arXiv subjects

Xiaofan Li

Publications and source records attributed to Xiaofan Li.

At least 55 records · Page 3Linked to original sources

arXiv2026

Enhancing Descriptive Captions with Visual Attributes for Multimodal Perception

Training Large Multimodality Models (LMMs) relies on descriptive image caption that connects image and language. Existing methods for generating such captions often rely on distilling the captions from pretrained LMMs, constructing them from publicly available internet images, or even generating them through human annotation. However, these strategies can fall short in terms of precision and granularity, particularly when dealing with complex visual reasoning tasks. In this paper, we propose to leverage off-the-shelf visual specialists, which were trained from annotated images initially not for image captioning, for enhancing the image caption. Our approach, named EDC, explores object low-level and fine-grained attributes (e.g., depth, emotion and fine-grained categories) and object relations (e.g., relative location and human-object-interaction (HOI)), and combine the attributes into the descriptive caption. By systematically integrating these rich attributes into the generated captions, EDC significantly improves the descriptive quality of the captions, providing a deeper and more nuanced understanding of the visual content. Experiments demonstrate that such visual specialists are able to improve the performance for visual understanding tasks as well as reasoning that benefits from more accurate visual understanding. The complete source code of EDC pipeline and datasets will be available at https://github.com/syp2ysy/DCE.

cs.CV↗

arXiv2025

FaithFusion: Harmonizing Reconstruction and Generation via Pixel-wise Information Gain

In controllable driving-scene reconstruction and 3D scene generation, maintaining geometric fidelity while synthesizing visually plausible appearance under large viewpoint shifts is crucial. However, effective fusion of geometry-based 3DGS and appearance-driven diffusion models faces inherent challenges, as the absence of pixel-wise, 3D-consistent editing criteria often leads to over-restoration and geometric drift. To address these issues, we introduce \textbf{FaithFusion}, a 3DGS-diffusion fusion framework driven by pixel-wise Expected Information Gain (EIG). EIG acts as a unified policy for coherent spatio-temporal synthesis: it guides diffusion as a spatial prior to refine high-uncertainty regions, while its pixel-level weighting distills the edits back into 3DGS. The resulting plug-and-play system is free from extra prior conditions and structural modifications.Extensive experiments on the Waymo dataset demonstrate that our approach attains SOTA performance across NTA-IoU, NTL-IoU, and FID, maintaining an FID of 107.47 even at 6 meters lane shift. Our code is available at https://github.com/wangyuanbiubiubiu/FaithFusion.

cs.CV↗

arXiv2025

FVAR: Visual Autoregressive Modeling via Next Focus Prediction

Visual autoregressive models achieve remarkable generation quality through next-scale predictions across multi-scale token pyramids. However, the conventional method uses uniform scale downsampling to build these pyramids, leading to aliasing artifacts that compromise fine details and introduce unwanted jaggies and moiré patterns. To tackle this issue, we present \textbf{FVAR}, which reframes the paradigm from \emph{next-scale prediction} to \emph{next-focus prediction}, mimicking the natural process of camera focusing from blur to clarity. Our approach introduces three key innovations: \textbf{1) Next-Focus Prediction Paradigm} that transforms multi-scale autoregression by progressively reducing blur rather than simply downsampling; \textbf{2) Progressive Refocusing Pyramid Construction} that uses physics-consistent defocus kernels to build clean, alias-free multi-scale representations; and \textbf{3) High-Frequency Residual Learning} that employs a specialized residual teacher network to effectively incorporate alias information during training while maintaining deployment simplicity. Specifically, we construct optical low-pass views using defocus point spread function (PSF) kernels with decreasing radius, creating smooth blur-to-clarity transitions that eliminate aliasing at its source. To further enhance detail generation, we introduce a High-Frequency Residual Teacher that learns from both clean structure and alias residuals, distilling this knowledge to a vanilla VAR deployment network for seamless inference. Extensive experiments on ImageNet demonstrate that FVAR substantially reduces aliasing artifacts, improves fine detail preservation, and enhances text readability, achieving superior performance with perfect compatibility to existing VAR frameworks.

