arXiv subjects

Sheng Wang

Publications and source records attributed to Sheng Wang.

At least 91 records · Page 5Linked to original sources

arXiv2025

VisionCAD: An Integration-Free Radiology Copilot Framework

Widespread clinical deployment of computer-aided diagnosis (CAD) systems is hindered by the challenge of integrating with existing hospital IT infrastructure. Here, we introduce VisionCAD, a vision-based radiological assistance framework that circumvents this barrier by capturing medical images directly from displays using a camera system. The framework operates through an automated pipeline that detects, restores, and analyzes on-screen medical images, transforming camera-captured visual data into diagnostic-quality images suitable for automated analysis and report generation. We validated VisionCAD across diverse medical imaging datasets, demonstrating that our modular architecture can flexibly utilize state-of-the-art diagnostic models for specific tasks. The system achieves diagnostic performance comparable to conventional CAD systems operating on original digital images, with an F1-score degradation typically less than 2\% across classification tasks, while natural language generation metrics for automated reports remain within 1\% of those derived from original images. By requiring only a camera device and standard computing resources, VisionCAD offers an accessible approach for AI-assisted diagnosis, enabling the deployment of diagnostic capabilities in diverse clinical settings without modifications to existing infrastructure.

cs.CV↗

arXiv2025

A Unified Approach for Multi-Granularity Search over Spatial Datasets

There has been increased interest in data search as a means to find relevant datasets or data points in data lakes and repositories. Although approaches have been proposed to support spatial dataset search and data point search, they consider the two types of searches independently. To enable search operations ranging from the coarse-grained dataset level to the fine-grained data point level, we provide an integrated one that supports diverse query types and distance metrics. In this paper, we focus on designing a multi-granularity spatial data search system, called Spadas, that supports both dataset and data point search operations. To address the challenges of the high cost of indexing and susceptibility to outliers, we propose a unified index that can drastically improve query efficiency in various scenarios by organizing data reasonably and removing outliers in datasets. Moreover, to accelerate all data search operations, we propose a set of pruning mechanisms based on the unified index, including fast bound estimation, approximation technique with error bound, and pruning in batch techniques, to effectively filter out non-relevant datasets and points. Finally, we report the results of a detailed experimental evaluation using six spatial data repositories, achieving orders of magnitude faster than the state-of-the-art algorithms and demonstrating the effectiveness by case study. An online spatial data search system of Spadas is also implemented and made accessible to users.

cs.DB↗

arXiv2025

Existence and qualitative properties of ground state solutions for the Schrödinger-Bopp-Podolsky system

This paper concerns the existence and related properties of solutions to the Schrödinger-Bopp-Podolsky system, which reduces to a nonlinear and nonlocal partial differential equation describing a Schrödinger field coupled with its electromagnetic field in Bopp-Podolsky theory under purely electrostatic conditions. Firstly, by applying the mountain-pass lemma, we obtain the existence of nontrivial solutions. Then, through some estimates of the ground state energy, we prove the existence of ground state solutions. By exploring the relationship between solutions and paths associated with critical points, we further demonstrate that the obtained solutions are ground states of mountain-pass type. Additionally, the positivity, radial symmetry, rotational invariance, and exponential decay of the ground state solutions are considered. Finally, in the radial case, we explore the asymptotic behavior of the obtained solutions with respect to $a$.

math.AP↗

arXiv2025

Demeter: A Parametric Model of Crop Plant Morphology from the Real World

Learning 3D parametric shape models of objects has gained popularity in vision and graphics and has showed broad utility in 3D reconstruction, generation, understanding, and simulation. While powerful models exist for humans and animals, equally expressive approaches for modeling plants are lacking. In this work, we present Demeter, a data-driven parametric model that encodes key factors of a plant morphology, including topology, shape, articulation, and deformation into a compact learned representation. Unlike previous parametric models, Demeter handles varying shape topology across various species and models three sources of shape variation: articulation, subcomponent shape variation, and non-rigid deformation. To advance crop plant modeling, we collected a large-scale, ground-truthed dataset from a soybean farm as a testbed. Experiments show that Demeter effectively synthesizes shapes, reconstructs structures, and simulates biophysical processes. Code and data is available at https://tianhang-cheng.github.io/Demeter/.

