arXiv subjects

Feng Liu

Publications and source records attributed to Feng Liu.

At least 217 records · Page 12Linked to original sources

arXiv2025

RadarMP: Motion Perception for 4D mmWave Radar in Autonomous Driving

Accurate 3D scene motion perception significantly enhances the safety and reliability of an autonomous driving system. Benefiting from its all-weather operational capability and unique perceptual properties, 4D mmWave radar has emerged as an essential component in advanced autonomous driving. However, sparse and noisy radar points often lead to imprecise motion perception, leaving autonomous vehicles with limited sensing capabilities when optical sensors degrade under adverse weather conditions. In this paper, we propose RadarMP, a novel method for precise 3D scene motion perception using low-level radar echo signals from two consecutive frames. Unlike existing methods that separate radar target detection and motion estimation, RadarMP jointly models both tasks in a unified architecture, enabling consistent radar point cloud generation and pointwise 3D scene flow prediction. Tailored to radar characteristics, we design specialized self-supervised loss functions guided by Doppler shifts and echo intensity, effectively supervising spatial and motion consistency without explicit annotations. Extensive experiments on the public dataset demonstrate that RadarMP achieves reliable motion perception across diverse weather and illumination conditions, outperforming radar-based decoupled motion perception pipelines and enhancing perception capabilities for full-scenario autonomous driving systems.

cs.CV↗

arXiv2025

CURENet: Combining Unified Representations for Efficient Chronic Disease Prediction

Electronic health records (EHRs) are designed to synthesize diverse data types, including unstructured clinical notes, structured lab tests, and time-series visit data. Physicians draw on these multimodal and temporal sources of EHR data to form a comprehensive view of a patient's health, which is crucial for informed therapeutic decision-making. Yet, most predictive models fail to fully capture the interactions, redundancies, and temporal patterns across multiple data modalities, often focusing on a single data type or overlooking these complexities. In this paper, we present CURENet, a multimodal model (Combining Unified Representations for Efficient chronic disease prediction) that integrates unstructured clinical notes, lab tests, and patients' time-series data by utilizing large language models (LLMs) for clinical text processing and textual lab tests, as well as transformer encoders for longitudinal sequential visits. CURENet has been capable of capturing the intricate interaction between different forms of clinical data and creating a more reliable predictive model for chronic illnesses. We evaluated CURENet using the public MIMIC-III and private FEMH datasets, where it achieved over 94\% accuracy in predicting the top 10 chronic conditions in a multi-label framework. Our findings highlight the potential of multimodal EHR integration to enhance clinical decision-making and improve patient outcomes.

cs.AI↗

arXiv2025

Design of a Six-band, 2.4-Octave (80--420 GHz) Hierarchically Summed Phased-Array Slot-Dipole Antenna Array for NEW-MUSIC

The Next-generation Extended Wavelength Multi-band Sub/millimeter Inductance Camera (NEW-MUSIC), located on the Leighton Chajnantor Telescope (LCT), will be the first six-band trans-millimeter wave polarimeter. This paper proposes a broadband, hierarchical phased-array antenna with integrated band-defining filters necessary to realize NEW-MUSIC. It covers a spectral bandwidth of 2.4 octaves from 80~GHz to 420~GHz, a frequency range ideal for studying trans-millimeter emission from a range of time-domain sources, using the Sunyaev-Zeldovich effects to study hot plasmas in galaxy clusters and galaxies, and to observe dusty sources, from star-forming regions in our galaxy to high-redshift dusty, star-forming galaxies. To achieve these goals, three groups of superconducting lumped-element on-chip low-pass/band-pass filter-banks were designed to hierarchically sum the superconducting, broadband, non-resonant, slot-dipole antenna arrays and band-pass filter the trans-mm light before outputting it on microstripline to detectors (KIDs in the case of NEW-MUSIC).

astro-ph.IM↗

arXiv2025

Enhancing Multimodal Medical Image Classification using Cross-Graph Modal Contrastive Learning

The classification of medical images is a pivotal aspect of disease diagnosis, often enhanced by deep learning techniques. However, traditional approaches typically focus on unimodal medical image data, neglecting the integration of diverse non-image patient data. This paper proposes a novel Cross-Graph Modal Contrastive Learning (CGMCL) framework for multimodal structured data from different data domains to improve medical image classification. The model effectively integrates both image and non-image data by constructing cross-modality graphs and leveraging contrastive learning to align multimodal features in a shared latent space. An inter-modality feature scaling module further optimizes the representation learning process by reducing the gap between heterogeneous modalities. The proposed approach is evaluated on two datasets: a Parkinson's disease (PD) dataset and a public melanoma dataset. Results demonstrate that CGMCL outperforms conventional unimodal methods in accuracy, interpretability, and early disease prediction. Additionally, the method shows superior performance in multi-class melanoma classification. The CGMCL framework provides valuable insights into medical image classification while offering improved disease interpretability and predictive capabilities.

