arXiv subjects

Xiao Zhou

Publications and source records attributed to Xiao Zhou.

At least 73 records · Page 4Linked to original sources

arXiv2025

Lost in Time? A Meta-Learning Framework for Time-Shift-Tolerant Physiological Signal Transformation

Translating non-invasive signals such as photoplethysmography (PPG) and ballistocardiography (BCG) into clinically meaningful signals like arterial blood pressure (ABP) is vital for continuous, low-cost healthcare monitoring. However, temporal misalignment in multimodal signal transformation impairs transformation accuracy, especially in capturing critical features like ABP peaks. Conventional synchronization methods often rely on strong similarity assumptions or manual tuning, while existing Learning with Noisy Labels (LNL) approaches are ineffective under time-shifted supervision, either discarding excessive data or failing to correct label shifts. To address this challenge, we propose ShiftSyncNet, a meta-learning-based bi-level optimization framework that automatically mitigates performance degradation due to time misalignment. It comprises a transformation network (TransNet) and a time-shift correction network (SyncNet), where SyncNet learns time offsets between training pairs and applies Fourier phase shifts to align supervision signals. Experiments on one real-world industrial dataset and two public datasets show that ShiftSyncNet outperforms strong baselines by 9.4%, 6.0%, and 12.8%, respectively. The results highlight its effectiveness in correcting time shifts, improving label quality, and enhancing transformation accuracy across diverse misalignment scenarios, pointing toward a unified direction for addressing temporal inconsistencies in multimodal physiological transformation.

cs.LG↗

arXiv2025

Advancing Ocean State Estimation with efficient and scalable AI

Accurate and efficient global ocean state estimation remains a grand challenge for Earth system science, hindered by the dual bottlenecks of computational scalability and degraded data fidelity in traditional data assimilation (DA) and deep learning (DL) approaches. Here we present an AI-driven Data Assimilation Framework for Ocean (ADAF-Ocean) that directly assimilates multi-source and multi-scale observations, ranging from sparse in-situ measurements to 4 km satellite swaths, without any interpolation or data thinning. Inspired by Neural Processes, ADAF-Ocean learns a continuous mapping from heterogeneous inputs to ocean states, preserving native data fidelity. Through AI-driven super-resolution, it reconstructs 0.25$^\circ$ mesoscale dynamics from coarse 1$^\circ$ fields, which ensures both efficiency and scalability, with just 3.7\% more parameters than the 1$^\circ$ configuration. When coupled with a DL forecasting system, ADAF-Ocean extends global forecast skill by up to 20 days compared to baselines without assimilation. This framework establishes a computationally viable and scientifically rigorous pathway toward real-time, high-resolution Earth system monitoring.

cs.LG↗

arXiv2025

Reachability of Independent Sets and Vertex Covers Under Extended Reconfiguration Rules

In reconfiguration problems, we are given two feasible solutions to a graph problem and asked whether one can be transformed into the other via a sequence of feasible intermediate solutions under a given reconfiguration rule. While earlier work focused on modifying a single element at a time, recent studies have started examining how different rules impact computational complexity. Motivated by recent progress, we study Independent Set Reconfiguration (ISR) and Vertex Cover Reconfiguration (VCR) under the $k$-Token Jumping ($k$-TJ) and $k$-Token Sliding ($k$-TS) models. In $k$-TJ, up to $k$ vertices may be replaced, while $k$-TS additionally requires a perfect matching between removed and added vertices. It is known that the complexity of ISR crucially depends on $k$, ranging from PSPACE-complete and NP-complete to polynomial-time solvable. In this paper, we further explore the gradient of computational complexity of the problems. We first show that ISR under $k$-TJ with $k = |I| - μ$ remains NP-hard when $μ$ is any fixed positive integer and the input graph is restricted to graphs of maximum degree 3 or planar graphs of maximum degree 4, where $|I|$ is the size of feasible solutions. In addition, we prove that the problem belongs to NP not only for $μ=O(1)$ but also for $μ= O(\log |I|)$. In contrast, we show that VCR under $k$-TJ is in XP when parameterized by $μ= |S| - k$, where $|S|$ is the size of feasible solutions. Furthermore, we establish the PSPACE-completeness of ISR and VCR under both $k$-TJ and $k$-TS on several graph classes, for fixed $k$ as well as superconstant $k$ relative to the size of feasible solutions.