cs.CV↗

arXiv2025

Video4Edit: Viewing Image Editing as a Degenerate Temporal Process

We observe that recent advances in multimodal foundation models have propelled instruction-driven image generation and editing into a genuinely cross-modal, cooperative regime. Nevertheless, state-of-the-art editing pipelines remain costly: beyond training large diffusion/flow models, they require curating massive high-quality triplets of \{instruction, source image, edited image\} to cover diverse user intents. Moreover, the fidelity of visual replacements hinges on how precisely the instruction references the target semantics. We revisit this challenge through the lens of temporal modeling: if video can be regarded as a full temporal process, then image editing can be seen as a degenerate temporal process. This perspective allows us to transfer single-frame evolution priors from video pre-training, enabling a highly data-efficient fine-tuning regime. Empirically, our approach matches the performance of leading open-source baselines while using only about one percent of the supervision demanded by mainstream editing models.

cs.CV↗

arXiv2025

Exploring Effective Factors for Improving Visual In-Context Learning

The In-Context Learning (ICL) is to understand a new task via a few demonstrations (aka. prompt) and predict new inputs without tuning the models. While it has been widely studied in NLP, it is still a relatively new area of research in computer vision. To reveal the factors influencing the performance of visual in-context learning, this paper shows that prompt selection and prompt fusion are two major factors that have a direct impact on the inference performance of visual context learning. Prompt selection is the process of identifying the most appropriate prompt or example to help the model understand new tasks. This is important because providing the model with relevant prompts can help it learn more effectively and efficiently. Prompt fusion involves combining knowledge from different positions within the large-scale visual model. By doing this, the model can leverage the diverse knowledge stored in different parts of the model to improve its performance on new tasks. Based these findings, we propose a simple framework prompt-SelF for visual in-context learning. Specifically, we first use the pixel-level retrieval method to select a suitable prompt, and then use different prompt fusion methods to activate all the knowledge stored in the large-scale model, and finally ensemble the prediction results obtained from different prompt fusion methods to obtain the final prediction results. And we conduct extensive experiments on single-object segmentation and detection tasks to demonstrate the effectiveness of prompt-SelF. Remarkably, the prompt-SelF has outperformed OSLSM based meta-learning in 1-shot segmentation for the first time. This indicated the great potential of visual in-context learning. The source code and models will be available at https://github.com/syp2ysy/prompt-SelF.

cs.CV↗

arXiv2025

VRP-SAM: SAM with Visual Reference Prompt

In this paper, we propose a novel Visual Reference Prompt (VRP) encoder that empowers the Segment Anything Model (SAM) to utilize annotated reference images as prompts for segmentation, creating the VRP-SAM model. In essence, VRP-SAM can utilize annotated reference images to comprehend specific objects and perform segmentation of specific objects in target image. It is note that the VRP encoder can support a variety of annotation formats for reference images, including \textbf{point}, \textbf{box}, \textbf{scribble}, and \textbf{mask}. VRP-SAM achieves a breakthrough within the SAM framework by extending its versatility and applicability while preserving SAM's inherent strengths, thus enhancing user-friendliness. To enhance the generalization ability of VRP-SAM, the VRP encoder adopts a meta-learning strategy. To validate the effectiveness of VRP-SAM, we conducted extensive empirical studies on the Pascal and COCO datasets. Remarkably, VRP-SAM achieved state-of-the-art performance in visual reference segmentation with minimal learnable parameters. Furthermore, VRP-SAM demonstrates strong generalization capabilities, allowing it to perform segmentation of unseen objects and enabling cross-domain segmentation. The source code and models will be available at https://github.com/syp2ysy/VRP-SAM

cs.CV↗

arXiv2025

Ocean Diviner: A Diffusion-Augmented Reinforcement Learning Framework for AUV Robust Control in Underwater Tasks

