arXiv subjects

Yuan Gao

Publications and source records attributed to Yuan Gao.

At least 217 records · Page 12Linked to original sources

arXiv2025

Enabling 6G Through Multi-Domain Channel Extrapolation: Opportunities and Challenges of Generative Artificial Intelligence

Channel extrapolation has attracted wide attention due to its potential to acquire channel state information (CSI) with high accuracy and minimal overhead. This is becoming increasingly crucial as the sixth-generation (6G) mobile networks aim to support complex scenarios, for example, high-mobility communications utilizing ultra-massive multiple-input multiple-output (MIMO) technologies and broad spectrum bands, necessitating multi-domain channel extrapolation. Current research predominantly addresses channel extrapolation within a single domain, lacking a comprehensive approach to multi-domain channel extrapolation. To bridge the gap, we propose the concept of multi-domain channel extrapolation, detailing the essential performance requirements for 6G networks. These include precise channel extrapolation, adaptability to varying scenarios, and manageable computational complexity during both training and inference stages. In light of these requirements, we elaborate the potential and challenges of incorporating generative artificial intelligence (GAI)-based models for effective multi-domain channel extrapolation. Given the ability of the Transformer to capture long-range dependencies and hidden patterns, we propose a novel Transformer encoder-like model by eliminating the positional encoding module and replacing the original multi-head attention with a multilayer perceptron (MLP) for multi-domain channel extrapolation. Simulation results indicate that this model surpasses existing baseline models in terms of extrapolation accuracy and inference speed. Ablation studies further demonstrate the effectiveness of the module design of the proposed design. Finally, we pose several open questions for the development of practical GAI-based multi-domain channel extrapolation models, including the issues of explainability, generalization, and dataset collection.

eess.SP↗

arXiv2025

Apple Intelligence Foundation Language Models: Tech Report 2025

We introduce two multilingual, multimodal foundation language models that power Apple Intelligence features across Apple devices and services: i a 3B-parameter on-device model optimized for Apple silicon through architectural innovations such as KV-cache sharing and 2-bit quantization-aware training; and ii a scalable server model built on a novel Parallel-Track Mixture-of-Experts PT-MoE transformer that combines track parallelism, mixture-of-experts sparse computation, and interleaved global-local attention to deliver high quality with competitive cost on Apple's Private Cloud Compute platform. Both models are trained on large-scale multilingual and multimodal datasets sourced via responsible web crawling, licensed corpora, and high-quality synthetic data, then further refined with supervised fine-tuning and reinforcement learning on a new asynchronous platform. The resulting models support several additional languages while understanding images and executing tool calls. In public benchmarks and human evaluations, both the server model and the on-device model match or surpass comparably sized open baselines. A new Swift-centric Foundation Models framework exposes guided generation, constrained tool calling, and LoRA adapter fine-tuning, allowing developers to integrate these capabilities with a few lines of code. The latest advancements in Apple Intelligence models are grounded in our Responsible AI approach with safeguards like content filtering and locale-specific evaluation, as well as our commitment to protecting our users' privacy with innovations like Private Cloud Compute.

cs.LG↗

arXiv2025

Fast-slow chemical reactions: convergence of Hamilton-Jacobi equation and variational representation

Microscopic behaviors of chemical reactions can be described by a random time-changed Poisson process, whose large-volume limit determines the macroscopic behaviors of species concentrations, including both typical and non-typical trajectories. When the reaction intensities (or fluxes) exhibit a separation of fast-slow scales, the macroscopic typical trajectory is governed by a system of $\varepsilon$-dependent nonlinear reaction rate equations (RRE), while the non-typical trajectories deviating from the typical ones are characterized by an $\varepsilon$-dependent exponentially nonlinear Hamilton-Jacobi equation (HJE). In this paper, for general chemical reactions, we study the fast-slow limit as $\varepsilon\to 0$ for the viscosity solutions of the associated HJE with a state-constrained boundary condition. We identify the limiting effective HJE on a slow manifold, along with an effective variational representation for the solution. Through the uniform convergence of the viscosity solutions and the $Γ$-convergence of the variational solution representations, we rigorously show that all non-typical (and also typical) trajectories are concentrated on the slow manifold and the effective macroscopic dynamics are described by the coarse-grained RRE and HJE, respectively. This approach for studying the fast-slow limit is applicable to, but not limited to, reversible chemical reactions described by gradient flows.