cs.CC↗

arXiv2025

From Retrieval to Generation: Unifying External and Parametric Knowledge for Medical Question Answering

Medical question answering (QA) requires extensive access to domain-specific knowledge. A promising direction is to enhance large language models (LLMs) with external knowledge retrieved from medical corpora or parametric knowledge stored in model parameters. Existing approaches typically fall into two categories: Retrieval-Augmented Generation (RAG), which grounds model reasoning on externally retrieved evidence, and Generation-Augmented Generation (GAG), which depends solely on the models internal knowledge to generate contextual documents. However, RAG often suffers from noisy or incomplete retrieval, while GAG is vulnerable to hallucinated or inaccurate information due to unconstrained generation. Both issues can mislead reasoning and undermine answer reliability. To address these challenges, we propose MedRGAG, a unified retrieval-generation augmented framework that seamlessly integrates external and parametric knowledge for medical QA. MedRGAG comprises two key modules: Knowledge-Guided Context Completion (KGCC), which directs the generator to produce background documents that complement the missing knowledge revealed by retrieval; and Knowledge-Aware Document Selection (KADS), which adaptively selects an optimal combination of retrieved and generated documents to form concise yet comprehensive evidence for answer generation. Extensive experiments on five medical QA benchmarks demonstrate that MedRGAG achieves a 12.5% improvement over MedRAG and a 4.5% gain over MedGENIE, highlighting the effectiveness of unifying retrieval and generation for knowledge-intensive reasoning. Our code and data are publicly available at https://anonymous.4open.science/r/MedRGAG

cs.CL↗

arXiv2025

Beyond Single Models: Mitigating Multimodal Hallucinations via Adaptive Token Ensemble Decoding

Large Vision-Language Models (LVLMs) have recently achieved impressive results in multimodal tasks such as image captioning and visual question answering. However, they remain prone to object hallucination -- generating descriptions of nonexistent or misidentified objects. Prior work has partially mitigated this via auxiliary training objectives or external modules, but challenges remain in terms of scalability, adaptability, and model independence. To address these limitations, we propose Adaptive Token Ensemble Decoding (ATED), a training-free, token-level ensemble framework that mitigates hallucination by aggregating predictions from multiple LVLMs during inference. ATED dynamically computes uncertainty-based weights for each model, reflecting their reliability at each decoding step. It also integrates diverse decoding paths to improve contextual grounding and semantic consistency. Experiments on standard hallucination detection benchmarks demonstrate that ATED significantly outperforms state-of-the-art methods, reducing hallucination without compromising fluency or relevance. Our findings highlight the benefits of adaptive ensembling and point to a promising direction for improving LVLM robustness in high-stakes applications. The code is available at https://github.com/jinlin2021/ATED.

cs.CV↗

arXiv2025

A Physics-Guided AI Cascaded Corrector Model Significantly Extends Madden-Julian Oscillation Prediction Skill

The Madden-Julian Oscillation (MJO) is an important driver of global weather and climate extremes, but its prediction in operational dynamical models remains challenging, with skillful forecasts typically limited to 3-4 weeks. Here, we introduce a novel deep learning framework, the Physics-guided Cascaded Corrector for MJO (PCC-MJO), which acts as a universal post-processor to correct MJO forecasts from dynamical models. This two-stage model first employs a physics-informed 3D U-Net to correct spatial-temporal field errors, then refines the MJO's RMM index using an LSTM optimized for forecast skill. When applied to three different operational forecasts from CMA, ECMWF and NCEP, our unified framework consistently extends the skillful forecast range (bivariate correlation > 0.5) by 2-8 days. Crucially, the model effectively mitigates the "Maritime Continent barrier", enabling more realistic eastward propagation and amplitude. Explainable AI analysis quantitatively confirms that the model's decision-making is spatially congruent with observed MJO dynamics (correlation > 0.93), demonstrating that it learns physically meaningful features rather than statistical fittings. Our work provides a promising physically consistent, computationally efficient, and highly generalizable pathway to break through longstanding barriers in subseasonal forecasting.