Autonomous Underwater Vehicles (AUVs) are essential for marine exploration, yet their control remains highly challenging due to nonlinear dynamics and uncertain environmental disturbances. This paper presents a diffusion-augmented Reinforcement Learning (RL) framework for robust AUV control, aiming to improve AUV's adaptability in dynamic underwater environments. The proposed framework integrates two core innovations: (1) A diffusion-based action generation framework that produces physically feasible and high-quality actions, enhanced by a high-dimensional state encoding mechanism combining current observations with historical states and actions through a novel diffusion U-Net architecture, significantly improving long-horizon planning capacity for robust control. (2) A sample-efficient hybrid learning architecture that synergizes diffusion-guided exploration with RL policy optimization, where the diffusion model generates diverse candidate actions and the RL critic selects the optimal action, achieving higher exploration efficiency and policy stability in dynamic underwater environments. Extensive simulation experiments validate the framework's superior robustness and flexibility, outperforming conventional control methods in challenging marine conditions, offering enhanced adaptability and reliability for AUV operations in underwater tasks. Finally, we will release the code publicly soon to support future research in this area.

cs.RO↗

arXiv2025

HAMMER: Hamiltonian Curiosity Augmented Large Language Model Reinforcement

Recent curriculum reinforcement learning for large language models (LLMs) typically rely on difficulty-based annotations for data filtering and ordering. However, such methods suffer from local optimization, where continual training on simple samples in the early steps can cause the policy to lose its exploration. We propose a novel schema, namely Hamiltonian curiosity augmented large language model reinforcement (HAMMER), that transfers diversity metrics, commonly used in dataset evaluation, into the dynamic reinforcement learning procedure, where training samples are ordered via a minimum-semantic Hamiltonian path making the initial training retrain more exploration. From a theoretical perspective of generalization bounds, diversity-driven ordering facilitates stable convergence. Empirical evaluations indicate that HAMMER stimulates model "curiosity" and consistently achieves a 3% to 4% average accuracy gain across diverse inference benchmark.

cs.LG↗

arXiv2025

Investigating Security Implications of Automatically Generated Code on the Software Supply Chain

In recent years, various software supply chain (SSC) attacks have posed significant risks to the global community. Severe consequences may arise if developers integrate insecure code snippets that are vulnerable to SSC attacks into their products. Particularly, code generation techniques, such as large language models (LLMs), have been widely utilized in the developer community. However, LLMs are known to suffer from inherent issues when generating code, including fabrication, misinformation, and reliance on outdated training data, all of which can result in serious software supply chain threats. In this paper, we investigate the security threats to the SSC that arise from these inherent issues. We examine three categories of threats, including eleven potential SSC-related threats, related to external components in source code, and continuous integration configuration files. We find some threats in LLM-generated code could enable attackers to hijack software and workflows, while some others might cause potential hidden threats that compromise the security of the software over time. To understand these security impacts and severity, we design a tool, SSCGuard, to generate 439,138 prompts based on SSC-related questions collected online, and analyze the responses of four popular LLMs from GPT and Llama. Our results show that all identified SSC-related threats persistently exist. To mitigate these risks, we propose a novel prompt-based defense mechanism, namely Chain-of-Confirmation, to reduce fabrication, and a middleware-based defense that informs users of various SSC threats.

