arXiv subjects

Chao Sun

Publications and source records attributed to Chao Sun.

At least 19 recordsLinked to original sources

arXiv2026

UniFLM: United Segmentation and Measurement on Fetal Limb Ultrasonic Image

Prenatal ultrasound examination is crucial for assessing fetal limb development and detecting congenital anomalies. However, existing artificial intelligence models often overlook fetal lethal skeletal dysplasias due to the lack of high-quality annotated data and a unified framework for multiple long bones. Moreover, generic segmentation models struggle with the inherent noise and semantic gaps in ultrasound images. To address these challenges, we construct the Fetal Limb Bones (FLB) dataset, comprising high-quality annotations for the humerus, femur, tibia-fibula, and radius-ulna. Furthermore, we propose UniFLM, a unified framework for automatic cross-plane segmentation and measurement. UniFLM incorporates a Semantic-Aware Skip Connection module to bridge the semantic gap between encoder and decoder features, and a Positive Sampling strategy to adaptively filter noise and extract essential semantic information. Finally, a Point Regression Mapping module is introduced to learn clinician annotation patterns for precise bone length measurement. Extensive experiments conducted on the FLB dataset demonstrate that the proposed UniFLM achieves superior accuracy and enhanced generalization capabilities in fetal long bone assessment compared to current state-of-the-art models.

cs.CV↗

arXiv2026

Multimodal Adaptive Control for Safe Robotic Craniotomy Under Partial Observability

Autonomous robotic craniotomy requires continuous regulation of tool-tissue interactions to mitigate mechanical overload and thermal damage while maintaining surgical efficiency. However, this process is inherently partially observable due to unknown, time-varying tissue properties and the inability to directly measure cutting temperatures under physical occlusion. To address these challenges, we propose RL-MACRO, a cybernetic closed-loop intelligence framework that couples multimodal perception, adaptive decision-making, and robotic execution. This framework empowers the surgical robot to autonomously perceive inaccessible states from partial sensory feedback and dynamically optimize its behaviors under uncertain environment. A CNN-LSTM observer first fuses force and sound feedback to reconstruct the hidden temperature state (R^2=0.939, MAE = 1.717 deg C). This reconstructed temperature, alongside multi-sensor features, forms the belief state for an offline Implicit Q-Learning (IQL) policy. A novel dual-head Actor dynamically coordinates the feed rate, spindle speed, and cutting depth to optimize efficiency within strict safety bounds. These decisions are seamlessly translated into spatial motions via online trajectory re-planning and velocity servoing. Experiments on bovine ribs and six ex vivo goat skulls validate the system's robust perception, adaptive recovery from force/temperature excursions, and smooth execution on irregular surfaces, establishing a data-driven cybernetic paradigm for safe and efficient autonomous bone cutting.

cs.RO↗

arXiv2026

Defer to Plan: Adaptive Multi-Agent Fusion for End-to-End V2X Driving

Vehicle-to-everything-aided autonomous driving (V2X-AD) significantly enhances driving performance through information sharing. However, existing collaborative perception methods only optimize module-level perception capabilities and fail to effectively serve the ultimate planning and control tasks. We propose an end-to-end collaborative driving system that directly optimizes planning task performance. The system employs MotionNetwork to fuse historical temporal information, utilizes attention mechanisms to efficiently compress spatial features into compact tokens, and adaptively fuses multi-agent features through an autoregressive decoder. Additionally, we introduce Mixture-of-Experts (MoE) architecture to enhance the model's representation capacity for heterogeneous features. Experiments demonstrate that our method achieves a driving score of 79.72, surpassing the state-of-the-art CoDriving baseline (77.15) by 3.33% in closed-loop evaluation while maintaining communication efficiency.

cs.RO↗

arXiv2026

UniSpine-GS: An Efficient Physics-Aware Gaussian Framework for Cross-Modality Multi-view Spine Image Synthesis

The diagnosis of spinal diseases is often assisted by 3D imaging techniques in clinical practice. However, precise 3D spinal assessment is limited by the high costs of 3D imaging hardware and the challenges posed by the physical differences between imaging modalities, which hinder the generalizability of models. To address these issues, we propose UniSpine-GS, an efficient, physics-aware Gaussian framework designed for novel-view projection rendering in multi-view spine imaging via a 3D-aware representation. Instead of performing explicit 3D reconstruction, our approach learns a geometry-aware Gaussian representation that ensures anatomical consistency across different views. We introduce SPWM, a structure-guided loss reweighting strategy to improve boundary fidelity and local details. We evaluate our method on the CTSpine3D dataset and a newly constructed 3D fetal ultrasound dataset, FeSpine3D. Our results demonstrate that UniSpine-GS significantly outperforms existing methods across all metrics, offering a practical and cost-effective solution for unified multi-view medical imaging. Our code is publicly available at https://github.com/orangeisland66/UniSpine-GS.