math.AP↗

arXiv2025

Homogenization of Wasserstein gradient flows

We prove the convergence of a Wasserstein gradient flow of a free energy in inhomogeneous media. Both the energy and media can depend on the spatial variable in a fast oscillatory manner. In particular, we show that the gradient-flow structure is preserved in the limit which is expressed in terms of an effective energy and Wasserstein metric. The gradient flow and its limiting behavior are analyzed through an energy dissipation inequality (EDI). The result is consistent with asymptotic analysis in the realm of homogenization. However, we note that the effective metric is in general different from that obtained from the Gromov-Hausdorff convergence of metric spaces. We apply our framework to a linear Fokker-Planck equation but we believe the approach is robust enough to be applicable in a broader context.

math.AP↗

arXiv2025

DualPhys-GS: Dual Physically-Guided 3D Gaussian Splatting for Underwater Scene Reconstruction

In 3D reconstruction of underwater scenes, traditional methods based on atmospheric optical models cannot effectively deal with the selective attenuation of light wavelengths and the effect of suspended particle scattering, which are unique to the water medium, and lead to color distortion, geometric artifacts, and collapsing phenomena at long distances. We propose the DualPhys-GS framework to achieve high-quality underwater reconstruction through a dual-path optimization mechanism. Our approach further develops a dual feature-guided attenuation-scattering modeling mechanism, the RGB-guided attenuation optimization model combines RGB features and depth information and can handle edge and structural details. In contrast, the multi-scale depth-aware scattering model captures scattering effects at different scales using a feature pyramid network and an attention mechanism. Meanwhile, we design several special loss functions. The attenuation scattering consistency loss ensures physical consistency. The water body type adaptive loss dynamically adjusts the weighting coefficients. The edge-aware scattering loss is used to maintain the sharpness of structural edges. The multi-scale feature loss helps to capture global and local structural information. In addition, we design a scene adaptive mechanism that can automatically identify the water-body-type characteristics (e.g., clear coral reef waters or turbid coastal waters) and dynamically adjust the scattering and attenuation parameters and optimization strategies. Experimental results show that our method outperforms existing methods in several metrics, especially in suspended matter-dense regions and long-distance scenes, and the reconstruction quality is significantly improved.

cs.GR↗

arXiv2025

Score Augmentation for Diffusion Models

Diffusion models have achieved remarkable success in generative modeling. However, this study confirms the existence of overfitting in diffusion model training, particularly in data-limited regimes. To address this challenge, we propose Score Augmentation (ScoreAug), a novel data augmentation framework specifically designed for diffusion models. Unlike conventional augmentation approaches that operate on clean data, ScoreAug applies transformations to noisy data, aligning with the inherent denoising mechanism of diffusion. Crucially, ScoreAug further requires the denoiser to predict the augmentation of the original target. This design establishes an equivariant learning objective, enabling the denoiser to learn scores across varied denoising spaces, thereby realizing what we term score augmentation. We also theoretically analyze the relationship between scores in different spaces under general transformations. In experiments, we extensively validate ScoreAug on multiple benchmarks including CIFAR-10, FFHQ, AFHQv2, and ImageNet, with results demonstrating significant performance improvements over baselines. Notably, ScoreAug effectively mitigates overfitting across diverse scenarios, such as varying data scales and model capacities, while exhibiting stable convergence properties. Another advantage of ScoreAug over standard data augmentation lies in its ability to circumvent data leakage issues under certain conditions. Furthermore, we show that ScoreAug can be synergistically combined with traditional data augmentation techniques to achieve additional performance gains.

cs.LG↗

arXiv2025

EvoMakeup: High-Fidelity and Controllable Makeup Editing with MakeupQuad

Facial makeup editing aims to realistically transfer makeup from a reference to a target face. Existing methods often produce low-quality results with coarse makeup details and struggle to preserve both identity and makeup fidelity, mainly due to the lack of structured paired data -- where source and result share identity, and reference and result share identical makeup. To address this, we introduce MakeupQuad, a large-scale, high-quality dataset with non-makeup faces, references, edited results, and textual makeup descriptions. Building on this, we propose EvoMakeup, a unified training framework that mitigates image degradation during multi-stage distillation, enabling iterative improvement of both data and model quality. Although trained solely on synthetic data, EvoMakeup generalizes well and outperforms prior methods on real-world benchmarks. It supports high-fidelity, controllable, multi-task makeup editing -- including full-face and partial reference-based editing, as well as text-driven makeup editing -- within a single model. Experimental results demonstrate that our method achieves superior makeup fidelity and identity preservation, effectively balancing both aspects. Code and dataset will be released upon acceptance.

