arXiv subjects

Yu Han

Publications and source records attributed to Yu Han.

At least 73 records · Page 4Linked to original sources

arXiv2025

A Systematic Review on Foundation Models for Electrocardiogram Analysis: Initial Strides and Expansive Horizons

Electrocardiogram (ECG) is widely used in healthcare applications, such as arrhythmia detection and sleep monitoring, making accurate ECG analysis critically essential. Traditional deep learning models for ECG are task-specific, with limited generalization and narrow functionality. Foundation models (FMs), or large pre-training models, have recently advanced representation learning, enabling strong performance across diverse tasks and motivating their adoption for ECG analysis. Here, we present the first comprehensive review dedicated to ECG foundation models (ECG-FMs). We map the current landscape of architectures, pre-training paradigms, and adaptation strategies, and critically examine their strengths, limitations, and clinical potential. By consolidating this emerging field, we aim to accelerate the development of robust, generalizable ECG-FMs and chart future directions for their integration into healthcare practice.

eess.SP↗

arXiv2025

Pioneering Scalable Prototyping for Mid-Band XL-MIMO Systems: Design and Implementation

The mid-band frequency range, combined with extra large-scale multiple-input multiple-output (XL-MIMO), is emerging as a key enabler for future communication systems. Thanks to the advent of new spectrum resources and degrees of freedom brought by the near-field propagation, the mid-band XL-MIMO system is expected to significantly enhance throughput and inherently support advanced functionalities such as integrated sensing and communication. Although theoretical studies have highlighted the benefits of mid-band XL-MIMO systems, the promised performance gains have yet to be validated in practical systems, posing a major challenge to the standardization. In this paper, preliminaries are first discussed, followed by an analysis of key challenges in constructing a real-time prototype system. Subsequently, the design and implementation of a real-time mid-band XL-MIMO prototype system are presented. Benefiting from the novel architecture, the proposed prototype system supports metrics aligned with standardization, including a bandwidth of 200 MHz, up to 1024 antenna elements, and up to 256 transceiver chains. Operating in time-division duplexing (TDD) mode, the prototype enables multiuser communication with support for up to 12 users, while retaining standard communication procedures. Built on software-defined radio (SDR) platforms, the system is programmable and allows for flexible deployment of advanced algorithms. Moreover, the modular architecture ensures high scalability, making the system adaptable to various configurations, including distributed deployments and decentralized signal processing. Experimental results with the proposed prototype system demonstrate real-time digital sample processing at 1167.85 Gbps, a peak data throughput of 15.81 Gbps for 12 users, and a maximal spectral efficiency approaching 80 bit/s/Hz.

eess.SP↗

arXiv2025

On the Slipperiness of Surfactants: Charge-Mediated Friction Control at the Molecular Scale

From soap-covered dishes to freshly cleaned floors, surfactants can make surfaces slippery; yet, the underlying mechanism remains poorly understood. Here, we identify the molecular origin behind this ubiquitous phenomenon using macroscopic tribology and surface molecular spectroscopy. We demonstrate that reducing friction through surfactants hinges on charge complementarity: surfactants of opposite charge to the solid surface reduce friction even at extreme contact pressures, whereas like-charged or neutral surfactants are ineffective. Oppositely charged surfactants self-assemble into dense and robust molecular brushes, creating a persistent lubrication beyond the limits of conventional mechanisms. This charge-mediated approach offers a universal and scalable framework for friction control across length scales without significant surface modification.

physics.chem-ph↗

arXiv2025

Standard Applicability Judgment and Cross-jurisdictional Reasoning: A RAG-based Framework for Medical Device Compliance

Identifying the appropriate regulatory standard applicability remains a critical yet understudied challenge in medical device compliance, frequently necessitating expert interpretation of fragmented and heterogeneous documentation across different jurisdictions. To address this challenge, we introduce a modular AI system that leverages a retrieval-augmented generation (RAG) pipeline to automate standard applicability determination. Given a free-text device description, our system retrieves candidate standards from a curated corpus and uses large language models to infer jurisdiction-specific applicability, classified as Mandatory, Recommended, or Not Applicable, with traceable justifications. We construct an international benchmark dataset of medical device descriptions with expert-annotated standard mappings, and evaluate our system against retrieval-only, zero-shot, and rule-based baselines. The proposed approach attains a classification accuracy of 73% and a Top-5 retrieval recall of 87%, demonstrating its effectiveness in identifying relevant regulatory standards. We introduce the first end-to-end system for standard applicability reasoning, enabling scalable and interpretable AI-supported regulatory science. Notably, our region-aware RAG agent performs cross-jurisdictional reasoning between Chinese and U.S. standards, supporting conflict resolution and applicability justification across regulatory frameworks.