cs.CV↗

arXiv2026

SonoCLIP: Mask-Guided Region-Aware Vision-Language Pretraining for Fetal Ultrasound Analysis

Vision-language foundation models have shown strong potential in medical image analysis. Although foundation models for ultrasound imaging have recently emerged, the domain remains particularly challenging due to severe speckle noise, acquisition variability, and subtle anatomical boundaries, leading to high inter-observer variability. Existing CLIP-based models rely primarily on global image-text alignment, limiting their sensitivity to clinically decisive local structures. We propose SonoCLIP, the first million-scale region-controllable fetal ultrasound vision-language foundation model that integrates segmentation masks as mask-channel visual prompts within the vision encoder, enabling joint global-local contrastive representation learning. To support scalable region-text alignment, we introduce a sigmoid-based pairwise contrastive loss that improves stability under large-scale supervision. We further curate a 1.44M-image multimodal fetal ultrasound dataset spanning 24 standard planes for large-scale pretraining. Extensive cross-center evaluations demonstrate that SonoCLIP achieves superior zero-shot transfer performance under both global and mask-guided inference, establishing a controllable and clinically oriented foundation model for fetal ultrasound analysis. Our code and data are available at https://github.com/Harrison-one/SonoCLIP.

cs.CV↗

arXiv2026

Median-of-Means for Nash Equilibrium Seeking in Heavy-Tailed Games

This paper studies the Nash equilibrium seeking problem for stochastic games under heavy-tailed noise. The gradient noise is considered to have a finite $\delta$-th moment ($1<\delta\le 2$), which generalizes the Gaussian noise and covers cases with infinite variance. In this work, we employ the classic method Median-of-Means (MoM) in robust estimation. MoM works by dividing samples into blocks, taking the average of each block, and then taking the median of these block averages, achieving a breakdown point of up to $1/2$. This makes the final estimate reliable even when some samples are very noisy or wrong, and thus is effective to handle the heavy-tailed noise. The method also naturally defends against malicious gradient attacks. Compared with gradient clipping, which is the most popular method to deal with the heavy-tailed noise, MoM requires no preset clipping threshold and is insensitive to the tail behavior of the noise. Under standard assumptions, we prove the almost sure convergence of the algorithm and derive its almost sure convergence rate. To address the systematic bias caused by asymmetric noise, we further design an online bias correction strategy. Simulation results show the effectiveness and efficiency of the proposed algorithms.

math.OC↗

arXiv2026

Random Reshuffling-Based Distributed Nash Equilibrium Seeking

This paper studies random reshuffling (RR)-based distributed Nash equilibrium seeking for noncooperative games. The game is motivated as a sample-average approximation of an underlying expected-value stochastic game, while the algorithmic focus is placed on the resulting finite-sum equilibrium problem. Unlike existing distributed stochastic Nash equilibrium methods that mainly rely on with-replacement sampling, the proposed approach incorporates without-replacement component updates into equilibrium computation over networks. We first consider a full-information benchmark, for which an intermediate reference trajectory and a shuffling variance are introduced to characterize the epoch-wise dynamics induced by RR. The method is then extended to the more practical partial-decision-information setting, where each player updates its action using local estimates of the joint action profile. For the full-information case, a descent-type bound is established for the RR iterates. For the distributed partial-decision-information case, it is shown that, under constant parameters, the proposed algorithm converges linearly to a neighborhood of the Nash equilibrium, while under diminishing parameters, it converges exactly to the Nash equilibrium almost surely and in mean square. Numerical experiments on an EV charging game and a nonquadratic edge resource admission game demonstrate that RR consistently outperforms the conventional with-replacement SGD baseline in both steady-state accuracy and long-horizon performance.

math.OC↗

arXiv2026

DeepFAN, a transformer-based deep learning model for human-artificial intelligence collaborative assessment of incidental pulmonary nodules in CT scans: a multi-reader, multi-case trial

