structures that are often weakened after ratio-based index compression. The index stream operates on vegetation-index maps with windowed self-attention to model local structure efficiently, state-space layers to propagate field-scale context without quadratic attention cost, and Slot Attention to form stable region descriptors that improve discrimination of sparse weeds under canopy mixing. To support supervised training and deployment-oriented evaluation, we introduce BAWSeg, a four-year UAV multispectral dataset collected over commercial barley paddocks in Western Australia, providing radiometrically calibrated blue, green, red, red edge, and near-infrared orthomosaics, derived vegetation indices, and dense crop, weed, and other labels with leakage-free block splits. On BAWSeg, VISA achieves 75.6% mIoU and 63.5% weed IoU with 22.8 M parameters, outperforming a multispectral SegFormer-B1 baseline by 1.2 mIoU and 1.9 weed IoU. Under cross-plot and cross-year protocols, VISA maintains 71.2% and 69.2% mIoU, respectively.

cs.CV↗
arXiv2026

Pressure-induced lattice instabilities and phonon softening in the orthorhombically distorted ferrimagnet Ni4Nb2O9

The ambient- and high-pressure behavior of the ferrimagnet Ni4Nb2O9 (orthorhombically distorted honeycomb structure), is investigated using NMR, Raman spectroscopy, and synchrotron XRD. Ambient-pressure NMR measurements reveal, despite its orthorhombic symmetry, the local environment of Ni4Nb2O9 closely resembles that of its trigonal analogue Mn4Nb2O9. In contrast, substantially different paramagnetic shifts observed in the two compounds reflect their distinct average crystal symmetries, governing orbital overlap and magnetic exchange pathways. Under external pressure, Ni4Nb2O9 exhibits pronounced sensitivity to lattice distortions and phonon instabilities. Three isostructural transitions are identified near 2, 6, and 10 GPa, manifested by mode splitting, frequency shifts, line broadenings, intensity anomalies, and slope changes in the evolution of lattice parameters. At higher pressure, around 13 GPa, signatures of an incipient long-range structural transition from orthorhombic Pbcn to monoclinic P2/c symmetry emerge, signaling the onset of a symmetry-lowering transformation. The anomalous softening of the 192 cm^-1 Raman mode, accompanied by multiple linewidth and spectral-weight anomalies, serve as a key fingerprint of these structural instabilities, linking local symmetry breaking at low pressures to the long-range transition into the P2/c phase. Notably, pronounced linewidth anomalies, strongly anisotropic pressure coefficients, together with a marked enhancement of the intensity of the low-frequency branch over the 2-13 GPa range, point toward a pressure-induced regime influenced by coupled spin, orbital, and lattice degrees of freedom. The close correspondence of transition pressures in Ni4Nb2O9 and those reported for Mn4Nb2O9 highlights a common mechanism rooted in their similar local structural environments, as revealed by NMR.

cond-mat.mtrl-sci↗
arXiv2026

Repurposing Synthetic Data for Fine-grained Search Agent Supervision

LLM-based search agents are increasingly trained on entity-centric synthetic data to solve complex, knowledge-intensive tasks. However, prevailing training methods like Group Relative Policy Optimization (GRPO) discard this rich entity information, relying instead on sparse, outcome-based rewards. This critical limitation renders them unable to distinguish informative "near-miss" samples-those with substantially correct reasoning but a flawed final answer-from complete failures, thus discarding valuable learning signals. We address this by leveraging the very entities discarded during training. Our empirical analysis reveals a strong positive correlation between the number of ground-truth entities identified during an agent's reasoning process and final answer accuracy. Building on this insight, we introduce Entity-aware Group Relative Policy Optimization (E-GRPO), a novel framework that formulates a dense entity-aware reward function. E-GRPO assigns partial rewards to incorrect samples proportional to their entity match rate, enabling the model to effectively learn from these "near-misses". Experiments on diverse question-answering (QA) and deep research benchmarks show that E-GRPO consistently and significantly outperforms the GRPO baseline. Furthermore, our analysis reveals that E-GRPO not only achieves superior accuracy but also induces more efficient reasoning policies that require fewer tool calls, demonstrating a more effective and sample-efficient approach to aligning search agents.

cs.CL↗
arXiv2026

Beyond Message Passing: A Symbolic Alternative for Expressive and Interpretable Graph Learning