cs.CV↗

arXiv2025

Cross-border offshore hydrogen trade and carbon mitigation for Europe's net zero transition

European countries are ambitious in both the net-zero transition and offshore energy resource development. The Irish and UK governments announced their commitments to offshore wind capacities - 37 and 125 GW, respectively, in 2050, more than two times higher than their projected power demands. While other continental countries, such as Germany, are calling for cleaner fuel resources. Exporting surplus offshore green hydrogen and bridging supply and demand could be pivotal in carbon emission mitigation for Europe. Yet, the potentials of these Island countries, are usually underestimated. This paper developed a bottom-up method to investigate the role of offshore hydrogen from Ireland and the UK in the decarbonisation of the entire Europe. We evaluate the future hydrogen/ammonia trading and the contributions of each country in carbon emission mitigation, considering their relative cost-competitiveness in offshore hydrogen production, domestic hourly power and gas system operation, and international shipping costs. Results indicate that the offshore green hydrogen could reduce 175.16 Mt/year of carbon dioxide emissions in Europe. The UK will be the largest hydrogen supplier from 2030 to 2040, while surpassed by Ireland in 2050, with 161 TWh of hydrogen exports to France and Spain. The offshore green hydrogen can contribute to 175.16 Mt of annual carbon dioxide emission reductions in total. This general flow of hydrogen from the West to the East not only facilitates Europe's net-zero progress, but also reshapes the energy supply structure and helps to ensure energy security across the European continent.

eess.SY↗

arXiv2025

ProReason: Multi-Modal Proactive Reasoning with Decoupled Eyesight and Wisdom

Large vision-language models (LVLMs) have witnessed significant progress on visual understanding tasks. However, they often prioritize language knowledge over image information on visual reasoning tasks, incurring performance degradation. To tackle this issue, we first identify the drawbacks of existing solutions (i.e., limited multi-modal reasoning capacities, and insufficient and irrelevant visual descriptions). We then decompose visual reasoning process into two stages: proactive visual perception (i.e., eyesight) and textual reasoning (i.e., wisdom), and introduce a novel visual reasoning framework named ProReason. This framework features decoupled vision-reasoning capabilities and multi-run proactive perception. Briefly, given a multi-modal question, ProReason iterates proactive information collection and reasoning until the answer can be concluded with necessary and sufficient visual descriptions. Notably, the disassociation of capabilities allows seamless integration of existing large language models (LLMs) to compensate for the reasoning deficits of LVLMs. Our extensive experiments demonstrate that ProReason outperforms existing multi-step reasoning frameworks on various benchmarks for both open-source and closed-source models, with the average performance gain reaching 13.2%. Besides, the integration of LLMs allows ProReason to produce high-quality visual reasoning data, which empowers ProReason-distilled models (i.e., ProReason-VL and ProReason-Q3) to achieve superior performance in downstream tasks. Our insights into existing solutions and the decoupled perspective for feasible integration of LLMs illuminate future research on visual reasoning techniques, especially LLM-assisted ones.