cs.CV↗

arXiv2025

Universal Magnetocaloric Effect near Quantum Critical Point of Magnon Bose-Einstein Condensation

Bose-Einstein condensation (BEC), a macroscopic quantum phenomenon arising from phase coherence and bosonic statistics, has been realized in quantum magnets. Here, we report the observation of a universal magnetocaloric effect (MCE) near a BEC quantum critical point (QCP) in copper sulfate crystal ($CuSO_4 \cdot 5H_2O$). By conducting magnetocaloric and nuclear magnetic resonance measurements, we uncover a field-driven BEC QCP, evidenced by the universal scaling law $T_c \propto (B_c - B)^{2/3}$ and the perfect data collapse of the magnetic Grüneisen ratio. Thermal excitation triggers a dimensional crossover to a 1D quantum-critical regime, where the MCE scaling strictly matches the universality class of 1D Fermi gases. Notably, the quantum-critical MCE enables cooling down to 12.8 mK without helium-3, with very fast thermal relaxation rate that is critical for high cooling power. This work demonstrates the universal MCE in magnon BEC systems, using a common copper sulfate compound as a paradigmatic example, and paves the way for next-generation sub-Kelvin cooling.

cond-mat.str-el↗

arXiv2025

AI-driven Wireless Positioning: Fundamentals, Standards, State-of-the-art, and Challenges

Wireless positioning technologies hold significant value for applications in autonomous driving, extended reality (XR), unmanned aerial vehicles (UAVs), and more. With the advancement of artificial intelligence (AI), leveraging AI to enhance positioning accuracy and robustness has emerged as a field full of potential. Driven by the requirements and functionalities defined in the 3rd Generation Partnership Project (3GPP) standards, AI/machine learning (ML)-based cellular positioning is becoming a key technology to overcome the limitations of traditional methods. This paper presents a comprehensive survey of AI-driven cellular positioning. We begin by reviewing the fundamentals of wireless positioning and AI models, analyzing their respective challenges and synergies. We provide a comprehensive review of the evolution of 3GPP positioning standards, with a focus on the integration of AI/ML in current and upcoming standard releases. Guided by the 3GPP-defined taxonomy, we categorize and summarize state-of-the-art (SOTA) research into two major classes: AI/ML-assisted positioning and direct AI/ML-based positioning. The former includes line-of-sight (LOS)/non-line-of-sight (NLOS) detection, time of arrival (TOA)/time difference of arrival (TDOA) estimation, and angle prediction; the latter encompasses fingerprinting, knowledge-assisted learning, and channel charting. Furthermore, we review representative public datasets and conduct performance evaluations of AI-based positioning algorithms using these datasets. Finally, we conclude by summarizing the challenges and opportunities of AI-driven wireless positioning.

eess.SP↗

arXiv2025

DepR: Depth Guided Single-view Scene Reconstruction with Instance-level Diffusion

We propose DepR, a depth-guided single-view scene reconstruction framework that integrates instance-level diffusion within a compositional paradigm. Instead of reconstructing the entire scene holistically, DepR generates individual objects and subsequently composes them into a coherent 3D layout. Unlike previous methods that use depth solely for object layout estimation during inference and therefore fail to fully exploit its rich geometric information, DepR leverages depth throughout both training and inference. Specifically, we introduce depth-guided conditioning to effectively encode shape priors into diffusion models. During inference, depth further guides DDIM sampling and layout optimization, enhancing alignment between the reconstruction and the input image. Despite being trained on limited synthetic data, DepR achieves state-of-the-art performance and demonstrates strong generalization in single-view scene reconstruction, as shown through evaluations on both synthetic and real-world datasets.

cs.CV↗

arXiv2025

Fuse Before Transfer: Knowledge Fusion for Heterogeneous Distillation

Most knowledge distillation (KD) methodologies predominantly focus on teacher-student pairs with similar architectures, such as both being convolutional neural networks (CNNs). However, the potential and flexibility of KD can be greatly improved by expanding it to novel Cross-Architecture KD (CAKD), where the knowledge of homogeneous and heterogeneous teachers can be transferred flexibly to a given student. The primary challenge in CAKD lies in the substantial feature gaps between heterogeneous models, originating from the distinction of their inherent inductive biases and module functions. To this end, we introduce an assistant model as a bridge to facilitate smooth feature knowledge transfer between heterogeneous teachers and students. More importantly, within our proposed design principle, the assistant model combines the advantages of cross-architecture inductive biases and module functions by merging convolution and attention modules derived from both student and teacher module functions. Furthermore, we observe that heterogeneous features exhibit diverse spatial distributions in CAKD, hindering the effectiveness of conventional pixel-wise mean squared error (MSE) loss. Therefore, we leverage a spatial-agnostic InfoNCE loss to align features after spatial smoothing, thereby improving the feature alignments in CAKD. Our proposed method is evaluated across some homogeneous model pairs and arbitrary heterogeneous combinations of CNNs, ViTs, and MLPs, achieving state-of-the-art performance for distilled models with a maximum gain of 11.47% on CIFAR-100 and 3.67% on ImageNet-1K. Our code and models will be released.