eess.IV↗

arXiv2025

Structural Priors and Modular Adapters in the Composable Fine-Tuning Algorithm of Large-Scale Models

This paper proposes a composable fine-tuning method that integrates graph structural priors with modular adapters to address the high computational cost and structural instability faced by large-scale pre-trained models in multi-task adaptation. The method introduces a relation matrix to model dependencies among tasks, explicitly encoding correlations between nodes and paths into graph structural priors, which provide unified structural constraints for adapter weight allocation and path selection. Modular adapters are embedded into different layers through low-rank mapping and a pluggable mechanism, enabling efficient cross-task composition and reuse under prior guidance. This mechanism not only improves parameter efficiency and training stability but also alleviates path conflicts and redundant computation in multi-task scenarios. Furthermore, experiments on hyperparameter sensitivity, environmental sensitivity, and data sensitivity are conducted to systematically analyze key factors such as routing temperature, gating thresholds, and relation matrix regularization strength, verifying the consistency and superior performance of the method under structural constraints. The results demonstrate that the proposed framework significantly enhances task prediction accuracy, adapter weight allocation precision, and overall computational efficiency while maintaining model lightweight design, highlighting the synergistic advantages of graph priors and modular mechanisms in composable fine-tuning.

cs.LG↗

arXiv2025

Enhancing Adversarial Transferability by Balancing Exploration and Exploitation with Gradient-Guided Sampling

Adversarial attacks present a critical challenge to deep neural networks' robustness, particularly in transfer scenarios across different model architectures. However, the transferability of adversarial attacks faces a fundamental dilemma between Exploitation (maximizing attack potency) and Exploration (enhancing cross-model generalization). Traditional momentum-based methods over-prioritize Exploitation, i.e., higher loss maxima for attack potency but weakened generalization (narrow loss surface). Conversely, recent methods with inner-iteration sampling over-prioritize Exploration, i.e., flatter loss surfaces for cross-model generalization but weakened attack potency (suboptimal local maxima). To resolve this dilemma, we propose a simple yet effective Gradient-Guided Sampling (GGS), which harmonizes both objectives through guiding sampling along the gradient ascent direction to improve both sampling efficiency and stability. Specifically, based on MI-FGSM, GGS introduces inner-iteration random sampling and guides the sampling direction using the gradient from the previous inner-iteration (the sampling's magnitude is determined by a random distribution). This mechanism encourages adversarial examples to reside in balanced regions with both flatness for cross-model generalization and higher local maxima for strong attack potency. Comprehensive experiments across multiple DNN architectures and multimodal large language models (MLLMs) demonstrate the superiority of our method over state-of-the-art transfer attacks. Code is made available at https://github.com/anuin-cat/GGS.

cs.LG↗

arXiv2025

M^3Detection: Multi-Frame Multi-Level Feature Fusion for Multi-Modal 3D Object Detection with Camera and 4D Imaging Radar

Recent advances in 4D imaging radar have enabled robust perception in adverse weather, while camera sensors provide dense semantic information. Fusing the these complementary modalities has great potential for cost-effective 3D perception. However, most existing camera-radar fusion methods are limited to single-frame inputs, capturing only a partial view of the scene. The incomplete scene information, compounded by image degradation and 4D radar sparsity, hinders overall detection performance. In contrast, multi-frame fusion offers richer spatiotemporal information but faces two challenges: achieving robust and effective object feature fusion across frames and modalities, and mitigating the computational cost of redundant feature extraction. Consequently, we propose M^3Detection, a unified multi-frame 3D object detection framework that performs multi-level feature fusion on multi-modal data from camera and 4D imaging radar. Our framework leverages intermediate features from the baseline detector and employs the tracker to produce reference trajectories, improving computational efficiency and providing richer information for second-stage. In the second stage, we design a global-level inter-object feature aggregation module guided by radar information to align global features across candidate proposals and a local-level inter-grid feature aggregation module that expands local features along the reference trajectories to enhance fine-grained object representation. The aggregated features are then processed by a trajectory-level multi-frame spatiotemporal reasoning module to encode cross-frame interactions and enhance temporal representation. Extensive experiments on the VoD and TJ4DRadSet datasets demonstrate that M^3Detection achieves state-of-the-art 3D detection performance, validating its effectiveness in multi-frame detection with camera-4D imaging radar fusion.