Graph Neural Networks (GNNs) have become essential in high-stakes domains such as drug discovery, yet their black-box nature remains a significant barrier to trustworthiness. While self-explainable GNNs attempt to bridge this gap, they often rely on standard message-passing backbones that inherit fundamental limitations, including the 1-Weisfeiler-Lehman (1-WL) expressivity barrier and a lack of fine-grained interpretability. To address these challenges, we propose SymGraph, a symbolic framework designed to transcend these constraints. By replacing continuous message passing with discrete structural hashing and topological role-based aggregation, our architecture theoretically surpasses the 1-WL barrier, achieving superior expressiveness without the overhead of differentiable optimization. Extensive empirical evaluations demonstrate that SymGraph achieves state-of-the-art performance, outperforming existing self-explainable GNNs. Notably, SymGraph delivers 10x to 100x speedups in training time using only CPU execution. Furthermore, SymGraph generates rules with superior semantic granularity compared to existing rule-based methods, offering great potential for scientific discovery and explainable AI.

cs.LG↗
arXiv2026

Finite-time flocking of an infinite set of Cucker-Smale particles with sublinear velocity couplings

We study finite-time flocking for an infinite set of Cucker-Smale particles with sublinear velocity coupling under fixed and switching sender networks. For this, we use a component-wise diameter framework and exploit sub-linear dissipation mechanisms, and derive sufficient conditions for finite-time flocking equipped with explicit alignment-time estimate. For a fixed sender network, we establish component-wise finite-time flocking results under both integrable and non-integrable communication weights. When communication weight function is non-integrable, finite-time flocking is guaranteed for any bounded initial configuration. We further extend the flocking analysis to switching sender networks and show that finite-time flocking persists under mild assumptions on the cumulative influence of time-varying sender weights. The proposed framework is also applicable to both finite and infinite systems, and it yields alignment-time estimates that do not depend on the number of agents.

math.DS↗
arXiv2026

Geodesics and Shadows in the Kerr-Bertotti-Robinson Black Hole Spacetime

In this work, we investigate geodesics and black hole shadows in the Kerr-Bertotti-Robinson spacetime. We show that the equations of motion for null geodesics are separable and admit analytical treatment, whereas timelike geodesics are generally non-separable. Approximate analytical expressions for the photon sphere and the innermost stable circular orbit are derived via perturbative expansions in the magnetic field strength. We further explore the black hole shadow using both numerical and analytical methods, examining the effects of the magnetic field, the observer's inclination angle and radial position. Deviations from the standard Kerr shadow are quantified, and a physical interpretation is provided by introducing asymptotic regimes defined relative to the magnetic field strength.

gr-qc↗
arXiv2026

LARV: Data-Free Layer-wise Adaptive Rescaling Veneer for Model Merging

Model merging aims to combine multiple fine-tuned models into a single multi-task model without access to training data. Existing task-vector merging methods such as TIES, TSV-M, and Iso-C/CTS differ in their aggregation rules but treat all layers nearly uniformly. This assumption overlooks the strong layer-wise heterogeneity in large vision transformers, where shallow layers are sensitive to interference while deeper layers encode stable task-specific features. We introduce LARV, a training-free, data-free, merger-agnostic Layer-wise Adaptive Rescaling Veneer that plugs into any task-vector merger and assigns a per-layer scale to each task vector before aggregation, and show it consistently boosts diverse merging rules. LARV adaptively suppresses shallow-layer interference and amplifies deeper-layer alignment using a simple deterministic schedule, requiring no retraining or modification to existing mergers. To our knowledge, this is the first work to perform layer-aware scaling for task-vector merging. LARV computes simple data-free layer proxies and turns them into scales through a lightweight rule; we study several instantiations within one framework (e.g., tiered two/three-level scaling with fixed values, or continuous mappings) and show that tiered choices offer the best robustness, while continuous mappings remain an ablation. LARV is orthogonal to the base merger and adds negligible cost. On FusionBench with Vision Transformers, LARV consistently improves all task-vector baselines across 8/14/20-task settings; for example, Iso-C + LARV reaches 85.9% on ViT-B/32, 89.2% on ViT-B/16, and 92.6% on ViT-L/14. Layerwise analysis and corruption tests further indicate that LARV suppresses shallow-layer interference while modestly amplifying deeper, task-stable features, turning model merging into a robust, layer-aware procedure rather than a uniform one.

cs.CV↗
arXiv2026

OJBKQ: Objective-Joint Babai-Klein Quantization

Post-training quantization (PTQ) is widely used to compress large language models without retraining. However, many existing weight-only methods rely on heuristic objectives and greedy rounding, thus leading to noticeable degradation under low-bit quantization. In this work, we introduce OJBKQ (Objective-Joint Babai-Klein Quantization with K-Best Sampling), a layer-wise PTQ method that formulates weight quantization as a joint optimization problem over activations and weights. This formulation results in a multiple-right-hand-side box-constrained integer least squares (BILS) problem in each layer, which is NP-hard. For each column of the weight matrix, we apply an extended Babai nearest-plane algorithm and an extended version of Klein's randomized Babai algorithm to find the minimum-residual Babai-Klein point, a sub-optimal solution to the BILS problem. Experimental results on large language models show that OJBKQ achieves lower perplexity at 3-4 bits compared to existing PTQ approaches, while maintaining comparable computational cost.

