arXiv subjects

Liang Li

Publications and source records attributed to Liang Li.

At least 109 records · Page 6Linked to original sources

arXiv2025

Hazy Pedestrian Trajectory Prediction via Physical Priors and Graph-Mamba

To address the issues of physical information degradation and ineffective pedestrian interaction modeling in pedestrian trajectory prediction under hazy weather conditions, we propose a deep learning model that combines physical priors of atmospheric scattering with topological modeling of pedestrian relationships. Specifically, we first construct a differentiable atmospheric scattering model that decouples haze concentration from light degradation through a network with physical parameter estimation, enabling the learning of haze-mitigated feature representations. Second, we design an adaptive scanning state space model for feature extraction. Our adaptive Mamba variant achieves a 78% inference speed increase over native Mamba while preserving long-range dependency modeling. Finally, to efficiently model pedestrian relationships, we develop a heterogeneous graph attention network, using graph matrices to model multi-granularity interactions between pedestrians and groups, combined with a spatio-temporal fusion module to capture the collaborative evolution patterns of pedestrian movements. Furthermore, we constructed a new pedestrian trajectory prediction dataset based on ETH/UCY to evaluate the effectiveness of the proposed method. Experiments show that our method reduces the minADE / minFDE metrics by 37.2% and 41.5%, respectively, compared to the SOTA models in dense haze scenarios (visibility < 30m), providing a new modeling paradigm for reliable perception in intelligent transportation systems in adverse environments.

cs.CV↗

arXiv2025

MEF: A Systematic Evaluation Framework for Text-to-Image Models

Rapid advances in text-to-image (T2I) generation have raised higher requirements for evaluation methodologies. Existing benchmarks center on objective capabilities and dimensions, but lack an application-scenario perspective, limiting external validity. Moreover, current evaluations typically rely on either ELO for overall ranking or MOS for dimension-specific scoring, yet both methods have inherent shortcomings and limited interpretability. Therefore, we introduce the Magic Evaluation Framework (MEF), a systematic and practical approach for evaluating T2I models. First, we propose a structured taxonomy encompassing user scenarios, elements, element compositions, and text expression forms to construct the Magic-Bench-377, which supports label-level assessment and ensures a balanced coverage of both user scenarios and capabilities. On this basis, we combine ELO and dimension-specific MOS to generate model rankings and fine-grained assessments respectively. This joint evaluation method further enables us to quantitatively analyze the contribution of each dimension to user satisfaction using multivariate logistic regression. By applying MEF to current T2I models, we obtain a leaderboard and key characteristics of the leading models. We release our evaluation framework and make Magic-Bench-377 fully open-source to advance research in the evaluation of visual generative models.

cs.AI↗

arXiv2025

On traveling wave solutions of water-wave equations in curved annular domains

This paper presents a pioneering investigation into the existence of traveling wave solutions for the two-dimensional Euler equations with constant vorticity in a curved annular domain, where gravity acts radially inward. This configuration is highly relevant to astrophysical and equatorial oceanic flows, such as those found in planetary rings and equatorial currents. Unlike traditional water wave models that assume flat beds and vertical gravity, our study more accurately captures the centripetal effects and boundary-driven vorticity inherent in these complex systems. Our main results establish both local and global bifurcation of traveling waves, marking a significant advancement in the field. First, through a local bifurcation analysis near a trivial solution, we identify a critical parameter $α_c$ and prove the existence of a smooth branch of small-amplitude solutions. The bifurcation is shown to be pitchfork-type, with its direction (subcritical or supercritical) determined by the sign of an explicit parameter $\mathcal{O}$. This finding provides a nuanced understanding of the wave behavior under varying conditions. Second, we obtain a global bifurcation result using a modified Leray-Schauder degree theory. This result demonstrates that the local branch extends to large-amplitude waves. This comprehensive analysis offers a holistic view of the traveling wave solutions in this complex domain. Finally, numerical examples illustrate the theoretical bifurcation types, confirming both supercritical and subcritical regimes. These examples highlight the practical applicability of our results.