cs.AI↗

arXiv2025

Electric Field Control of Spin Orbit Coupling and Circular Photogalvanic Effect in a True Ferrielectric Crystal

Materials possessing long range ordering of magnetic spins or electric dipoles have been the focus of condensed matter research. Among them, ferri-systems with two sublattices of unequal/noncollinear spins or electric dipoles are expected to combine the properties of ferro- and antiferro-systems, but lack experimental observations in single phase materials. This is particularly true for the ferrielectric system, since the electric dipoles usually can be redefined to incorporate the two sublattices into one, making it indistinguishable from ferroelectric. This raises doubts about whether or not ferrielectricity can be considered as an independent ferroic order. Here we report the observation of true ferrielectric behaviors in a hybrid single crystal (MV)[SbBr5] (MV2+ = N,N'-dimethyl-4,4'-bipyridinium or methylviologen), where the two electric dipole sublattices switch asynchronously, thus cannot be reduced to ferroelectric by redefining the unit cell. Furthermore, the complex dipole configuration imparts circularly polarized light sensitivity to the system. An electric field can modulate the non-collinear dipole sublattices and even induce a transition from ferrielectric to ferroelectric state, thereby tuning the helicity-dependent photocurrent. This study opens a new paradigm for the study of true irreducible ferrielectricity (a new class of polar system) and provides an effective approach to the electric field control of spin-orbit coupling and circular photogalvanic effect.

cond-mat.mtrl-sci↗

arXiv2025

AI for Regulatory Affairs: Balancing Accuracy, Interpretability, and Computational Cost in Medical Device Classification

Regulatory affairs, which sits at the intersection of medicine and law, can benefit significantly from AI-enabled automation. Classification task is the initial step in which manufacturers position their products to regulatory authorities, and it plays a critical role in determining market access, regulatory scrutiny, and ultimately, patient safety. In this study, we investigate a broad range of AI models -- including traditional machine learning (ML) algorithms, deep learning architectures, and large language models -- using a regulatory dataset of medical device descriptions. We evaluate each model along three key dimensions: accuracy, interpretability, and computational cost.

cs.AI↗

arXiv2025

A Review of Stop-and-Go Traffic Wave Suppression Strategies: Variable Speed Limit vs. Jam-Absorption Driving

The main form of freeway traffic congestion is the familiar stop-and-go wave, characterized by wide moving jams that propagate indefinitely upstream provided enough traffic demand. They cause severe, long-lasting adverse effects, such as reduced traffic efficiency, increased driving risks, and higher vehicle emissions. This underscores the crucial importance of artificial intervention in the propagation of stop-and-go waves. Over the past two decades, two prominent strategies for stop-and-go wave suppression have emerged: variable speed limit (VSL) and jam-absorption driving (JAD). Although they share similar research motivations, objectives, and theoretical foundations, the development of these strategies has remained relatively disconnected. To synthesize fragmented advances and drive the field forward, this paper first provides a comprehensive review of the achievements in the stop-and-go wave suppression-oriented VSL and JAD, respectively. It then focuses on bridging the two areas and identifying research opportunities from the following perspectives: fundamental diagrams, secondary waves, generalizability, traffic state estimation and prediction, robustness to randomness, scenarios for strategy validation, and field tests and practical deployment. We expect that through this review, one area can effectively address its limitations by identifying and leveraging the strengths of the other, thus promoting the overall research goal of freeway stop-and-go wave suppression.

physics.soc-ph↗

arXiv2025

Extract the Best, Discard the Rest: CSI Feedback with Offline Large AI Models

Large AI models (LAMs) have shown strong potential in wireless communication tasks, but their practical deployment remains hindered by latency and computational constraints. In this work, we focus on the challenge of integrating LAMs into channel state information (CSI) feedback for frequency-division duplex (FDD) massive multiple-intput multiple-output (MIMO) systems. To this end, we propose two offline frameworks, namely site-specific LAM-enhanced CSI feedback (SSLCF) and multi-scenario LAM-enhanced CSI feedback (MSLCF), that incorporate LAMs into the codebook-based CSI feedback paradigm without requiring real-time inference. Specifically, SSLCF generates a site-specific enhanced codebook through fine-tuning on locally collected CSI data, while MSLCF improves generalization by pre-generating a set of environment-aware codebooks. Both of these frameworks build upon the LAM with vision-based backbone, which is pre-trained on large-scale image datasets and fine-tuned with CSI data to generate customized codebooks. This resulting network named LVM4CF captures the structural similarity between CSI and image, allowing the LAM to refine codewords tailored to the specific environments. To optimize the codebook refinement capability of LVM4CF under both single- and dual-side deployment modes, we further propose corresponding training and inference algorithms. Simulation results show that our frameworks significantly outperform existing schemes in both reconstruction accuracy and system throughput, without introducing additional inference latency or computational overhead. These results also support the core design methodology of our proposed frameworks, extracting the best and discarding the rest, as a promising pathway for integrating LAMs into future wireless systems.