cs.CV↗

arXiv2025

Optimizing Storage Overhead of User Behavior Log for ML-embedded Mobile Apps

Machine learning (ML) models are increasingly integrated into modern mobile apps to enable personalized and intelligent services. These models typically rely on rich input features derived from historical user behaviors to capture user intents. However, as ML-driven services become more prevalent, recording necessary user behavior data imposes substantial storage cost on mobile apps, leading to lower system responsiveness and more app uninstalls. To address this storage bottleneck, we present AdaLog, a lightweight and adaptive system designed to improve the storage efficiency of user behavior log in ML-embedded mobile apps, without compromising model inference accuracy or latency. We identify two key inefficiencies in current industrial practices of user behavior log: (i) redundant logging of overlapping behavior data across different features and models, and (ii) sparse storage caused by storing behaviors with heterogeneous attribute descriptions in a single log file. To solve these issues, AdaLog first formulates the elimination of feature-level redundant data as a maximum weighted matching problem in hypergraphs, and proposes a hierarchical algorithm for efficient on-device deployment. Then, AdaLog employs a virtually hashed attribute design to distribute heterogeneous behaviors into a few log files with physically dense storage. Finally, to ensure scalability to dynamic user behavior patterns, AdaLog designs an incremental update mechanism to minimize the I/O operations needed for adapting outdated behavior log. We implement a prototype of AdaLog and deploy it into popular mobile apps in collaboration with our industry partner. Evaluations on real-world user data show that AdaLog reduces behavior log size by 19% to 44% with minimal system overhead (only 2 seconds latency and 15 MB memory usage), providing a more efficient data foundation for broader adoption of on-device ML.

cs.LG↗

arXiv2025

Pathology-CoT: Learning Visual Chain-of-Thought Agent from Expert Whole Slide Image Diagnosis Behavior

Diagnosing a whole-slide image is an interactive, multi-stage process of changing magnification and moving between fields. Although recent pathology foundation models demonstrated superior performances, practical agentic systems that decide what field to examine next, adjust magnification, and deliver explainable diagnoses are still lacking. Such limitation is largely bottlenecked by data: scalable, clinically aligned supervision of expert viewing behavior that is tacit and experience-based, not documented in textbooks or internet, and therefore absent from LLM training. Here we introduce a framework designed to address this challenge through three key breakthroughs. First, the AI Session Recorder seamlessly integrates with standard whole-slide image viewers to unobtrusively record routine navigation and convert the viewer logs into standardized behavioral commands and bounding boxes. Second, a lightweight human-in-the-loop review turns AI-drafted rationales for behavioral commands into the Pathology-CoT dataset, a form of paired "where to look" and "why it matters", enabling six-fold faster labeling compared to manual constructing such Chain-of-Thought dataset. Using this behavioral data, we build Pathology-o3, a two-stage agent that first proposes important ROIs and then performs behavior-guided reasoning. On the gastrointestinal lymph-node metastasis detection task, our method achieved 100 recall on the internal validation from Stanford Medicine and 97.6 recall on an independent external validation from Sweden, exceeding the state-of-the-art OpenAI o3 model and generalizing across backbones. To our knowledge, Pathology-CoT constitutes one of the first behavior-grounded agentic systems in pathology. Turning everyday viewer logs into scalable, expert-validated supervision, our framework makes agentic pathology practical and establishes a path to human-aligned, upgradeable clinical AI.

cs.CV↗

arXiv2025

QSpec: Speculative Decoding with Complementary Quantization Schemes

Quantization is widely adopted to accelerate inference and reduce memory consumption in large language models (LLMs). While activation-weight joint quantization enables efficient low-precision decoding, it suffers from substantial performance degradation on multi-step reasoning tasks. We propose QSpec, a novel quantization paradigm that decouples efficiency from quality by integrating two complementary schemes via speculative decoding: low-precision joint quantization for fast drafting and high-precision weight-only quantization for accurate verification. QSpec reuses both weights and KV cache across stages, enabling near-zero-cost switching without retraining or auxiliary models. Compared to high-precision baselines, QSpec achieves up to 1.64x speedup without quality degradation, and outperforms state-of-the-art speculative decoding methods by up to 1.55x in batched settings. Furthermore, QSpec supports plug-and-play deployment and generalizes well across model scales, quantization methods, and workloads. These properties make QSpec a practical and scalable solution for high-fidelity quantized LLM serving under memory-constrained scenarios. Our code is available at https://github.com/hku-netexplo-lab/QSpec.