math.AP↗

arXiv2025

RewardDance: Reward Scaling in Visual Generation

Reward Models (RMs) are critical for improving generation models via Reinforcement Learning (RL), yet the RM scaling paradigm in visual generation remains largely unexplored. It primarily due to fundamental limitations in existing approaches: CLIP-based RMs suffer from architectural and input modality constraints, while prevalent Bradley-Terry losses are fundamentally misaligned with the next-token prediction mechanism of Vision-Language Models (VLMs), hindering effective scaling. More critically, the RLHF optimization process is plagued by Reward Hacking issue, where models exploit flaws in the reward signal without improving true quality. To address these challenges, we introduce RewardDance, a scalable reward modeling framework that overcomes these barriers through a novel generative reward paradigm. By reformulating the reward score as the model's probability of predicting a "yes" token, indicating that the generated image outperforms a reference image according to specific criteria, RewardDance intrinsically aligns reward objectives with VLM architectures. This alignment unlocks scaling across two dimensions: (1) Model Scaling: Systematic scaling of RMs up to 26 billion parameters; (2) Context Scaling: Integration of task-specific instructions, reference examples, and chain-of-thought (CoT) reasoning. Extensive experiments demonstrate that RewardDance significantly surpasses state-of-the-art methods in text-to-image, text-to-video, and image-to-video generation. Crucially, we resolve the persistent challenge of "reward hacking": Our large-scale RMs exhibit and maintain high reward variance during RL fine-tuning, proving their resistance to hacking and ability to produce diverse, high-quality outputs. It greatly relieves the mode collapse problem that plagues smaller models.

cs.CV↗

arXiv2025

Is Artificial Intelligence Reshaping the Landscape of the International Academic Community of Geosciences?

Through bibliometric analysis and topic modeling, we find that artificial intelligence (AI) is positively transforming geosciences research, with a notable increase in AI-related scientific output in recent years. We are encouraged to observe that earth scientists from developing countries have gained better visibility in the recent AI for Science (AI4S) paradigm and that AI is also improving the landscape of international collaboration in geoscience-related research.

cs.DL↗

arXiv2025

FlowDubber: Movie Dubbing with LLM-based Semantic-aware Learning and Flow Matching based Voice Enhancing

Movie Dubbing aims to convert scripts into speeches that align with the given movie clip in both temporal and emotional aspects while preserving the vocal timbre of a given brief reference audio. Existing methods focus primarily on reducing the word error rate while ignoring the importance of lip-sync and acoustic quality. To address these issues, we propose a large language model (LLM) based flow matching architecture for dubbing, named FlowDubber, which achieves high-quality audio-visual sync and pronunciation by incorporating a large speech language model and dual contrastive aligning while achieving better acoustic quality via the proposed voice-enhanced flow matching than previous works. First, we introduce Qwen2.5 as the backbone of LLM to learn the in-context sequence from movie scripts and reference audio. Then, the proposed semantic-aware learning focuses on capturing LLM semantic knowledge at the phoneme level. Next, dual contrastive aligning (DCA) boosts mutual alignment with lip movement, reducing ambiguities where similar phonemes might be confused. Finally, the proposed Flow-based Voice Enhancing (FVE) improves acoustic quality in two aspects, which introduces an LLM-based acoustics flow matching guidance to strengthen clarity and uses affine style prior to enhance identity when recovering noise into mel-spectrograms via gradient vector field prediction. Extensive experiments demonstrate that our method outperforms several state-of-the-art methods on two primary benchmarks.

cs.MM↗

arXiv2025

Spectral-Prior Guided Multistage Physics-Informed Neural Networks for Highly Accurate PDE Solutions