cs.CV↗

arXiv2025

Hydrodynamic Insight Drives Multimodal Light_Field Dynamics via Streamline Engineering

Since the 1970s, analogies between laser dynamics and fluid systems have provided insight into phenomena such as chaos, multistability, and turbulence. Building on this perspective, we model the optical field as an energy fluid and interpret Poynting-vector trajectories as energy streamlines, yielding a unified, three_dimensional map of light's free-space dynamics. By sculpting these streamlines, we develop an approach to talior vortex-beam propagation dynamics that suppresses both diffraction- and OAM-induced broadening. Extending this method to general structured modes, we enable a single field to exhibit customizable multimodal dynamics that integrate features from primary structured light families: the diffraction-free, self-healing behavior of Bessel beams; the tunable self-similarity of Laguerre-Gaussian beams and adjustable self-acceleration of Airy beams. Additionally, it allows for adjustable propagating energy-density profiles to counteract losses. Optical-tweezer experiments,analogous to particle-tracking velocimetry in fluid dynamics, show that trapped microspheres closely follow the designed streamlines, validating the streamline geometries and indicating a potential route toward precision 3D optomechanical control. In a proof-of-principle free-space communication experiment, vortex beams with customized multimodal dynamics demonstrate several improvements, including more independent channels, reduced turbulence-induced mode scattering, and robust non-line-of-sight transmission. Together, the streamline-engineering approach offers a unified and adaptable strategy for tailoring light's propagation dynamics, with potential applications in precision optomechanics, optofluidics, and advanced optical networking.

physics.optics↗

arXiv2025

From Conditional to Unconditional Independence: Testing Conditional Independence via Transport Maps

Testing conditional independence between two random vectors given a third is a fundamental and challenging problem in statistics, particularly in multivariate nonparametric settings due to the complexity of conditional structures. We propose a novel method for testing conditional independence by transforming it to an unconditional independence test problem. We achieve this by constructing two transport maps that transform conditional independence into unconditional independence, this substantially simplifies the problem. These transport maps are estimated from data using conditional continuous normalizing flow models. Within this framework, we derive a test statistic and prove its asymptotic validity under both the null and alternative hypotheses. A permutation-based procedure is employed to evaluate the significance of the test. We validate the proposed method through extensive simulations and real-data analysis. Our numerical studies demonstrate the practical effectiveness of the proposed method for conditional independence

stat.ML↗

arXiv2025

Vortex-Free Intrinsic Orbital Angular Momentum

Optical orbital angular momentum (OAM) has traditionally relied on vortex beams with helical phase fronts imparting quantized intrinsic OAM. Here, we introduce a fundamentally vortex_free framework where intrinsic OAM arises from the natural curvature of lights energy flow, specifically, the caustic geometry of self_accelerating beams whose curved trajectories act as orbital highways for photons. This OAM generation mechanism is independent of phase vortices but mirrors celestial orbital motion. Through numerical simulations, experimental characterization, and optomechanical measurements using optical tweezers, we demonstrate intrinsic vortex_free OAM rooted solely in beam intensity architecture. Generalizing beyond geometric caustics to arbitrary optical fields, we demonstrate OAM via curved Poynting_vector energy streamlines, unifying conventional vortex and novel vortex_free OAM under a single quantitative framework. Streamline engineering enables customizable rotational dynamics, including hybrid orbital_cyclonic motions reminiscent of tropical storms, with promising applications in precision optomechanics, optofluidics, and optical analogues of fluid dynamics. This energy-flow perspective offers a versatile platform for designing and quantifying OAM across structured light.

physics.optics↗

arXiv2025

From Words to Collisions: LLM-Guided Evaluation and Adversarial Generation of Safety-Critical Driving Scenarios