cs.CV↗

arXiv2025

Highly Undersampled MRI Reconstruction via a Single Posterior Sampling of Diffusion Models

Incoherent k-space undersampling and deep learning-based reconstruction methods have shown great success in accelerating MRI. However, the performance of most previous methods will degrade dramatically under high acceleration factors, e.g., 8$\times$ or higher. Recently, denoising diffusion models (DM) have demonstrated promising results in solving this issue; however, one major drawback of the DM methods is the long inference time due to a dramatic number of iterative reverse posterior sampling steps. In this work, a Single Step Diffusion Model-based reconstruction framework, namely SSDM-MRI, is proposed for restoring MRI images from highly undersampled k-space. The proposed method achieves one-step reconstruction by first training a conditional DM and then iteratively distilling this model four times using an iterative selective distillation algorithm, which works synergistically with a shortcut reverse sampling strategy for model inference. Comprehensive experiments were carried out on both publicly available fastMRI brain and knee images, as well as an in-house multi-echo GRE (QSM) subject. Overall, the results showed that SSDM-MRI outperformed other methods in terms of numerical metrics (e.g., PSNR and SSIM), error maps, image fine details, and latent susceptibility information hidden in MRI phase images. In addition, the reconstruction time for a 320$\times$320 brain slice of SSDM-MRI is only 0.45 second, which is only comparable to that of a simple U-net, making it a highly effective solution for MRI reconstruction tasks.

eess.IV↗

arXiv2025

Revealing Multimodal Causality with Large Language Models

Uncovering cause-and-effect mechanisms from data is fundamental to scientific progress. While large language models (LLMs) show promise for enhancing causal discovery (CD) from unstructured data, their application to the increasingly prevalent multimodal setting remains a critical challenge. Even with the advent of multimodal LLMs (MLLMs), their efficacy in multimodal CD is hindered by two primary limitations: (1) difficulty in exploring intra- and inter-modal interactions for comprehensive causal variable identification; and (2) insufficiency to handle structural ambiguities with purely observational data. To address these challenges, we propose MLLM-CD, a novel framework for multimodal causal discovery from unstructured data. It consists of three key components: (1) a novel contrastive factor discovery module to identify genuine multimodal factors based on the interactions explored from contrastive sample pairs; (2) a statistical causal structure discovery module to infer causal relationships among discovered factors; and (3) an iterative multimodal counterfactual reasoning module to refine the discovery outcomes iteratively by incorporating the world knowledge and reasoning capabilities of MLLMs. Extensive experiments on both synthetic and real-world datasets demonstrate the effectiveness of the proposed MLLM-CD in revealing genuine factors and causal relationships among them from multimodal unstructured data.

cs.LG↗

arXiv2025

Coherently Enhanced Axion-Photon Conversion via Seeded Photons for Short-Pulse Axion Detection

We propose a seeded axion-photon conversion scheme to enhance the sensitivity of light-shining-through-a-wall (LSW) experiments for axion detection, where the axions are generated from short pulse lasers and the usual resonant cavity is not applicable. By injecting a weak, coherent seed electromagnetic (EM) field into the axion-photon conversion region, the axion-induced EM field can constructively interfere with the seed field, amplifying the number of regenerated photons to a level exceeding that of the unseeded scenario. We evaluate the expected signal enhancement, statistical limits from Poisson counting with seed fluctuations and background, and the potential improvement in coupling sensitivity. Compared to a standard LSW setup, the seeded scheme can achieve orders-of-magnitude higher photon yield per axion, potentially surpassing resonance-enhanced experiments in certain parameter regimes. This approach presents a promising pathway to extend the reach of laboratory axion searches, particularly in scenarios where the resonant cavities are impractical.

hep-ph↗

arXiv2025

Single mode lasing and spectral narrowing in photonic crystal line-defect cavities via spatially selected Bloch modes

The demand for high-efficiency and miniaturized on-chip light sources drives continuous innovation in photonic crystal (PhC) microcavity lasers. The presence of slow-light effects in PhC microcavities leads to the mode competition between Bloch modes resulting in multi-mode lasing, which obstructs the dense integration of PhC lasers. Here, we theoretically verify a technical scheme for the single-mode lasing of PhC line-defect-cavity lasers by spatially pumping a certain Bloch mode via optical interference.We demonstrate the capability to select a specific longitudinal mode to lase with a side mode suppression ratio (SMSR) exceeding 30 dB. The interaction between optical interference fringes and the vacuum electromagnetic field inside the PhC cavity improves the linewidth and noise characteristics of lasers. This scheme of Bloch mode selection provides a novel and viable tool for the manipulation of PhC microcavity lasers.