eess.SP↗

arXiv2025

Reconfigurable Intelligent Surface Aided Integrated Communication and Localization with a Single Access Point

Reconfigurable intelligent surfaces (RISs) not only assist communication but also help the localization of user equipment (UE). This study focuses on the indoor localization of UE with a single access point (AP) aided by multiple RISs. First, we propose a two-stage channel estimation scheme where the phase shifts of RIS elements are tuned to obtain multiple channel soundings. In the first stage, the newtonized orthogonal matching pursuit algorithm extracts the parameters of multiple paths from the received signals. Then, the LOS path and RIS-reflected paths are identified. In the second stage, the estimated path gains of RIS-reflected paths with different phase shifts are utilized to determine the angle of arrival (AOA) at the RIS by obtaining the angular pseudo spectrum. Consequently, by taking the AP and RISs as reference points, the linear least squares estimator can locate UE with the estimated AOAs. Simulation results show that the proposed algorithm can realize centimeter-level localization accuracy in the discussed scenarios. Moreover, the higher accuracy of pseudo spectrum, a larger number of channel soundings, and a larger number of reference points can realize higher localization accuracy of UE.

cs.IT↗

arXiv2025

Toward Automated Regulatory Decision-Making: Trustworthy Medical Device Risk Classification with Multimodal Transformers and Self-Training

Accurate classification of medical device risk levels is essential for regulatory oversight and clinical safety. We present a Transformer-based multimodal framework that integrates textual descriptions and visual information to predict device regulatory classification. The model incorporates a cross-attention mechanism to capture intermodal dependencies and employs a self-training strategy for improved generalization under limited supervision. Experiments on a real-world regulatory dataset demonstrate that our approach achieves up to 90.4% accuracy and 97.9% AUROC, significantly outperforming text-only (77.2%) and image-only (54.8%) baselines. Compared to standard multimodal fusion, the self-training mechanism improved SVM performance by 3.3 percentage points in accuracy (from 87.1% to 90.4%) and 1.4 points in macro-F1, suggesting that pseudo-labeling can effectively enhance generalization under limited supervision. Ablation studies further confirm the complementary benefits of both cross-modal attention and self-training.

cs.LG↗

arXiv2025

Distributed U6G ELAA Communication Systems: Channel Measurement and Small-Scale Fading Characterization

The distributed upper 6 GHz (U6G) extra-large scale antenna array (ELAA) is a key enabler for future wireless communication systems, offering higher throughput and wider coverage, similar to existing ELAA systems, while effectively mitigating unaffordable complexity and hardware overhead. Uncertain channel characteristics, however, present significant bottleneck problems that hinder the hardware structure and algorithm design of the distributed U6G ELAA system. In response, we construct a U6G channel sounder and carry out extensive measurement campaigns across various typical scenarios. Initially, U6G channel characteristics, particularly small-scale fading characteristics, are unveiled and compared across different scenarios. Subsequently, the U6G ELAA channel characteristics are analyzed using a virtual array comprising 64 elements. Furthermore, inspired by the potential for distributed processing, we investigate U6G ELAA channel characteristics from the perspectives of subarrays and sub-bands, including subarray-wise nonstationarities, consistencies, far-field approximations, and sub-band characteristics. Through a combination of analysis and measurement validation, several insights and benefits, particularly suitable for distributed processing in U6G ELAA systems, are revealed, which provides practical validation for the deployment of U6G ELAA systems.

eess.SP↗

arXiv2025

Vehicular Road Crack Detection with Deep Learning: A New Online Benchmark for Comprehensive Evaluation of Existing Algorithms