The widespread adoption of CT has notably increased the number of detected lung nodules. However, current deep learning methods for classifying benign and malignant nodules often fail to comprehensively integrate global and local features, and most of them have not been validated through clinical trials. To address this, we developed DeepFAN, a transformer-based model trained on over 10K pathology-confirmed nodules and further conducted a multi-reader, multi-case clinical trial to evaluate its efficacy in assisting junior radiologists. DeepFAN achieved diagnostic area under the curve (AUC) of 0.939 (95% CI 0.930-0.948) on an internal test set and 0.954 (95% CI 0.934-0.973) on the clinical trial dataset involving 400 cases across three independent medical institutions. Explainability analysis indicated higher contributions from global than local features. Twelve readers' average performance significantly improved by 10.9% (95% CI 8.3%-13.5%) in AUC, 10.0% (95% CI 8.9%-11.1%) in accuracy, 7.6% (95% CI 6.1%-9.2%) in sensitivity, and 12.6% (95% CI 10.9%-14.3%) in specificity (P<0.001 for all). Nodule-level inter-reader diagnostic consistency improved from fair to moderate (overall k: 0.313 vs. 0.421; P=0.019). In conclusion, DeepFAN effectively assisted junior radiologists and may help homogenize diagnostic quality and reduce unnecessary follow-up of indeterminate pulmonary nodules. Chinese Clinical Trial Registry: ChiCTR2400084624.

cs.CV↗

arXiv2026

Experimental investigation into Lagrangian statistics of droplets in homogeneous isotropic turbulence

We experimentally investigate the Lagrangian dynamics of finite-sized, neutrally buoyant droplets in homogeneous isotropic turbulence. The droplet size follows a log-normal distribution whose average value decreases with increasing Reynolds number, reflecting enhanced turbulent breakup. While size-conditioned velocity and acceleration statistics show only weak finite-size dependence, temporal measures reveal clear size-dependent dynamics: larger droplets exhibit longer Lagrangian velocity integral times and an extended ballistic regime in their mean squared displacement. These findings indicate that though droplets exhibit mild deformation and internal circulation, they behave similarly to finite-size rigid particles in terms of Lagrangian dynamics. Our study opens the way to study droplet-laden turbulence and droplet-flow interactions.

physics.flu-dyn↗

arXiv2026

Emergence of transverse optical spin in a subwavelength grating ring resonator

The local polarization of the electromagnetic field plays a crucial role in the interaction of light with spin- and valley-polarized quantum sources. Unlike free-space electromagnetic waves, whose polarization degeneracy enables flexible polarization manipulation, planar integrated optical structures lack such degree of freedom owing to intrinsic structural anisotropy. Here, we propose a planar optical ring resonator based on a subwavelength grating waveguide that supports two quasi-degenerate modes. We demonstrate that coupling of these modes in the ring resonator leads to the formation of the resonances with a predominant direction of electric-field rotation in the vicinity of the resonator, resulting in the non-zero transverse optical spin. The average degree of circular polarization in the proposed structures reaches values of up to 70%. The theoretical predictions are corroborated by experimental validation in the microwave spectral range. Our findings suggest a viable route toward realization of on-chip optical spintronic and valleytronic interfaces.

physics.optics↗

arXiv2026

Asymmetric velocity boundary conditions lead to zonal flow in centrifugal convection

We perform direct numerical simulations of rapidly rotating annular centrifugal convection to investigate how mixed (asymmetric) velocity boundary conditions and geometric curvature shape the flow organisation and heat transfer. Motivated by the quasi-two-dimensionalisation under strong rotation and the long spin-up required for large-scale states, we employ two-dimensional simulations and consider four boundary-condition sets: no-slip/no-slip (INON), no-slip/stress-free (INOS), stress-free/no-slip (ISON) and stress-free/stress-free (ISOS). For fixed geometry with the radius ratio $\eta=0.5$ and over the Rayleigh number $Ra\in[10^6,10^9]$, the heat transfer is strongest for ISOS, followed by INOS and INON, while ISON exhibits a pronounced suppression as a strong zonal flow aligned with the rotation develops. In the three cases dominated by large-scale circulation, the Nusselt number $Nu$ follows an effective classical-type scaling close to $Nu\sim Ra^{0.27}$, whereas the zonal-flow branch displays a much weaker scaling $Nu\sim Ra^{0.1}$ and strong flow anisotropy with the large difference between the radial and azimuthal Reynolds numbers $Re_r\ll Re_\varphi$. A dissipation analysis shows that zonal-flow formation is accompanied by a transition from boundary-layer-dominated dissipation to a relatively low and uniform bulk dissipation, consistent with shear-induced plume suppression. By varying the radius ratio $\eta$, we demonstrate that increasing $\eta$ weakens curvature asymmetry and destabilises the zonal-flow state, leading to roll-dominated convection in the planar limit, and we relate the accompanying bulk-temperature asymmetry to the boundary heat flux asymmetry using a free-convective boundary-layer model.