cs.LG↗

arXiv2025

On Convergence Rates of General $N$-Player Stackelberg Games to their Mean Field Limits

In this article, we establish precise convergence rates of a general class of $N$-Player Stackelberg games to their mean field limits, which allows the response time delay of information, empirical distribution based interactions, and the control-dependent diffusion coefficients. All these features makes our problem nonstandard, barely been touched in the literature, and they complicate the analysis and therefore reduce the convergence rate. We first justify the same convergence rate for both the followers and the leader. Specifically, for the most general case, the convergence rate is shown to be $\mathcal{O}\left(N^{-\frac{2(q-2)}{n_1(3q-4)}}\right)$ when $n_1>4$ where $n_1$ is the dimension of the follower's state, and $q$ is the order of the integration of the initial; and this rate has yet been shown in the literature, to the best of our knowledge. Moreover, by classifying cases according to the state dimension $n_1$, the nature of the delay, and the assumptions of the coefficients, we provide several subcases where faster convergence rates can be obtained; for instance the $\mathcal{O}\left(N^{-\frac{2}{3n_1}}\right)$-convergence when the diffusion coefficients are independent of control variable. Our result extends the standard $o(1)$-convergence result for the mean field Stackelberg games in the literature, together with the $\mathcal{O}(N^{-\frac{1}{n_1+4}})$-convergence for the mean field games with major and minor players. We also discuss the special case where our coefficients are linear in distribution argument while nonlinear in state and control arguments, and we establish an $\mathcal{O}(1/\sqrt{N})$ convergence rate, which extends the linear quadratic cases in the literature.

math.OC↗

arXiv2025

A Besov-based integration-by-parts method for the incompressible Navier-Stokes equations

This note introduces a novel numerical analysis framework for the incompressible Navier-Stokes equations based on Besov spaces. The key contribution of this note is to establish the stability and convergence of a semi-implicit time-stepping scheme by deriving precise error estimates in the $B^0_{\infty,1}$ and $B^0_{\infty,2}$ spaces. Another contribution of our analysis is the detailed treatment of the $B^0_{\infty,2}$ case, where a crucial integration-by-parts technique is employed to adeptly handle the nonlinear advection term. This technique allows for a refined estimate that effectively transfers derivatives onto the test functions, mitigating the inherent analytical challenges posed by the low regularity of these spaces. Our results provide sharper, more localized error bounds than in classical Sobolev spaces, directly linking the scheme's convergence to the critical regularity of the continuous solution. This work underscores the advantage of Besov spaces for the numerical analysis of nonlinear fluid PDEs.

math.NA↗

arXiv2025

A co-evolving agentic AI system for medical imaging analysis

Agentic AI is rapidly advancing in healthcare and biomedical research. However, in medical image analysis, their performance and adoption remain limited due to the lack of a robust ecosystem, insufficient toolsets, and the absence of real-time interactive expert feedback. Here we present "TissueLab", a co-evolving agentic AI system that allows researchers to ask direct questions, automatically plan and generate explainable workflows, and conduct real-time analyses where experts can visualize intermediate results and refine them. TissueLab integrates tool factories across pathology, radiology, and spatial omics domains. By standardizing inputs, outputs, and capabilities of diverse tools, the system determines when and how to invoke them to address research and clinical questions. Across diverse tasks with clinically meaningful quantifications that inform staging, prognosis, and treatment planning, TissueLab achieves state-of-the-art performance compared with end-to-end vision-language models (VLMs) and other agentic AI systems such as GPT-5. Moreover, TissueLab continuously learns from clinicians, evolving toward improved classifiers and more effective decision strategies. With active learning, it delivers accurate results in unseen disease contexts within minutes, without requiring massive datasets or prolonged retraining. Released as a sustainable open-source ecosystem, TissueLab aims to accelerate computational research and translational adoption in medical imaging while establishing a foundation for the next generation of medical AI.