physics.optics↗

arXiv2025

Polymorphic self-poisoning in poly(lactic acid): a new phenomenon in polymer crystallization

Self-poisoning (SP) is ubiquitous in polymer crystallization, but has so far manifested itself visibly only as minima in growth rate vs. temperature in either monodisperse systems where e.g. unstable folded chains obstruct crystallization of stable extended chains, or in periodically segmented chains where unstable stems with n-1 segments disturb deposition of stable stems with n segments. Here we report a new type of self-poisoning found in poly(lactic acid) (PLA), where a less stable crystal form (alpha') disturbs growth of the stable form (alpha). While alpha requires strict up-down order of the polar chains, alpha' does not, hence is kinetically favoured. Unexpectedly, below the temperature of the growth rate minimum the lamellar thickness increases rather than drops, as in all other reported cases of polymer crystallization with decreasing temperature. A growth rate equation model is developed, giving good match with experiments, but revealing an unexpectedly low fold surface free energy of alpha' form. Due to SP of alpha, most practical fast-cooling processing gives the low-modulus alpha'-form grown close to Tg, explaining generally poor mechanical properties of the bio-friendly PLA.

cond-mat.soft↗

arXiv2025

SFGFusion: Surface Fitting Guided 3D Object Detection with 4D Radar and Camera Fusion

3D object detection is essential for autonomous driving. As an emerging sensor, 4D imaging radar offers advantages as low cost, long-range detection, and accurate velocity measurement, making it highly suitable for object detection. However, its sparse point clouds and low resolution limit object geometric representation and hinder multi-modal fusion. In this study, we introduce SFGFusion, a novel camera-4D imaging radar detection network guided by surface fitting. By estimating quadratic surface parameters of objects from image and radar data, the explicit surface fitting model enhances spatial representation and cross-modal interaction, enabling more reliable prediction of fine-grained dense depth. The predicted depth serves two purposes: 1) in an image branch to guide the transformation of image features from perspective view (PV) to a unified bird's-eye view (BEV) for multi-modal fusion, improving spatial mapping accuracy; and 2) in a surface pseudo-point branch to generate dense pseudo-point cloud, mitigating the radar point sparsity. The original radar point cloud is also encoded in a separate radar branch. These two point cloud branches adopt a pillar-based method and subsequently transform the features into the BEV space. Finally, a standard 2D backbone and detection head are used to predict object labels and bounding boxes from BEV features. Experimental results show that SFGFusion effectively fuses camera and 4D radar features, achieving superior performance on the TJ4DRadSet and view-of-delft (VoD) object detection benchmarks.

cs.CV↗

arXiv2025

Neural Network Reprogrammability: A Unified Theme on Model Reprogramming, Prompt Tuning, and Prompt Instruction

As large-scale pre-trained foundation models continue to expand in size and capability, efficiently adapting them to specific downstream tasks has become increasingly critical. Despite substantial progress, existing adaptation approaches have evolved largely in isolation, without a clear understanding of their interrelationships. This survey introduces neural network reprogrammability as a unifying framework that bridges mainstream model adaptation techniques--model reprogramming, prompt tuning, and prompt instruction--previously fragmented research areas yet converges on a shared principle: repurposing a pre-trained model by manipulating information at the interfaces while keeping the model parameters frozen. These methods exploit neural networks' sensitivity to manipulation on different interfaces, be it through perturbing inputs, inserting tokens into intermediate layers, or providing task-specific examples in context, to redirect model behaviors towards desired outcomes. We then present a taxonomy that categorizes such information manipulation-based adaptation approaches across four key dimensions: manipulation format (fixed or learnable), location (interfaces where manipulations occur), operator (how they are applied), and output alignment requirement (post-processing needed to align outputs with downstream tasks). Notably, this framework applies consistently across data modalities, independent of specific model architectures. Moreover, viewing established techniques like in-context learning and chain-of-thought prompting through this lens reveals both their theoretical connections and practical distinctions. We further analyze remaining technical challenges and ethical considerations, positioning neural network reprogrammability as a fundamental paradigm for efficient model adaptation. We lastly identify promising research directions emerging from this integrative viewpoint.