physics.flu-dyn↗

arXiv2025

HeatV2X: Scalable Heterogeneous Collaborative Perception via Efficient Alignment and Interaction

Vehicle-to-Everything (V2X) collaborative perception extends sensing beyond single vehicle limits through transmission. However, as more agents participate, existing frameworks face two key challenges: (1) the participating agents are inherently multi-modal and heterogeneous, and (2) the collaborative framework must be scalable to accommodate new agents. The former requires effective cross-agent feature alignment to mitigate heterogeneity loss, while the latter renders full-parameter training impractical, highlighting the importance of scalable adaptation. To address these issues, we propose Heterogeneous Adaptation (HeatV2X), a scalable collaborative framework. We first train a high-performance agent based on heterogeneous graph attention as the foundation for collaborative learning. Then, we design Local Heterogeneous Fine-Tuning and Global Collaborative Fine-Tuning to achieve effective alignment and interaction among heterogeneous agents. The former efficiently extracts modality-specific differences using Hetero-Aware Adapters, while the latter employs the Multi-Cognitive Adapter to enhance cross-agent collaboration and fully exploit the fusion potential. These designs enable substantial performance improvement of the collaborative framework with minimal training cost. We evaluate our approach on the OPV2V-H and DAIR-V2X datasets. Experimental results demonstrate that our method achieves superior perception performance with significantly reduced training overhead, outperforming existing state-of-the-art approaches. Our implementation will be released soon.

cs.CV↗

arXiv2025

Freezing and ice aging dynamics in saline water under natural convection

Understanding the coupled dynamics of liquid-solid phase change and fluid flows is crucial in a wide range of geophysical and industrial applications. When freezing occurs in saline water, the newly formed ice is mushy, with a porous structure that traps the brine within the ice. In this work, which combines experiments and theoretical analyses, we investigate the long-term evolution of saline ice, comprehensively accounting for the coupled dynamics of multiscale fluid flow, heat and mass transfer, and phase change. We show that in a closed convective system the rapid formation of a mushy ice layer is followed by desalination (i.e, the expulsion of salt from the ice) processes that might lead to a slow asymptotic decrease of the ice thickness. Desalination of mushy ice reduces its porosity, which alters the dynamic thermal equilibrium and ice thickness by weakening buoyancy-driven convection within the mushy layer. In turn, changes in brine convection and ice thickness affect the further desalination of the ice. The long-term dynamics of the system can be accurately predicted by a one-dimensional model based on appropriate parameterizations of global heat and mass transfer properties. Furthermore, within the same theoretical model we explore the ice dynamics across a broader parameter space. Our findings advance the understanding of the coupled phase-change physics of saline solutions in the presence of convective fluid flows and provide a basis for explaining and predicting real-world phenomena such as the aging of sea ice.

physics.flu-dyn↗

arXiv2025

Conditional Representation Learning for Customized Tasks

Conventional representation learning methods learn a universal representation that primarily captures dominant semantics, which may not always align with customized downstream tasks. For instance, in animal habitat analysis, researchers prioritize scene-related features, whereas universal embeddings emphasize categorical semantics, leading to suboptimal results. As a solution, existing approaches resort to supervised fine-tuning, which however incurs high computational and annotation costs. In this paper, we propose Conditional Representation Learning (CRL), aiming to extract representations tailored to arbitrary user-specified criteria. Specifically, we reveal that the semantics of a space are determined by its basis, thereby enabling a set of descriptive words to approximate the basis for a customized feature space. Building upon this insight, given a user-specified criterion, CRL first employs a large language model (LLM) to generate descriptive texts to construct the semantic basis, then projects the image representation into this conditional feature space leveraging a vision-language model (VLM). The conditional representation better captures semantics for the specific criterion, which could be utilized for multiple customized tasks. Extensive experiments on classification and retrieval tasks demonstrate the superiority and generality of the proposed CRL. The code is available at https://github.com/XLearning-SCU/2025-NeurIPS-CRL.

cs.CV↗

arXiv2025

Lattice-Based Dynamic $k$-Times Anonymous Authentication with Attribute-Based Credentials