cs.CV↗

arXiv2025

Online Learning Based Efficient Resource Allocation for LoRaWAN Network

The deployment of large-scale LoRaWAN networks requires jointly optimizing conflicting metrics like Packet Delivery Ratio (PDR) and Energy Efficiency (EE) by dynamically allocating transmission parameters, including Carrier Frequency, Spreading Factor, and Transmission Power. Existing methods often oversimplify this challenge, focusing on a single metric or lacking the adaptability needed for dynamic channel environments, leading to suboptimal performance. To address this, we propose two online learning-based resource allocation frameworks that intelligently navigate the PDR-EE trade-off. Our foundational proposal, D-LoRa, is a fully distributed framework that models the problem as a Combinatorial Multi-Armed Bandit. By decomposing the joint parameter selection and employing specialized, disaggregated reward functions, D-LoRa dramatically reduces learning complexity and enables nodes to autonomously adapt to network dynamics. To further enhance performance in LoRaWAN networks, we introduce CD-LoRa, a hybrid framework that integrates a lightweight, centralized initialization phase to perform a one-time, quasi-optimal channel assignment and action space pruning, thereby accelerating subsequent distributed learning. Extensive simulations and real-world field experiments demonstrate the superiority of our frameworks, showing that D-LoRa excels in non-stationary environments while CD-LoRa achieves the fastest convergence in stationary conditions. In physical deployments, our methods outperform state-of-the-art baselines, improving PDR by up to 10.8% and EE by 26.1%, confirming their practical effectiveness for scalable and efficient LoRaWAN networks.

cs.NI↗

arXiv2025

NeurStore: Efficient In-database Deep Learning Model Management System

With the prevalence of in-database AI-powered analytics, there is an increasing demand for database systems to efficiently manage the ever-expanding number and size of deep learning models. However, existing database systems typically store entire models as monolithic files or apply compression techniques that overlook the structural characteristics of deep learning models, resulting in suboptimal model storage overhead. This paper presents NeurStore, a novel in-database model management system that enables efficient storage and utilization of deep learning models. First, NeurStore employs a tensor-based model storage engine to enable fine-grained model storage within databases. In particular, we enhance the hierarchical navigable small world (HNSW) graph to index tensors, and only store additional deltas for tensors within a predefined similarity threshold to ensure tensor-level deduplication. Second, we propose a delta quantization algorithm that effectively compresses delta tensors, thus achieving a superior compression ratio with controllable model accuracy loss. Finally, we devise a compression-aware model loading mechanism, which improves model utilization performance by enabling direct computation on compressed tensors. Experimental evaluations demonstrate that NeurStore achieves superior compression ratios and competitive model loading throughput compared to state-of-the-art approaches.

cs.DB↗

arXiv2025

zERExtractor:An Automated Platform for Enzyme-Catalyzed Reaction Data Extraction from Scientific Literature

The rapid expansion of enzyme kinetics literature has outpaced the curation capabilities of major biochemical databases, creating a substantial barrier to AI-driven modeling and knowledge discovery. We present zERExtractor, an automated and extensible platform for comprehensive extraction of enzyme-catalyzed reaction and activity data from scientific literature. zERExtractor features a unified, modular architecture that supports plug-and-play integration of state-of-the-art models, including large language models (LLMs), as interchangeable components, enabling continuous system evolution alongside advances in AI. Our pipeline combines domain-adapted deep learning, advanced OCR, semantic entity recognition, and prompt-driven LLM modules, together with human expert corrections, to extract kinetic parameters (e.g., kcat, Km), enzyme sequences, substrate SMILES, experimental conditions, and molecular diagrams from heterogeneous document formats. Through active learning strategies integrating AI-assisted annotation, expert validation, and iterative refinement, the system adapts rapidly to new data sources. We also release a large benchmark dataset comprising over 1,000 annotated tables and 5,000 biological fields from 270 P450-related enzymology publications. Benchmarking demonstrates that zERExtractor consistently outperforms existing baselines in table recognition (Acc 89.9%), molecular image interpretation (up to 99.1%), and relation extraction (accuracy 94.2%). zERExtractor bridges the longstanding data gap in enzyme kinetics with a flexible, plugin-ready framework and high-fidelity extraction, laying the groundwork for future AI-powered enzyme modeling and biochemical knowledge discovery.