cs.LG↗

arXiv2025

The bimodal distribution of donor stars in X-ray binaries

The classification of X-ray binaries into high- and low-mass types has historically lacked a unified, data-driven quantitative criterion, and large-scale statistical studies of the donor star population have been limited. In this work, we address this gap by compiling data for 3,964 XRBs and deriving a plentiful set of physical parameters (mass, radius, age, and evolutionary stage) for a sub-sample of 288 donor stars using Gaia DR3 spectral data and stellar evolution models. We find a statistically bimodal distribution in the donor star parameters, which is characterized by a valley at approximately 3 $M_{\odot}$ or 11,000 K. We uncover the physical mechanism behind this bimodality: a previously unreported ``parallel tracks'' phenomenon observed in the relationship between the donor's evolutionary stage and its fundamental parameters, such as luminosity and radius. These two tracks represent distinct main-sequence populations, and the valley between them corresponds to the sparsely populated pre- and post-main-sequence evolutionary phases.

astro-ph.SR↗

arXiv2025

WildSpeech-Bench: Benchmarking End-to-End SpeechLLMs in the Wild

Recent multi-modal Large Language Models (LLMs) such as GPT-4o have demonstrated strong capabilities of direct speech interaction. However, the lack of specialized and comprehensive benchmarks for end-to-end speech LLM evaluation hinders optimizing the user experience of Audio LLMs in real-world applications. Existing evaluation methods often adapt text-based benchmarks, overlooking speech's unique characteristics and challenges, including prosody, homophones, stuttering, and differing user expectations. Here, we introduce the first comprehensive benchmark designed to systematically evaluate end-to-end speechLLMs in practical speech conversations. We systematically curate real-world chat data relevant to spoken scenarios, introduce diversity in speaker attributes and acoustic conditions, and augment the dataset with speech-specific phenomena. We further design a query-aware evaluation method to use customized evaluation checklists and prompts to enhance the accuracy of automatic evaluation. We conduct comprehensive testing and detailed analysis of various mainstream speech models, revealing significant differences in model performance across different speech scenarios. The use of query-aware evaluation further enables a finer-grained assessment under various speech-specific scenarios. Our benchmark can provide valuable insights for speech model development and evaluation.

cs.CL↗

arXiv2025

SocialTraj: Two-Stage Socially-Aware Trajectory Prediction for Autonomous Driving via Conditional Diffusion Model

Accurate trajectory prediction of surrounding vehicles (SVs) is crucial for autonomous driving systems to avoid misguided decisions and potential accidents. However, achieving reliable predictions in highly dynamic and complex traffic scenarios remains a significant challenge. One of the key impediments lies in the limited effectiveness of current approaches to capture the multi-modal behaviors of drivers, which leads to predicted trajectories that deviate from actual future motions. To address this issue, we propose SocialTraj, a novel trajectory prediction framework integrating social psychology principles through social value orientation (SVO). By utilizing Bayesian inverse reinforcement learning (IRL) to estimate the SVO of SVs, we obtain the critical social context to infer the future interaction trend. To ensure modal consistency in predicted behaviors, the estimated SVOs of SVs are embedded into a conditional denoising diffusion model that aligns generated trajectories with historical driving styles. Additionally, the planned future trajectory of the ego vehicle (EV) is explicitly incorporated to enhance interaction modeling. Extensive experiments on NGSIM and HighD datasets demonstrate that SocialTraj is capable of adapting to highly dynamic and interactive scenarios while generating socially compliant and behaviorally consistent trajectory predictions, outperforming existing baselines. Ablation studies demonstrate that dynamic SVO estimation and explicit ego-planning components notably improve prediction accuracy and substantially reduce inference time.