cs.CR↗

arXiv2025

Gate-Based and Annealing-Based Quantum Algorithms for the Maximum K-Plex Problem

The $ k $-plex model, which allows each vertex to miss connections with up to $ k $ neighbors, serves as a relaxation of the clique. Its adaptability makes it more suitable for analyzing real-world graphs where noise and imperfect data are common and the ideal clique model is often impractical. The problem of identifying the maximum $ k $-plex (MKP, which is NP-hard) is gaining attention in fields such as social network analysis, community detection, terrorist network identification, and graph clustering. Recent works have focused on optimizing the time complexity of MKP algorithms. The state-of-the-art has reduced the complexity from a trivial $ O^*(2^n) $ to $ O^*(c_k^n) $, with $ c_k > 1.94 $ for $ k \geq 3 $, where $ n $ denotes the vertex number. This paper investigates the MKP using two quantum models: gate-based model and annealing-based model. Two gate-based algorithms, qTKP and qMKP, are proposed to achieve $ O^*(1.42^n) $ time complexity. qTKP integrates quantum search with graph encoding, degree counting, degree comparison, and size determination to find a $ k $-plex of a given size; qMKP uses binary search to progressively identify the maximum solution. Furthermore, by reformulating MKP as a quadratic unconstrained binary optimization problem, we propose qaMKP, the first annealing-based approximation algorithm, which utilizes qubit resources more efficiently than gate-based algorithms. To validate the practical performance, proof-of-principle experiments were conducted using the latest IBM gate-based quantum simulator and D-Wave adiabatic quantum computer. This work holds potential to be applied to a wide range of clique relaxations, e.g., $ n $-clan and $ n $-club.

cs.DB↗

arXiv2025

Autonomous Driving with RSMA-Enabled Finite Blocklength Transmissions: Ergodic Performance Analysis and Optimization

Rate-splitting multiple access (RSMA) is a key technology for next-generation multiple access systems due to its robustness against imperfect channel state information (CSI). This makes RSMA particularly suitable for high-mobility autonomous driving, where ultra-reliable and low-latency communication (URLLC) is essential. To address the stringent requirements, this study enables RSMA finite blocklength (FBL) transmissions and explicitly evaluates the ergodic performance. We derive the closed-form lower bound for the ergodic sum-rate of RSMA, considering vital factors such as the vehicle velocities, vehicle positions, power allocation of each stream, blocklengths, and block error rates (BLERs). To further enhance the ergodic sum-rate while complying with quality of service (QoS) rate constraints, we jointly optimize the global power coefficient, private power distribution, and common rate splitting. Guided by gradient descent, we first adjust the global power coefficient based on its sum-rate solution. This parameter regulates the power state of the common stream, allowing for dynamic activation or deactivation: if active, we optimize the private power distribution and adjust the common rate splitting to meet minimum transmission constraints; if inactive, we use the sequential quadratic programming for private power distribution optimization. Simulation results confirm that our RSMA scheme significantly improves the ergodic performance, reduces blocklength and BLER, surpassing the RSMA counterpart with average private power and space division multiple access (SDMA). Furthermore, our approach is validated to guarantee the rates for users with the poorest channel conditions, thereby enhancing fairness across the network.

eess.SP↗

arXiv2025

CoopTrack: Exploring End-to-End Learning for Efficient Cooperative Sequential Perception

Cooperative perception aims to address the inherent limitations of single-vehicle autonomous driving systems through information exchange among multiple agents. Previous research has primarily focused on single-frame perception tasks. However, the more challenging cooperative sequential perception tasks, such as cooperative 3D multi-object tracking, have not been thoroughly investigated. Therefore, we propose CoopTrack, a fully instance-level end-to-end framework for cooperative tracking, featuring learnable instance association, which fundamentally differs from existing approaches. CoopTrack transmits sparse instance-level features that significantly enhance perception capabilities while maintaining low transmission costs. Furthermore, the framework comprises two key components: Multi-Dimensional Feature Extraction, and Cross-Agent Association and Aggregation, which collectively enable comprehensive instance representation with semantic and motion features, and adaptive cross-agent association and fusion based on a feature graph. Experiments on both the V2X-Seq and Griffin datasets demonstrate that CoopTrack achieves excellent performance. Specifically, it attains state-of-the-art results on V2X-Seq, with 39.0\% mAP and 32.8\% AMOTA. The project is available at https://github.com/zhongjiaru/CoopTrack.