Physics-Informed Neural Networks (PINNs) are becoming a popular method for solving PDEs, due to their mesh-free nature and their ability to handle high-dimensional problems where traditional numerical solvers often struggle. Despite their promise, the practical application of PINNs is still constrained by several fac- tors, a primary one being their often-limited accuracy. This paper is dedicated to enhancing the accuracy of PINNs by introducing spectral-prior guided multistage strategy. We propose two methods: Spectrum- Informed Multistage Physics-Informed Neural Networks (SI-MSPINNs) and Multistage Physics-Informed Neural Networks with Spectrum Weighted Random Fourier Features (RFF-MSPINNs). The SI-MSPINNs integrate the core mechanism of Spectrum-Informed Multistage Neural Network (SI-MSNNs) and PINNs, in which we extract the Dominant Spectral Pattern (DSP) of residuals by the discrete Fourier transform. This DSP guides the network initialization to alleviate spectral bias, and gradually optimizes the resolution accuracy using a multistage strategy. The RFF-MSPINNs combines random Fourier features with spectral weighting methods, dynamically adjusting the frequency sampling distribution based on the residual power spectral density, allowing the network to prioritize learning high-energy physical modes. Through experimental verification of the Burgers equation and the Helmholtz equation, we show that both models significantly improve the accuracy of the original PINNs.

math.NA↗

arXiv2025

Instantaneous optical selection rule for independent control of valley currents

We reveal an instantaneous optical valley selection rule that illuminates the coupling between the instantaneous optical chirality of the driving laser field and the chirality of valley systems. Building on this principle, we propose and demonstrate that a single chirality-separated optical field, in which oppositely signed instantaneous optical chiralities are separated within an optical cycle, enables independent manipulation of currents from K and K' valleys. Based on this scheme, we highlight two key example applications: (1) complete separation of currents from different valleys, yielding 100%-purity valley-polarized currents, and (2) generation of pure valley current with zero net charge flow. Our work offers a robust and highly controllable all-optical strategy for ultrafast engineering valley currents in the optical cycle timescale, paving a new avenue for valleytronics and quantum information technologies.

physics.optics↗

arXiv2025

PEdger++: Practical Edge Detection via Assembling Cross Information

Edge detection serves as a critical foundation for numerous computer vision applications, including object detection, semantic segmentation, and image editing, by extracting essential structural cues that define object boundaries and salient edges. To be viable for broad deployment across devices with varying computational capacities, edge detectors shall balance high accuracy with low computational complexity. While deep learning has evidently improved accuracy, they often suffer from high computational costs, limiting their applicability on resource-constrained devices. This paper addresses the challenge of achieving that balance: \textit{i.e.}, {how to efficiently capture discriminative features without relying on large-size and sophisticated models}. We propose PEdger++, a collaborative learning framework designed to reduce computational costs and model sizes while improving edge detection accuracy. The core principle of our PEdger++ is that cross-information derived from heterogeneous architectures, diverse training moments, and multiple parameter samplings, is beneficial to enhance learning from an ensemble perspective. Extensive experimental results on the BSDS500, NYUD and Multicue datasets demonstrate the effectiveness of our approach, both quantitatively and qualitatively, showing clear improvements over existing methods. We also provide multiple versions of the model with varying computational requirements, highlighting PEdger++'s adaptability with respect to different resource constraints. Codes are accessible at https://github.com/ForawardStar/EdgeDetectionviaPEdgerPlus/.

cs.CV↗

arXiv2025

Efficient Inference for Large Reasoning Models: A Survey

Large Reasoning Models (LRMs) significantly improve the reasoning ability of Large Language Models (LLMs) by learning to reason, exhibiting promising performance in solving complex tasks. However, their deliberative reasoning process leads to inefficiencies in token usage, memory consumption, and inference time. Thus, this survey provides a review of efficient inference methods designed specifically for LRMs, focusing on mitigating token inefficiency while preserving the reasoning quality. The overview structure of this paper is shown in Figure~\ref{fig:paper_structure}. First, we introduce a taxonomy to group the recent methods into two main categories: (a) explicit compact Chain-of-Thought (CoT), which reduces tokens while keeping the explicit reasoning structure, and (b) implicit latent CoT, which encodes reasoning steps within hidden representations instead of explicit tokens. Meanwhile, we discuss their strengths and weaknesses. Then, we conduct empirical analyses on existing methods from reasoning scenarios, object functions, and performance \& efficiency aspects. Besides, we present open challenges in this field, including human-centric controllable reasoning, trade-off between interpretability and efficiency of reasoning, ensuring the safety of efficient reasoning, and broader applications of efficient reasoning. In addition, we highlight key insights for enhancing LRMs' inference efficiency via techniques such as model merging, new architectures, and agent routers. We hope this work serves as a valuable guide, helping researchers overcome challenges in this vibrant field. A collection of efficient reasoning methods for LRMs (papers and codes) is provided at this link: https://github.com/yueliu1999/Awesome-Efficient-Inference-for-LRMs.