In the emerging field of urban digital twins (UDTs), advancing intelligent road inspection (IRI) vehicles with automatic road crack detection systems is essential for maintaining civil infrastructure. Over the past decade, deep learning-based road crack detection methods have been developed to detect cracks more efficiently, accurately, and objectively, with the goal of replacing manual visual inspection. Nonetheless, there is a lack of systematic reviews on state-of-the-art (SoTA) deep learning techniques, especially data-fusion and label-efficient algorithms for this task. This paper thoroughly reviews the SoTA deep learning-based algorithms, including (1) supervised, (2) unsupervised, (3) semi-supervised, and (4) weakly-supervised methods developed for road crack detection. Also, we create a dataset called UDTIRI-Crack, comprising $2,500$ high-quality images from seven public annotated sources, as the first extensive online benchmark in this field. Comprehensive experiments are conducted to compare the detection performance, computational efficiency, and generalizability of public SoTA deep learning-based algorithms for road crack detection. In addition, the feasibility of foundation models and large language models (LLMs) for road crack detection is explored. Afterwards, the existing challenges and future development trends of deep learning-based road crack detection algorithms are discussed. We believe this review can serve as practical guidance for developing intelligent road detection vehicles with the next-generation road condition assessment systems. The released benchmark UDTIRI-Crack is available at https://udtiri.com/submission/.

cs.CV↗

arXiv2025

Comment on "Attractor solutions in scalar-field cosmology" and "How many e-folds should we expect from high-scale inflation?"

In Ref. [1], it was claimed that in the spatially flat cosmological case there exists a unique conserved measure (up to normalization) on the $(ϕ,\dotϕ)$ phase space for scalar field with $m^2ϕ^2$ potential by finding a unique solution to the differential equation (44) (in Ref. [1]) in the low-energy regime. In Ref. [2], it was also claimed that a unique solution to the same differential equation was found in the high-energy regime and using this solution the authors calculated the expected total number of e-folds of inflation. In this comment, we reanalyze the differential equation (44) and obtain general solutions both in the low-energy and high-energy regime, which can include the solution in Ref. [1] and the solution in Ref. [2] as a special case in the corresponding energy regime. In this way, we find that following the constructions in Ref. [1] there actually exist infinitely many nonequivalent conserved measures for the scalar-field cosmology with $m^2ϕ^2$ potential on the $(ϕ,\dotϕ)$ phase space. Moreover, through specific calculations, we also show that different choices of measures can lead to quite different predictions of the expected total number of e-folds of inflation.

gr-qc↗

arXiv2025

Keypoint Detection Empowered Near-Field User Localization and Channel Reconstruction

In the near-field region of an extremely large-scale multiple-input multiple-output (XL MIMO) system, channel reconstruction is typically addressed through sparse parameter estimation based on compressed sensing (CS) algorithms after converting the received pilot signals into the transformed domain. However, the exhaustive search on the codebook in CS algorithms consumes significant computational resources and running time, particularly when a large number of antennas are equipped at the base station (BS). To overcome this challenge, we propose a novel scheme to replace the high-cost exhaustive search procedure. We visualize the sparse channel matrix in the transformed domain as a channel image and design the channel keypoint detection network (CKNet) to locate the user and scatterers in high speed. Subsequently, we use a small-scale newtonized orthogonal matching pursuit (NOMP) based refiner to further enhance the precision. Our method is applicable to both the Cartesian domain and the Polar domain. Additionally, to deal with scenarios with a flexible number of propagation paths, we further design FlexibleCKNet to predict both locations and confidence scores. Our experimental results validate that the CKNet and FlexibleCKNet-empowered channel reconstruction scheme can significantly reduce the computational complexity while maintaining high accuracy in both user and scatterer localization and channel reconstruction tasks.

eess.SP↗

arXiv2025

aiXcoder-7B: A Lightweight and Effective Large Language Model for Code Processing

Large Language Models (LLMs) have been widely used in code completion, and researchers are focusing on scaling up LLMs to improve their accuracy. However, larger LLMs have lower inference efficiency, affecting developers' experience and productivity. In this paper, we propose a lightweight and effective LLM for code completion named aiXcoder-7B. Compared to existing LLMs, aiXcoder-7B achieves higher code completion accuracy while having smaller scales (i.e., 7 billion parameters). We attribute the superiority of aiXcoder-7B to three key factors: (1) Multi-objective training. We employ three training objectives, one of which is our proposed Structured Fill-In-the-Middle (SFIM). SFIM considers the syntax structures in code and effectively improves the performance of LLMs for code. (2) Diverse data sampling strategies. They consider inter-file relationships and enhance the capability of LLMs in understanding cross-file contexts. (3) Extensive high-quality data. We establish a rigorous data collection pipeline and consume a total of 1.2 trillion unique tokens for training aiXcoder-7B. This vast volume of data enables aiXcoder-7B to learn a broad distribution of code. We evaluate aiXcoder-7B in five popular code completion benchmarks and a new benchmark collected by this paper. The results show that aiXcoder-7B outperforms the latest six LLMs with similar sizes and even surpasses four larger LLMs (e.g., StarCoder2-15B and CodeLlama-34B), positioning aiXcoder-7B as a lightweight and effective LLM for academia and industry. Finally, we summarize three valuable insights for helping practitioners train the next generations of LLMs for code. aiXcoder-7B has been open-souced and gained significant attention. Until January 2025, aiXcoder-7B has received 2,226 GitHub Stars.