cs.CV↗

arXiv2025

One-for-More: Continual Diffusion Model for Anomaly Detection

With the rise of generative models, there is a growing interest in unifying all tasks within a generative framework. Anomaly detection methods also fall into this scope and utilize diffusion models to generate or reconstruct normal samples when given arbitrary anomaly images. However, our study found that the diffusion model suffers from severe ``faithfulness hallucination'' and ``catastrophic forgetting'', which can't meet the unpredictable pattern increments. To mitigate the above problems, we propose a continual diffusion model that uses gradient projection to achieve stable continual learning. Gradient projection deploys a regularization on the model updating by modifying the gradient towards the direction protecting the learned knowledge. But as a double-edged sword, it also requires huge memory costs brought by the Markov process. Hence, we propose an iterative singular value decomposition method based on the transitive property of linear representation, which consumes tiny memory and incurs almost no performance loss. Finally, considering the risk of ``over-fitting'' to normal images of the diffusion model, we propose an anomaly-masked network to enhance the condition mechanism of the diffusion model. For continual anomaly detection, ours achieves first place in 17/18 settings on MVTec and VisA. Code is available at https://github.com/FuNz-0/One-for-More

cs.CV↗

arXiv2025

CodeAgents: A Token-Efficient Framework for Codified Multi-Agent Reasoning in LLMs

Effective prompt design is essential for improving the planning capabilities of large language model (LLM)-driven agents. However, existing structured prompting strategies are typically limited to single-agent, plan-only settings, and often evaluate performance solely based on task accuracy - overlooking critical factors such as token efficiency, modularity, and scalability in multi-agent environments. To address these limitations, we introduce CodeAgents, a prompting framework that codifies multi-agent reasoning and enables structured, token-efficient planning in multi-agent systems. In CodeAgents, all components of agent interaction - Task, Plan, Feedback, system roles, and external tool invocations - are codified into modular pseudocode enriched with control structures (e.g., loops, conditionals), boolean logic, and typed variables. This design transforms loosely connected agent plans into cohesive, interpretable, and verifiable multi-agent reasoning programs. We evaluate the proposed framework across three diverse benchmarks - GAIA, HotpotQA, and VirtualHome - using a range of representative LLMs. Results show consistent improvements in planning performance, with absolute gains of 3-36 percentage points over natural language prompting baselines. On VirtualHome, our method achieves a new state-of-the-art success rate of 56%. In addition, our approach reduces input and output token usage by 55-87% and 41-70%, respectively, underscoring the importance of token-aware evaluation metrics in the development of scalable multi-agent LLM systems. The code and resources are available at: https://anonymous.4open.science/r/CodifyingAgent-5A86

cs.AI↗

arXiv2025

LT-PINN: Lagrangian Topology-conscious Physics-informed Neural Network for Boundary-focused Engineering Optimization

Physics-informed neural networks (PINNs) have emerged as a powerful meshless tool for topology optimization, capable of simultaneously determining optimal topologies and physical solutions. However, conventional PINNs rely on density-based topology descriptions, which necessitate manual interpolation and limit their applicability to complex geometries. To address this, we propose Lagrangian topology-conscious PINNs (LT-PINNs), a novel framework for boundary-focused engineering optimization. By parameterizing the control variables of topology boundary curves as learnable parameters, LT-PINNs eliminate the need for manual interpolation and enable precise boundary determination. We further introduce specialized boundary condition loss function and topology loss function to ensure sharp and accurate boundary representations, even for intricate topologies. The accuracy and robustness of LT-PINNs are validated via two types of partial differential equations (PDEs), including elastic equation with Dirichlet boundary conditions and Laplace's equation with Neumann boundary conditions. Furthermore, we demonstrate effectiveness of LT-PINNs on more complex time-dependent and time-independent flow problems without relying on measurement data, and showcase their engineering application potential in flow velocity rearrangement, transforming a uniform upstream velocity into a sine-shaped downstream profile. The results demonstrate (1) LT-PINNs achieve substantial reductions in relative L2 errors compared with the state-of-art density topology-oriented PINNs (DT-PINNs), (2) LT-PINNs can handle arbitrary boundary conditions, making them suitable for a wide range of PDEs, and (3) LT-PINNs can infer clear topology boundaries without manual interpolation, especially for complex topologies.