cs.CL↗

arXiv2025

Stability and bifurcation of Navier-Stokes equations in an annular domain with mixed boundary conditions

We study the existence and stability of non-trivial steady-state solutions to the two-dimensional incompressible Navier-Stokes equations in an annular domain $Ω= B(0,b) \setminus \overline{B(0,a)}$ with radii $b>a>0$.The outer boundary $\partial B(0, b)$ is subject to the free condition, while the inner boundary $\partial B(0, a)$ obeys a Navier-slip condition with effective slip length $α> 0$. Our main results are fourfold. First, we establish global-in-time strong solutions and derive a sharp energy estimate that underpins the subsequent nonlinear instability analysis. Second, for $α> 0$, we compute an explicit critical viscosity $μ_c:=μ_c(α, a,b,μ)$ that separates qualitatively different dynamical behaviours. Third, we precisely characterize the stability properties of the trivial solution in three distinct regimes. The zero solution is globally asymptotically stable in $H^2$ if $μ> μ_c$. If $μ= μ_c$, we prove an alternative theorem that completely describes the local dynamics near the trivial state. If $μ< μ_c$, the trivial solution is nonlinearly unstable in every $L^p (p \geq 1)$.Finally, we demonstrate that for $μ< μ_c$, the system undergoes a pitchfork bifurcation that generates an infinite family of non-trivial steady states. For generic choices of $(α,b)$, this bifurcation is supercritical; a measurable subset of parameter space yields subcritical transitions. Notably, all bifurcating solutions share the same topological pattern-a single row of counter-rotating vortices-despite their mathematical non-uniqueness.

math.AP↗

arXiv2025

PAE MobiLLM: Privacy-Aware and Efficient LLM Fine-Tuning on the Mobile Device via Additive Side-Tuning

There is a huge gap between numerous intriguing applications fostered by on-device large language model (LLM) fine-tuning (FT) from fresh mobile data and the limited resources of a mobile device. While existing server-assisted methods (e.g., split learning or side-tuning) may enable LLM FT on the local mobile device, they suffer from heavy communication burdens of activation transmissions, and may disclose data and labels to the server. To address those issues, we develop PAE MobiLLM, a a privacy-aware and efficient LLM FT method which can be deployed on the mobile device via server-assisted additive side-tuning. To further accelerate FT convergence and improve computing efficiency, PAE MobiLLM integrates activation caching on the server side, which allows the server to reuse historical activations and saves the mobile device from repeatedly computing forward passes for the recurring data samples. Besides, to reduce communication cost, PAE MobiLLM develops an activation shortcut that transmits only the token involved in the loss calculation instead of full activation matrices to guide the side network tuning. Last but not least, PAE MobiLLM introduces the additive adapter side-network design which makes the server train the adapter modules based on device-defined prediction differences rather than raw ground-truth labels. In this way, the server can only assist device-defined side-network computing, and learn nothing about data and labels. Extensive experimental results demonstrate PAE MobiLLM's superiority.

cs.LG↗

arXiv2025

Fed MobiLLM: Efficient Federated LLM Fine-Tuning over Heterogeneous Mobile Devices via Server Assisted Side-Tuning