cs.RO↗

arXiv2025

POINTS-Reader: Distillation-Free Adaptation of Vision-Language Models for Document Conversion

High-quality labeled data is essential for training accurate document conversion models, particularly in domains with complex formats such as tables, formulas, and multi-column text. However, manual annotation is both costly and time-consuming, while automatic labeling using existing models often lacks accuracy in handling such challenging scenarios. Consequently, training student models by distilling outputs from teacher models can significantly limit their performance in real-world applications. In this paper, we propose a fully automated, distillation-free framework comprising two stages for constructing high-quality document extraction datasets and models capable of handling diverse document formats and layouts. In the first stage, we introduce a method for generating large-scale, diverse synthetic data, which enables a model to extract key elements in a unified format with strong initial performance. In the second stage, we present a self-improvement approach that further adapts the model, initially trained on synthetic data, to real-world documents. Specifically, we first use the fine-tuned model to annotate real documents, then apply a suite of filtering strategies to verify annotation quality, and finally retrain the model on the verified dataset. By iteratively repeating this process, we progressively enhance both the model's conversion capabilities and the quality of the generated data. We train a public POINTS-1.5 model to obtain POINTS-Reader, which surpasses many existing public and proprietary models of comparable or larger size. Our model is available at https://github.com/Tencent/POINTS-Reader.

cs.CV↗

arXiv2025

Boosting Pathology Foundation Models via Few-shot Prompt-tuning for Rare Cancer Subtyping

Rare cancers comprise 20-25% of all malignancies but face major diagnostic challenges due to limited expert availability-especially in pediatric oncology, where they represent over 70% of cases. While pathology vision-language (VL) foundation models show promising zero-shot capabilities for common cancer subtyping, their clinical performance for rare cancers remains limited. Existing multi-instance learning (MIL) methods rely only on visual features, overlooking cross-modal knowledge and compromising interpretability critical for rare cancer diagnosis. To address this limitation, we propose PathPT, a novel framework that fully exploits the potential of vision-language pathology foundation models through spatially-aware visual aggregation and task-specific prompt tuning. Unlike conventional MIL, PathPT converts WSI-level supervision into fine-grained tile-level guidance by leveraging the zero-shot capabilities of VL models, thereby preserving localization on cancerous regions and enabling cross-modal reasoning through prompts aligned with histopathological semantics. We benchmark PathPT on eight rare cancer datasets(four adult and four pediatric) spanning 56 subtypes and 2,910 WSIs, as well as three common cancer datasets, evaluating four state-of-the-art VL models and four MIL frameworks under three few-shot settings. Results show that PathPT consistently delivers superior performance, achieving substantial gains in subtyping accuracy and cancerous region grounding ability. This work advances AI-assisted diagnosis for rare cancers, offering a scalable solution for improving subtyping accuracy in settings with limited access to specialized expertise.

cs.CV↗

arXiv2025

The Liouville-type equation and an Onofri-type inequality on closed 4-manifolds

In this paper, we study the Liouville-type equation \[Δ^2 u-λ_1κΔu+λ_2κ^2(1-\mathrm e^{4u})=0\] on a closed Riemannian manifold $(M^4,g)$ with $\operatorname{Ric}\geqslant 3κg$ and $κ>0$. Using the method of invariant tensors, we derive a differential identity to classify solutions within certain ranges of the parameters $λ_1,λ_2$. A key step in our proof is a second-order derivative estimate, which is established via the continuity method. As an application of the classification results, we derive an Onofri-type inequality on the 4-sphere and prove its rigidity.