q-bio.BM↗

arXiv2025

Towards a deeper GCN: Alleviate over-smoothing with iterative training and fine-tuning

Graph Convolutional Networks (GCNs) suffer from severe performance degradation in deep architectures due to over-smoothing. While existing studies primarily attribute the over-smoothing to repeated applications of graph Laplacian operators, our empirical analysis reveals a critical yet overlooked factor: trainable linear transformations in GCNs significantly exacerbate feature collapse, even at moderate depths (e.g., 8 layers). In contrast, Simplified Graph Convolution (SGC), which removes these transformations, maintains stable feature diversity up to 32 layers, highlighting linear transformations' dual role in facilitating expressive power and inducing over-smoothing. However, completely removing linear transformations weakens the model's expressive capacity. To address this trade-off, we propose Layer-wise Gradual Training (LGT), a novel training strategy that progressively builds deep GCNs while preserving their expressiveness. LGT integrates three complementary components: (1) layer-wise training to stabilize optimization from shallow to deep layers, (2) low-rank adaptation to fine-tune shallow layers and accelerate training, and (3) identity initialization to ensure smooth integration of new layers and accelerate convergence. Extensive experiments on benchmark datasets demonstrate that LGT achieves state-of-the-art performance on vanilla GCN, significantly improving accuracy even in 32-layer settings. Moreover, as a training method, LGT can be seamlessly combined with existing methods such as PairNorm and ContraNorm, further enhancing their performance in deeper networks. LGT offers a general, architecture-agnostic training framework for scalable deep GCNs. The code is available at [https://github.com/jfklasdfj/LGT_GCN].

cs.LG↗

arXiv2025

Approximate Vector Set Search Inspired by Fly Olfactory Neural System

Vector set search, an underexplored similarity search paradigm, aims to find vector sets similar to a query set. This search paradigm leverages the inherent structural alignment between sets and real-world entities to model more fine-grained and consistent relationships for diverse applications. This task, however, faces more severe efficiency challenges than traditional single-vector search due to the combinatorial explosion of pairings in set-to-set comparisons. In this work, we aim to address the efficiency challenges posed by the combinatorial explosion in vector set search, as well as the curse of dimensionality inherited from single-vector search. To tackle these challenges, we present an efficient algorithm for vector set search, BioVSS (Bio-inspired Vector Set Search). BioVSS simulates the fly olfactory circuit to quantize vectors into sparse binary codes and then designs an index based on the set membership property of the Bloom filter. The quantization and indexing strategy enables BioVSS to efficiently perform vector set search by pruning the search space. Experimental results demonstrate over 50 times speedup compared to linear scanning on million-scale datasets while maintaining a high recall rate of up to 98.9%, making it an efficient solution for vector set search.

cs.DB↗

arXiv2025

Imaginary potential and thermal width in the spinning black hole background from holography

In this study, we investigate the imaginary potential of heavy quarkonium in the spinning black hole background. Then we estimate the thermal width, which is determined by the imaginary part of the finite temperature potential. In the ultra-local description, the boosted fluid represents a globally rotating fluid. Using a holographic approach, we systematically analyze how boost parameter influences these quantities. Our results reveal that increasing boost parameter causes the imaginary potential to emerge at smaller interquark distances, suggesting that boost parameter accelerates quarkonium melting. Furthermore, we find that boost parameter enhances the thermal width, indicating greater instability of the bound state at higher boost parameter. Notably, we observe that the effect of boost parameter on quarkonium dissociation is more pronounced when the axis of the quark-antiquark pair is transverse to the direction of boost parameter.

hep-ph↗