Collaboratively fine-tuning (FT) large language models (LLMs) over heterogeneous mobile devices fosters immense potential applications of personalized intelligence. However, such a vision faces critical system challenges. Conventional federated LLM FT approaches place prohibitive computational and memory burdens on mobile hardware, and their synchronous model aggregation protocols stall for slower devices. In this paper, we propose Fed MobiLLM, a novel design to facilitate efficient federated LLM FT across mobile devices with diverse computing/communication speeds and local model architectures. In particular, Fed MobiLLM implements a pioneering server-assisted federated side-tuning paradigm. Briefly, mobile devices perform lightweight forward propagation computations on local data using their frozen pre-scaled backbone LLMs, and then upload selected intermediate activations. The server trains a shared side-network independently, eliminating client-side backpropagation and enabling asynchronous updates. To bridge model heterogeneity across different devices, we introduce an adaptive layer-wise feature alignment method, which ensures consistent representations for collaboratively tuning a shared side network. Extensive experimental results demonstrate that Fed MobiLLM can maintain robust fine-tuning performance while achieving extremely low on-device memory, with at least 95.2% reduction in computation overhead, 93.2% reduction in communication costs and 5.1x faster convergence compared to existing methods, validating its efficacy for practical LLM adaptation over heterogeneous mobile devices.

cs.LG↗

arXiv2025

Backward Joint Model for the Joint Dynamic Prediction of Time-to-Event and Longitudinal Data: Basic Formulation and New Developments

Dynamic prediction of future clinical outcomes based on longitudinally measured predictors plays a crucial role in disease management and patient counseling, particularly when conventional static models are inadequate. Joint modeling of longitudinal and time-to-event data provides a useful framework for addressing this challenge. In this paper, we present a comprehensive development of the recently proposed backward joint model (BJM; Shen and Li 2021}, which factorizes the likelihood into the distribution of time-to-event data and the conditional distribution of longitudinal data given the event time. This structure facilitates computation and is well-suited for multivariate longitudinal data. We introduce several novel developments to the BJM, including the extrapolation and two-part specifications, as well as the incorporation of competing risks. We also address an important yet underexplored problem in the literature: predicting future longitudinal trajectories conditional on predicted event times. Additionally, we explore the connection between BJM and existing joint modeling approaches. All these extensions preserve the computational advantages of the basic BJM formulation, including one-dimensional numerical integration, convex optimization via the EM algorithm, and a quick procedure for consistent estimation using standard software. We evaluate the method's performance through simulation studies and illustrate its utility in a chronic kidney disease application.

stat.ME↗

arXiv2025

Progressive Homeostatic and Plastic Prompt Tuning for Audio-Visual Multi-Task Incremental Learning

Audio-visual multi-task incremental learning aims to continuously learn from multiple audio-visual tasks without the need for joint training on all tasks. The challenge of the problem is how to preserve the old task knowledge while facilitating the learning of new task with previous experiences. To address these challenges, we introduce a three-stage Progressive Homeostatic and Plastic audio-visual prompt (PHP) method. In the shallow phase, we design the task-shared modality aggregating adapter to foster cross-task and cross-modal audio-visual representation learning to enhance shared understanding between tasks. In the middle phase, we propose the task-specific modality-shared dynamic generating adapter, which constructs prompts that are tailored to individual tasks while remaining general across modalities, which balances the models ability to retain knowledge against forgetting with its potential for versatile multi-task transferability. In the deep phase, we introduce the task-specific modality-independent prompts to further refine the understand ability by targeting individual information for each task and modality. By incorporating these three phases, PHP retains task-specific prompts while adapting shared parameters for new tasks to effectively balance knowledge sharing and specificity. Our method achieves SOTA performance in different orders of four tasks (AVE, AVVP, AVS and AVQA). Our code can be available at https://github.com/ENJOY-Yin-jiong/PHP.