cs.CL↗

arXiv2025

ViPOcc: Leveraging Visual Priors from Vision Foundation Models for Single-View 3D Occupancy Prediction

Inferring the 3D structure of a scene from a single image is an ill-posed and challenging problem in the field of vision-centric autonomous driving. Existing methods usually employ neural radiance fields to produce voxelized 3D occupancy, lacking instance-level semantic reasoning and temporal photometric consistency. In this paper, we propose ViPOcc, which leverages the visual priors from vision foundation models (VFMs) for fine-grained 3D occupancy prediction. Unlike previous works that solely employ volume rendering for RGB and depth image reconstruction, we introduce a metric depth estimation branch, in which an inverse depth alignment module is proposed to bridge the domain gap in depth distribution between VFM predictions and the ground truth. The recovered metric depth is then utilized in temporal photometric alignment and spatial geometric alignment to ensure accurate and consistent 3D occupancy prediction. Additionally, we also propose a semantic-guided non-overlapping Gaussian mixture sampler for efficient, instance-aware ray sampling, which addresses the redundant and imbalanced sampling issue that still exists in previous state-of-the-art methods. Extensive experiments demonstrate the superior performance of ViPOcc in both 3D occupancy prediction and depth estimation tasks on the KITTI-360 and KITTI Raw datasets. Our code is available at: \url{https://mias.group/ViPOcc}.

cs.CV↗

arXiv2024

Adaptive Training Meets Progressive Scaling: Elevating Efficiency in Diffusion Models

Diffusion models have demonstrated remarkable efficacy in various generative tasks with the predictive prowess of denoising model. Currently, diffusion models employ a uniform denoising model across all timesteps. However, the inherent variations in data distributions at different timesteps lead to conflicts during training, constraining the potential of diffusion models. To address this challenge, we propose a novel two-stage divide-and-conquer training strategy termed TDC Training. It groups timesteps based on task similarity and difficulty, assigning highly customized denoising models to each group, thereby enhancing the performance of diffusion models. While two-stage training avoids the need to train each model separately, the total training cost is even lower than training a single unified denoising model. Additionally, we introduce Proxy-based Pruning to further customize the denoising models. This method transforms the pruning problem of diffusion models into a multi-round decision-making problem, enabling precise pruning of diffusion models. Our experiments validate the effectiveness of TDC Training, demonstrating improvements in FID of 1.5 on ImageNet64 compared to original IDDPM, while saving about 20\% of computational resources.

cs.LG↗

arXiv2024

Transforming Medical Regulations into Numbers: Vectorizing a Decade of Medical Device Regulatory Shifts in the USA, EU, and China

Navigating the regulatory frameworks that ensure the safety and efficacy of medical devices can be challenging, especially across different regions. These frameworks often require redundant testing, slowing down the process of getting innovations to patients. This study leverages Natural Language Processing (NLP) to analyze 664 regulations and guidelines from the USA, EU, and China over the past decade, covering over 200 million tokens. We categorize regulations into key phases, such as animal studies, clinical trials, and other testing stages, and use Bidirectional Encoder Representations from Transformers (BERT) to perform Named Entity Recognition (NER), identifying key regulatory terms and entities. By converting these texts into numerical representations and segmenting them by phase, country, and year, we compare jurisdictional requirements and assess their alignment. Additionally, we apply Latent Dirichlet Allocation (LDA) for theme analysis to observe changes in regulatory focus over time, reflecting evolving priorities and challenges. Our analysis reveals notable semantic similarities and differences between countries and phases. For instance, the closest alignment in animal study regulations is between China and the USA, with a mean cosine distance of 0.33. These findings highlight the computational potential in regulatory science, offering valuable insights for researchers, policymakers, and industry professionals.

cs.CY↗