Ensuring the safety of autonomous vehicles requires virtual scenario-based testing, which depends on the robust evaluation and generation of safety-critical scenarios. So far, researchers have used scenario-based testing frameworks that rely heavily on handcrafted scenarios as safety metrics. To reduce the effort of human interpretation and overcome the limited scalability of these approaches, we combine Large Language Models (LLMs) with structured scenario parsing and prompt engineering to automatically evaluate and generate safety-critical driving scenarios. We introduce Cartesian and Ego-centric prompt strategies for scenario evaluation, and an adversarial generation module that modifies trajectories of risk-inducing vehicles (ego-attackers) to create critical scenarios. We validate our approach using a 2D simulation framework and multiple pre-trained LLMs. The results show that the evaluation module effectively detects collision scenarios and infers scenario safety. Meanwhile, the new generation module identifies high-risk agents and synthesizes realistic, safety-critical scenarios. We conclude that an LLM equipped with domain-informed prompting techniques can effectively evaluate and generate safety-critical driving scenarios, reducing dependence on handcrafted metrics. We release our open-source code and scenarios at: https://github.com/TUM-AVS/From-Words-to-Collisions.

cs.AI↗

arXiv2025

LongDocURL: a Comprehensive Multimodal Long Document Benchmark Integrating Understanding, Reasoning, and Locating

Large vision language models (LVLMs) have improved the document understanding capabilities remarkably, enabling the handling of complex document elements, longer contexts, and a wider range of tasks. However, existing document understanding benchmarks have been limited to handling only a small number of pages and fail to provide a comprehensive analysis of layout elements locating. In this paper, we first define three primary task categories: Long Document Understanding, numerical Reasoning, and cross-element Locating, and then propose a comprehensive benchmark, LongDocURL, integrating above three primary tasks and comprising 20 sub-tasks categorized based on different primary tasks and answer evidences. Furthermore, we develop a semi-automated construction pipeline and collect 2,325 high-quality question-answering pairs, covering more than 33,000 pages of documents, significantly outperforming existing benchmarks. Subsequently, we conduct comprehensive evaluation experiments on both open-source and closed-source models across 26 different configurations, revealing critical performance gaps in this field.

cs.AI↗

arXiv2025

DpDNet: An Dual-Prompt-Driven Network for Universal PET-CT Segmentation

PET-CT lesion segmentation is challenging due to noise sensitivity, small and variable lesion morphology, and interference from physiological high-metabolic signals. Current mainstream approaches follow the practice of one network solving the segmentation of multiple cancer lesions by treating all cancers as a single task. However, this overlooks the unique characteristics of different cancer types. Considering the specificity and similarity of different cancers in terms of metastatic patterns, organ preferences, and FDG uptake intensity, we propose DpDNet, a Dual-Prompt-Driven network that incorporates specific prompts to capture cancer-specific features and common prompts to retain shared knowledge. Additionally, to mitigate information forgetting caused by the early introduction of prompts, prompt-aware heads are employed after the decoder to adaptively handle multiple segmentation tasks. Experiments on a PET-CT dataset with four cancer types show that DpDNet outperforms state-of-the-art models. Finally, based on the segmentation results, we calculated MTV, TLG, and SUVmax for breast cancer survival analysis. The results suggest that DpDNet has the potential to serve as a valuable tool for personalized risk stratification, supporting clinicians in optimizing treatment strategies and improving outcomes. Code is available at https://github.com/XinglongLiang08/DpDNet.

eess.IV↗

arXiv2025

PLUS: Plug-and-Play Enhanced Liver Lesion Diagnosis Model on Non-Contrast CT Scans

Focal liver lesions (FLL) are common clinical findings during physical examination. Early diagnosis and intervention of liver malignancies are crucial to improving patient survival. Although the current 3D segmentation paradigm can accurately detect lesions, it faces limitations in distinguishing between malignant and benign liver lesions, primarily due to its inability to differentiate subtle variations between different lesions. Furthermore, existing methods predominantly rely on specialized imaging modalities such as multi-phase contrast-enhanced CT and magnetic resonance imaging, whereas non-contrast CT (NCCT) is more prevalent in routine abdominal imaging. To address these limitations, we propose PLUS, a plug-and-play framework that enhances FLL analysis on NCCT images for arbitrary 3D segmentation models. In extensive experiments involving 8,651 patients, PLUS demonstrated a significant improvement with existing methods, improving the lesion-level F1 score by 5.66%, the malignant patient-level F1 score by 6.26%, and the benign patient-level F1 score by 4.03%. Our results demonstrate the potential of PLUS to improve malignant FLL screening using widely available NCCT imaging substantially.

eess.IV↗