cs.LG↗

arXiv2025

Structural Generalization for Microservice Routing Using Graph Neural Networks

This paper focuses on intelligent routing in microservice systems and proposes an end-to-end optimization framework based on graph neural networks. The goal is to improve routing decision efficiency and overall system performance under complex topologies. The method models invocation relationships among microservices as a graph. In this graph, service nodes and communication links are treated as graph nodes and edges. Multi-dimensional features such as node states, link latency, and call frequency are used as input. A multi-layer graph neural network is employed to perform high-order information aggregation and structural modeling. The model outputs a score for each candidate service path. These scores are then used to guide dynamic routing decisions. To improve the model's ability to assess path quality, an edge-aware attention mechanism is introduced. This mechanism helps the model capture instability and bottleneck risks in service communications more accurately. The paper also conducts a systematic analysis of the model's performance under different network depths, topology densities, and service scales. It evaluates the effectiveness of the method in terms of routing accuracy, prediction error, and system stability. Experimental results show that the proposed method outperforms existing mainstream strategies across multiple key metrics. It handles highly dynamic and concurrent microservice environments effectively and demonstrates strong performance, robustness, and structural generalization.

cs.NI↗

arXiv2025

Spatiotemporal Traffic Prediction in Distributed Backend Systems via Graph Neural Networks

This paper addresses the problem of traffic prediction in distributed backend systems and proposes a graph neural network based modeling approach to overcome the limitations of traditional models in capturing complex dependencies and dynamic features. The system is abstracted as a graph with nodes and edges, where node features represent traffic and resource states, and adjacency relations describe service interactions. A graph convolution mechanism enables multi order propagation and aggregation of node features, while a gated recurrent structure models historical sequences dynamically, thus integrating spatial structures with temporal evolution. A spatiotemporal joint modeling module further fuses graph representation with temporal dependency, and a decoder generates future traffic predictions. The model is trained with mean squared error to minimize deviations from actual values. Experiments based on public distributed system logs construct combined inputs of node features, topology, and sequences, and compare the proposed method with mainstream baselines using MSE, RMSE, MAE, and MAPE. Results show that the proposed method achieves stable performance and low error across different prediction horizons and model depths, significantly improving the accuracy and robustness of traffic forecasting in distributed backend systems and verifying the potential of graph neural networks in complex system modeling.

cs.DC↗

arXiv2025

An Iterative Problem-Driven Scenario Reduction Framework for Stochastic Optimization with Conditional Value-at-Risk

Scenario reduction (SR) alleviates the computational complexity of scenario-based stochastic optimization with conditional value-at-risk (SBSO-CVaR) by identifying representative scenarios to depict the underlying uncertainty and tail risks. Existing distribution-driven SR methods emphasize statistical similarity but often exclude extreme scenarios, leading to weak tail-risk awareness and insufficient problem-specific representativeness. Instead, this paper proposes an iterative problem-driven scenario reduction framework. Specifically, we integrate the SBSO-CVaR problem structure into SR process and project the original scenario set from the distribution space onto the problem space. Subsequently, to minimize the SR optimality gap with acceptable computation complexity, we propose a tractable iterative problem-driven scenario reduction (IPDSR) method that selects representative scenarios that best approximate the optimality distribution of the original scenario set while preserving tail risks. Furthermore, the iteration process is rendered as a mixed-integer program to enable scenario partitioning and representative scenarios selection. And ex-post problem-driven evaluation indices are proposed to evaluate the SR performance. Numerical experiments show IPDSR significantly outperforms existing SR methods by achieving an optimality gap of less than 1% within an acceptable computation time.

math.OC↗

arXiv2025

HatLLM: Hierarchical Attention Masking for Enhanced Collaborative Modeling in LLM-based Recommendation

Recent years have witnessed a surge of research on leveraging large language models (LLMs) for sequential recommendation. LLMs have demonstrated remarkable potential in inferring users' nuanced preferences through fine-grained semantic reasoning. However, they also exhibit a notable limitation in effectively modeling collaborative signals, i.e., behavioral correlations inherent in users' historical interactions. Our empirical analysis further reveals that the attention mechanisms in LLMs tend to disproportionately focus on tokens within the same item, thereby impeding the capture of cross-item correlations. To address this limitation, we propose a novel hierarchical attention masking strategy for LLM-based recommendation, termed HatLLM. Specifically, in shallow layers, HatLLM masks attention between tokens from different items, facilitating intra-item semantic understanding; in contrast, in deep layers, HatLLM masks attention within items, thereby compelling the model to capture cross-item correlations. This progressive, layer-wise approach enables LLMs to jointly model both token-level and item-level dependencies. Extensive experiments on three real-world datasets demonstrate that HatLLM achieves significant performance gains (9.13% on average) over existing LLM-based methods.

cs.IR↗