With the development of Internet, privacy has become a primary concern of users. Anonymous authentication plays an important role in privacy-preserving systems. A $k$-times anonymous authentication ($k$-TAA) scheme allows a group member to anonymously authenticate to the application provider up to $k$ times. Considering quantum computing attacks, lattice-based $k$-TAA was introduced. However, existing schemes neither support the dynamic granting and revocation of users nor enable users to control the release of their attributes. In this paper, we construct the first lattice-based dynamic $k$-TAA, which offers limited times anonymous authentication, dynamic member management, attribute-based authentication, and post-quantum security. We present a concrete construction, and reduce its security to standard complexity assumptions. Notably, compared with existing lattice-based $k$-TAA, our scheme is efficient in terms of communication cost.

cs.CR↗

arXiv2025

Physical Mechanism behind the Early Onset of the Ultimate State in Supergravitational Centrifugal Thermal Convection

We present a combined experimental and numerical investigation of the transition from the classical to the ultimate regime of thermal turbulence in a supergravitational centrifugal convection system. The transition is found to be robust, with the critical Rayleigh number decreasing systematically as the Froude number, defined as the ratio of centrifugal to Earth's gravity, decreases, highlighting the effect of residual gravity. Once the Rayleigh number reaches the transition threshold, the Stewartson layer induced by residual Earth gravity becomes comparable in thickness to the viscous boundary layer, and their interaction results in a coupled flow that distorts the viscous boundary layer, triggering its transition from laminar to turbulent flow and leading to a sharp increase in heat transport. These findings demonstrate the key role of the Stewartson layer induced by residual gravity in facilitating the transition to the ultimate regime in supergravitational centrifugal thermal convection.

physics.flu-dyn↗

arXiv2025

Non-reciprocal coalescence-breakup dynamics in concentrated emulsions

Dense stabilized emulsions are mixtures of immiscible fluids where the high-volume fraction droplet dispersed phase is stabilized against coalescence by steric interactions. The production of emulsions involves high-shear flows and it is well known that at a critical volume fraction the emulsion loses stability, undergoing an extremely rapid process where the fluid components in the emulsion exchange roles. This process, called catastrophic phase inversion, which resembles in several respects a dynamical phase transition, has remained widely elusive from an experimental and theoretical point of view. In this work, we present state-of-the-art experimental and numerical data to support a dynamical-system framework capable of precisely highlighting the dynamics occurring in the system as it approaches the catastrophic phase inversion. Our study clearly highlights that at high volume fractions, dynamical changes in the emulsion morphology, due to coalescence and breakup of droplets, play a critical role in determining emulsion's rheology and stability. Additionally, we show that at approaching the critical volume fractions, the dynamics can be simplified as being controlled by the dynamics of a correlation length represented, in our systems, by the size of the largest droplet. This dynamics shares a close connection with non-reciprocal phase transition where two different physical mechanisms, coalescence and breakups, can get out of balance leading to large non-symmetric periodic excursions in phase space. We clarify the phenomenology observed and quantitatively explain the essential aspect of the highly complex dynamics of stabilized emulsions undergoing catastrophic phase inversion.

cond-mat.soft↗

arXiv2025

Inner-product Functional Encryption with Fine-grained Revocation for Flexible EHR Sharing

E-health record (EHR) contains a vast amount of continuously growing medical data and enables medical institutions to access patient health data conveniently.This provides opportunities for medical data mining which has important applications in identifying high-risk patients and improving disease diagnosis, etc.Since EHR contains sensitive patient information, how to protect patient privacy and enable mining on EHR data is important and challenging.Traditional public key encryption (PKE) can protect patient privacy, but cannot support flexible selective computation on encrypted EHR data.Functional encryption (FE) allows authorised users to compute function values of encrypted data without releasing other information, hence supporting selective computation on encrypted data. Nevertheless, existing FE schemes do not support fine-grained revocation and update, so they are unsuitable for EHR system. In this paper,we first propose an inner-product functional encryption with fine-grained revocation (IPFE-FR) scheme, and then apply it to a flexible EHR sharing system. Our scheme possesses the following features:(1) a group manager can revoke a specific function computation of medical institutions on encrypted EHR data,instead of all function computation rights. (2) a revoked medical institution is not allowed to compute the function value of encrypted EHR data not only generated after the revocation, but also generated before the revocation. (3) secret keys issued to the same medical institution are bound together to prevent collusion attacks. The formal definition and security model of the IPFE-FR scheme are proposed.Furthermore, we present a concrete construction and reduce its security to the Learning with Errors (LWE) assumption which is quantum-resistant. Finally, the theoretical analysis and experimental implementation of our scheme are conducted to show its efficiency.

cs.CR↗