math.AP↗

arXiv2025

AutoSchemaKG: Autonomous Knowledge Graph Construction through Dynamic Schema Induction from Web-Scale Corpora

We present AutoSchemaKG, a framework for fully autonomous knowledge graph construction that eliminates the need for predefined schemas. Our system leverages large language models to simultaneously extract knowledge triples and induce comprehensive schemas directly from text, modeling both entities and events while employing conceptualization to organize instances into semantic categories. Processing over 50 million documents, we construct ATLAS (Automated Triple Linking And Schema induction), a family of knowledge graphs with 900+ million nodes and 5.9 billion edges. This approach outperforms state-of-the-art baselines on multi-hop QA tasks and enhances LLM factuality. Notably, our schema induction achieves 92\% semantic alignment with human-crafted schemas with zero manual intervention, demonstrating that billion-scale knowledge graphs with dynamically induced schemas can effectively complement parametric knowledge in large language models.

cs.CL↗

arXiv2025

An extremely low mass ratio contact binary NW Aps with a potential compact companion star

The photometric solutions of NW Aps reveal that it is a low mass ratio ($q = 0.086$) contact binary system. Investigation of orbital period shows that its orbital period is increasing continuously at a rate of $dP/dt=+1.117(\pm0.005)\times{10^{-6}}day\cdot year^{-1}$, which may be caused by mass transfer from the less massive component to the more massive one at a rate of $\frac{dM_{2}}{dt}=-3.36(\pm0.02)\times{10^{-8}}M_\odot/year$. A cyclic variation of $P_3 = 22.9(\pm0.1)$ is also found in the O - C curve. There may be a potential compact object orbiting around NW Aps, with its minimum mass to be $M_3 = 0.436(\pm0.007)M_\odot$. However, the magnetic activity of the primary star may also account for the cyclic change. NW Aps is a stellar merger candidate with the longest orbital period among all stellar merge candidates with mass ratio $q < 0.1$. It is still in a stable state since the ratio of orbital angular momentum ($J_{orb}$) to spin angular momentum ($J_{spin}$) is $\frac{J_{orb}}{J_{spin}}$ = 3.257. Both of its primary and secondary star are oversized to main sequence stars, and the surface gravity of the primary and secondary stars are significant lower than main sequence stars. The P - log g relationship is fitted with parabola for low mass ratio contact binary systems. More targets laid in the gap are needed to confirm the P - log g relationship and reveal the final evolutionary state of low mass ratio contact binary system.

astro-ph.SR↗

arXiv2025

Agentic Satellite-Augmented Low-Altitude Economy and Terrestrial Networks: A Survey on Generative Approaches

The development of satellite-augmented low-altitude economy and terrestrial networks (SLAETNs) demands intelligent and autonomous systems that can operate reliably across heterogeneous, dynamic, and mission-critical environments. To address these challenges, this survey focuses on enabling agentic artificial intelligence (AI), that is, artificial agents capable of perceiving, reasoning, and acting, through generative AI (GAI) and large language models (LLMs). We begin by introducing the architecture and characteristics of SLAETNs, and analyzing the challenges that arise in integrating satellite, aerial, and terrestrial components. Then, we present a model-driven foundation by systematically reviewing five major categories of generative models: variational autoencoders (VAEs), generative adversarial networks (GANs), generative diffusion models (GDMs), transformer-based models (TBMs), and LLMs. Moreover, we provide a comparative analysis to highlight their generative mechanisms, capabilities, and deployment trade-offs within SLAETNs. Building on this foundation, we examine how these models empower agentic functions across three domains: communication enhancement, security and privacy protection, and intelligent satellite tasks. Finally, we outline key future directions for building scalable, adaptive, and trustworthy generative agents in SLAETNs. This survey aims to provide a unified understanding and actionable reference for advancing agentic AI in next-generation integrated networks.