cs.AI↗

arXiv2025

Flavor Physics at the CEPC: a General Perspective

We discuss the landscape of flavor physics at the Circular Electron-Positron Collider (CEPC), based on the nominal luminosity outlined in its Technical Design Report. The CEPC is designed to operate in multiple modes to address a variety of tasks. At the $Z$ pole, the expected production of 4 Tera $Z$ bosons will provide unique and highly precise measurements of $Z$ boson couplings, while the substantial number of boosted heavy-flavored quarks and leptons produced in clean $Z$ decays will facilitate investigations into their flavor physics with unprecedented precision. We investigate the prospects of measuring various physics benchmarks and discuss their implications for particle theories and phenomenological models. Our studies indicate that, with its highlighted advantages and anticipated excellent detector performance, the CEPC can explore beauty and $τ$ physics in ways that are superior to or complementary with the Belle II and Large-Hadron-Collider-beauty experiments, potentially enabling the detection of new physics at energy scales of 10 TeV and above. This potential also extends to the observation of yet-to-be-discovered rare and exotic processes, as well as testing fundamental principles such as lepton flavor universality, lepton and baryon number conservation, etc., making the CEPC a vibrant platform for flavor physics research. The $WW$ threshold scan, Higgs-factory operation and top-pair productions of the CEPC further enhance its merits in this regard, especially for measuring the Cabibbo-Kobayashi-Maskawa matrix elements, and Flavor-Changing-Neutral-Current physics of Higgs boson and top quarks. We outline the requirements for detector performance and considerations for future development to achieve the anticipated scientific goals.

hep-ex↗

arXiv2025

AF-RLIO: Adaptive Fusion of Radar-LiDAR-Inertial Information for Robust Odometry in Challenging Environments

In robotic navigation, maintaining precise pose estimation and navigation in complex and dynamic environments is crucial. However, environmental challenges such as smoke, tunnels, and adverse weather can significantly degrade the performance of single-sensor systems like LiDAR or GPS, compromising the overall stability and safety of autonomous robots. To address these challenges, we propose AF-RLIO: an adaptive fusion approach that integrates 4D millimeter-wave radar, LiDAR, inertial measurement unit (IMU), and GPS to leverage the complementary strengths of these sensors for robust odometry estimation in complex environments. Our method consists of three key modules. Firstly, the pre-processing module utilizes radar data to assist LiDAR in removing dynamic points and determining when environmental conditions are degraded for LiDAR. Secondly, the dynamic-aware multimodal odometry selects appropriate point cloud data for scan-to-map matching and tightly couples it with the IMU using the Iterative Error State Kalman Filter. Lastly, the factor graph optimization module balances weights between odometry and GPS data, constructing a pose graph for optimization. The proposed approach has been evaluated on datasets and tested in real-world robotic environments, demonstrating its effectiveness and advantages over existing methods in challenging conditions such as smoke and tunnels.

cs.RO↗

arXiv2025

Lunar Orbital VLBI Experiment: motivation, scientific purposes and status

The Lunar Orbital VLBI Experiment (LOVEX) is a scientific component of the Chinese Lunar Exploration Project (CLEP) Chang'E-7. The spaceborne component of LOVEX is implemented onboard the relay satellite QueQiao-2, which was launched on 2024 March 20, and later placed into an elliptical selenocentric orbit. The LOVEX-specific payload consists of an X-band cryogenic receiver, a hydrogen maser frequency standard, and VLBI data formatting and acquisition electronics. Several components of the QueQiao-2 nominal onboard instrumentation, such as the 4.2-meter antenna, the data storage device, and the downlink communication system, contribute to the overall spaceborne VLBI instrumentation. This allows us to form a space radio telescope capable of co-observing with Earth-based radio telescopes in VLBI mode. In this space VLBI system, the length of the baseline extends up to approximately 380,000 km. This paper presents the LOVEX scientific objectives, architecture, instrumentation, pre-launch tests, in-flight verification and calibration, and the first in-flight detections of interferometric response (''fringes'') achieved through observations of the quasar AO 0235+164 and the Chang'E-6 orbital module, positioned at the Sun-Earth Lagrange point L2. These initial results demonstrate the successful performance of LOVEX, verifying its capability for both astronomical and spacecraft tracking observations at ultra-long VLBI baselines.

astro-ph.IM↗