cs.LG↗
arXiv2026

MonkeyOCR: Document Parsing with a Structure-Recognition-Relation Triplet Paradigm

We introduce MonkeyOCR, a document parsing model that advances the state of the art by leveraging a Structure-Recognition-Relation (SRR) triplet paradigm. This design simplifies what would otherwise be a complex multi-tool pipeline and avoids the inefficiencies of processing full pages with giant end-to-end models. In SRR, document parsing is abstracted into three fundamental questions - ``Where is it?'' (structure), ``What is it?'' (recognition), and ``How is it organized?'' (relation) - corresponding to structure detection, content recognition, and relation prediction. To support this paradigm, we present MonkeyDoc, a comprehensive dataset with 4.5 million bilingual instances spanning over ten document types, which addresses the limitations of existing datasets that often focus on a single task, language, or document type. Leveraging the SRR paradigm and MonkeyDoc, we trained a 3B-parameter document foundation model. We further identify parameter redundancy in this model and propose contiguous parameter degradation (CPD), enabling the construction of models from 0.6B to 1.2B parameters that run faster with acceptable performance drop. MonkeyOCR achieves state-of-the-art performance, surpassing previous open-source and closed-source methods, including Gemini 2.5-Pro. Additionally, the model can be efficiently deployed for inference on a single RTX 3090 GPU. Code and models will be released at https://github.com/Yuliang-Liu/MonkeyOCR.

cs.CV↗
arXiv2026

LiDAR, GNSS and IMU Sensor Fine Alignment through Dynamic Time Warping to Construct 3D City Maps

LiDAR-based 3D mapping suffers from cumulative drift causing global misalignment, particularly in GNSS-constrained environments. To address this, we propose a unified framework that fuses LiDAR, GNSS, and IMU data for high-resolution city-scale mapping. The method performs velocity-based temporal alignment using Dynamic Time Warping and refines GNSS and IMU signals via extended Kalman filtering. Local maps are built using Normal Distributions Transform-based registration and pose graph optimization with loop closure detection, while global consistency is enforced using GNSS-constrained anchors followed by fine registration of overlapping segments. We also introduce a large-scale multimodal dataset captured in Perth, Western Australia to facilitate future research in this direction. Our dataset comprises 144,000 frames acquired with a 128-channel Ouster LiDAR, synchronized RTK-GNSS trajectories, and MEMS-IMU measurements across 21 urban loops. To assess geometric consistency, we evaluated our method using alignment metrics based on road centerlines and intersections to capture both global and local accuracy. The proposed framework reduces the average global alignment error from 3.32m to 1.24m, achieving a 61.4% improvement, and significantly decreases the intersection centroid offset from 13.22m to 2.01m, corresponding to an 84.8% enhancement. The constructed high-fidelity map and raw dataset are publicly available through https://ieee-dataport.org/documents/perth-cbd-high-resolution-lidar-map-gnss-and-imu-calibration, and its visualization can be viewed at https://www.youtube.com/watch?v=-ZUgs1KyMks. The source code is available at https://github.com/HaitianWang/LiDAR-GNSS-and-IMU-Sensor-Fine-Alignment-through-Dynamic-Time-Warping-to-Construct-3D-City-Maps. This dataset and method together establish a new benchmark for evaluating 3D city mapping in GNSS-constrained environments.

cs.RO↗
arXiv2026

Paint by Odor: An Exploration of Odor Visualization through Large Language Model and Generative AI

Odor visualization translates odor information and perception into visual outcomes and arouses the corresponding olfactory synesthesia, surpassing the spatial limitation that odors can only be perceived where they are present. Traditional odor visualization has typically relied on unidimensional mappings, such as odor-to-color associations, and has required extensive manual design efforts. However, the advent of generative AI (Gen AI) and large language models (LLMs) presents a new opportunity for automatic odor visualization. Nonetheless, gaps remain in bridging olfactory perception with generative tools to produce odor images. To address these gaps, this paper introduces Paint by Odor, a pipeline that leverages Gen AI and LLMs to transform olfactory perceptions into rich, aesthetically engaging visual representations. Two experiments were conducted, where 30 participants smelled real-world odors and provided descriptive data and 28 participants evaluated 560 generated odor images through seven systematically designed prompts. Our findings explored the capability of LLMs in producing olfactory perception by comparing it with human responses and revealed the underlying mechanisms and effects of language-based descriptions and several abstraction styles on odor visualization. Our work further discussed the possibility of automatic odor visualization without human participation. These explorations and results have bridged the research gap in odor visualization using LLMs and Gen AI, offering valuable design insights and various possibilities for future applications.

cs.HC↗