cs.NI↗

arXiv2025

Large-Vocabulary Segmentation for Medical Images with Text Prompts

This paper aims to build a model that can Segment Anything in 3D medical images, driven by medical terminologies as Text prompts, termed as SAT. Our main contributions are three-fold: (i) We construct the first multimodal knowledge tree on human anatomy, including 6502 anatomical terminologies; Then, we build the largest and most comprehensive segmentation dataset for training, collecting over 22K 3D scans from 72 datasets, across 497 classes, with careful standardization on both image and label space; (ii) We propose to inject medical knowledge into a text encoder via contrastive learning and formulate a large-vocabulary segmentation model that can be prompted by medical terminologies in text form; (iii) We train SAT-Nano (110M parameters) and SAT-Pro (447M parameters). SAT-Pro achieves comparable performance to 72 nnU-Nets -- the strongest specialist models trained on each dataset (over 2.2B parameters combined) -- over 497 categories. Compared with the interactive approach MedSAM, SAT-Pro consistently outperforms across all 7 human body regions with +7.1% average Dice Similarity Coefficient (DSC) improvement, while showing enhanced scalability and robustness. On 2 external (cross-center) datasets, SAT-Pro achieves higher performance than all baselines (+3.7% average DSC), demonstrating superior generalization ability.

eess.IV↗

arXiv2025

Nonlinear Spectral Fusion Super-Resolution Fluorescence Microscopy based on Progressively Saturated Upconversion Nanoparticles

Single-beam scanning microscopy (SBSM) is one of the most robust strategies for commercial optical systems. Although structured illumination combined with Fourier-domain spatial spectrum fusion can enhance SBSM resolution beyond the diffraction limit, a sophisticated detection system is still required to optimize both effective resolution and signal-to-noise ratio.Here, we report that the diverse nonlinear responses of upconversion nanoparticles can unlock a new mode of Computational Progressively Emission Saturated Nanoscopy (CPSN), which employs a single doughnut-shaped excitation beam assisted by deep learning to simplify conventional microscopy. By modulating the excitation power, the smooth transition of the point spread function (PSF) from doughnut-shaped to Gaussian can be achieved, allowing for accessing different spatial frequency components of the sample. Then, in order to enhance time resolution, the doughnut-shaped beam at low power and the saturated Gaussian-like image were predicted by the doughnut-shaped beam at low saturation threshold based on the power dependence curve. Furthermore, a deep recursive residual network (DRRN) is employed to fusion these progressively complementary spatial frequency information into a final super-resolved image that encompasses the full frequency wwinformation. This approach can achieve high-quality super-resolution imaging with a spatial resolution of 33 nm, corresponding to 1/29th of the excitation wavelength, 55 dB of SNR ratio contracted to 7 dB in Gaussian imaging and applicable to any wavelength. The unique combination of nonlinear saturation and deep learning computational reconstruction could open a new avenue for simplifying the optical system and enhancing imaging quality in single-beam super-resolution nanoscopy.

physics.optics↗

arXiv2025

RadIR: A Scalable Framework for Multi-Grained Medical Image Retrieval via Radiology Report Mining

Developing advanced medical imaging retrieval systems is challenging due to the varying definitions of `similar images' across different medical contexts. This challenge is compounded by the lack of large-scale, high-quality medical imaging retrieval datasets and benchmarks. In this paper, we propose a novel methodology that leverages dense radiology reports to define image-wise similarity ordering at multiple granularities in a scalable and fully automatic manner. Using this approach, we construct two comprehensive medical imaging retrieval datasets: MIMIC-IR for Chest X-rays and CTRATE-IR for CT scans, providing detailed image-image ranking annotations conditioned on diverse anatomical structures. Furthermore, we develop two retrieval systems, RadIR-CXR and model-ChestCT, which demonstrate superior performance in traditional image-image and image-report retrieval tasks. These systems also enable flexible, effective image retrieval conditioned on specific anatomical structures described in text, achieving state-of-the-art results on 77 out of 78 metrics.

cs.CV↗