cs.LG↗

arXiv2025

Drag Your Gaussian: Effective Drag-Based Editing with Score Distillation for 3D Gaussian Splatting

Recent advancements in 3D scene editing have been propelled by the rapid development of generative models. Existing methods typically utilize generative models to perform text-guided editing on 3D representations, such as 3D Gaussian Splatting (3DGS). However, these methods are often limited to texture modifications and fail when addressing geometric changes, such as editing a character's head to turn around. Moreover, such methods lack accurate control over the spatial position of editing results, as language struggles to precisely describe the extent of edits. To overcome these limitations, we introduce DYG, an effective 3D drag-based editing method for 3D Gaussian Splatting. It enables users to conveniently specify the desired editing region and the desired dragging direction through the input of 3D masks and pairs of control points, thereby enabling precise control over the extent of editing. DYG integrates the strengths of the implicit triplane representation to establish the geometric scaffold of the editing results, effectively overcoming suboptimal editing outcomes caused by the sparsity of 3DGS in the desired editing regions. Additionally, we incorporate a drag-based Latent Diffusion Model into our method through the proposed Drag-SDS loss function, enabling flexible, multi-view consistent, and fine-grained editing. Extensive experiments demonstrate that DYG conducts effective drag-based editing guided by control point prompts, surpassing other baselines in terms of editing effect and quality, both qualitatively and quantitatively. Visit our project page at https://quyans.github.io/Drag-Your-Gaussian.

cs.GR↗

arXiv2025

Low-Complexity Hybrid Beamforming for Multi-Cell mmWave Massive MIMO: A Primitive Kronecker Decomposition Approach

To circumvent the high path loss of mmWave propagation and reduce the hardware cost of massive multiple-input multiple-output antenna systems, full-dimensional hybrid beamforming is critical in 5G and beyond wireless communications. Concerning an uplink multi-cell system with a large-scale uniform planar antenna array, this paper designs an efficient hybrid beamformer using primitive Kronecker decomposition and dynamic factor allocation, where the analog beamformer applies to null the inter-cell interference and simultaneously enhances the desired signals. In contrast, the digital beamformer mitigates the intra-cell interference using the minimum mean square error (MMSE) criterion. Then, due to the low accuracy of phase shifters inherent in the analog beamformer, a low-complexity hybrid beamformer is developed to slow its adjustment speed. Next, an optimality analysis from a subspace perspective is performed, and a sufficient condition for optimal antenna configuration is established. Finally, simulation results demonstrate that the achievable sum rate of the proposed beamformer approaches that of the optimal pure digital MMSE scheme, yet with much lower computational complexity and hardware cost.

cs.IT↗

arXiv2025

BEVWorld: A Multimodal World Simulator for Autonomous Driving via Scene-Level BEV Latents

World models have attracted increasing attention in autonomous driving for their ability to forecast potential future scenarios. In this paper, we propose BEVWorld, a novel framework that transforms multimodal sensor inputs into a unified and compact Bird's Eye View (BEV) latent space for holistic environment modeling. The proposed world model consists of two main components: a multi-modal tokenizer and a latent BEV sequence diffusion model. The multi-modal tokenizer first encodes heterogeneous sensory data, and its decoder reconstructs the latent BEV tokens into LiDAR and surround-view image observations via ray-casting rendering in a self-supervised manner. This enables joint modeling and bidirectional encoding-decoding of panoramic imagery and point cloud data within a shared spatial representation. On top of this, the latent BEV sequence diffusion model performs temporally consistent forecasting of future scenes, conditioned on high-level action tokens, enabling scene-level reasoning over time. Extensive experiments demonstrate the effectiveness of BEVWorld on autonomous driving benchmarks, showcasing its capability in realistic future scene generation and its benefits for downstream tasks such as perception and motion prediction.

cs.CV↗