arXiv subjects

Xinyi Wang

Publications and source records attributed to Xinyi Wang.

At least 91 records · Page 5Linked to original sources

arXiv2025

AffordBot: 3D Fine-grained Embodied Reasoning via Multimodal Large Language Models

Effective human-agent collaboration in physical environments requires understanding not only what to act upon, but also where the actionable elements are and how to interact with them. Existing approaches often operate at the object level or disjointedly handle fine-grained affordance reasoning, lacking coherent, instruction-driven grounding and reasoning. In this work, we introduce a new task: Fine-grained 3D Embodied Reasoning, which requires an agent to predict, for each referenced affordance element in a 3D scene, a structured triplet comprising its spatial location, motion type, and motion axis, based on a task instruction. To solve this task, we propose AffordBot, a novel framework that integrates Multimodal Large Language Models (MLLMs) with a tailored chain-of-thought (CoT) reasoning paradigm. To bridge the gap between 3D input and 2D-compatible MLLMs, we render surround-view images of the scene and project 3D element candidates into these views, forming a rich visual representation aligned with the scene geometry. Our CoT pipeline begins with an active perception stage, prompting the MLLM to select the most informative viewpoint based on the instruction, before proceeding with step-by-step reasoning to localize affordance elements and infer plausible interaction motions. Evaluated on the SceneFun3D dataset, AffordBot achieves state-of-the-art performance, demonstrating strong generalization and physically grounded reasoning with only 3D point cloud input and MLLMs.

cs.CV↗

arXiv2025

CAMP-VQA: Caption-Embedded Multimodal Perception for No-Reference Quality Assessment of Compressed Video

The prevalence of user-generated content (UGC) on platforms such as YouTube and TikTok has rendered no-reference (NR) perceptual video quality assessment (VQA) vital for optimizing video delivery. Nonetheless, the characteristics of non-professional acquisition and the subsequent transcoding of UGC video on sharing platforms present significant challenges for NR-VQA. Although NR-VQA models attempt to infer mean opinion scores (MOS), their modeling of subjective scores for compressed content remains limited due to the absence of fine-grained perceptual annotations of artifact types. To address these challenges, we propose CAMP-VQA, a novel NR-VQA framework that exploits the semantic understanding capabilities of large vision-language models. Our approach introduces a quality-aware prompting mechanism that integrates video metadata (e.g., resolution, frame rate, bitrate) with key fragments extracted from inter-frame variations to guide the BLIP-2 pretraining approach in generating fine-grained quality captions. A unified architecture has been designed to model perceptual quality across three dimensions: semantic alignment, temporal characteristics, and spatial characteristics. These multimodal features are extracted and fused, then regressed to video quality scores. Extensive experiments on a wide variety of UGC datasets demonstrate that our model consistently outperforms existing NR-VQA methods, achieving improved accuracy without the need for costly manual fine-grained annotations. Our method achieves the best performance in terms of average rank and linear correlation (SRCC: 0.928, PLCC: 0.938) compared to state-of-the-art methods. The source code and trained models, along with a user-friendly demo, are available at: https://github.com/xinyiW915/CAMP-VQA.

eess.IV↗

arXiv2025

Pre-equalization Design for ISAC-OTFS Air-Ground Transmission: A Deep Learning Approach

Despite the strong Doppler resilience capability, orthogonal time-frequency space (OTFS) modulation suffers from high channel estimation and equalization complexity at the receiver, hindering its applicability in air-ground transmission. In this paper, we propose a pre-equalization-based integrated sensing and communications-OTFS downlink transmission framework in which the terrestrial access point executes pre-equalization using the predicted channel state information (CSI), so that the unmanned aerial vehicle can perform direct symbol detection without channel equalization. In particular, the mean square error of OTFS symbol demodulation and Cramer-Rao lower bound of sensing parameter estimation are considered, with their weighted sum utilized as the metric for optimizing the pre-equalization matrix. To address the time-varying CSI, we develop a deep learning based framework composed of channel prediction and pre-equalization. In particular, a parameter-level channel prediction module is utilized to decouple OTFS channel parameters, and a low-dimensional prediction network is leveraged to correct outdated CSI, which is then used to initialize the input of the pre-equalization module. Finally, a dual-branch residual-structured deep neural network is cascaded to execute pre-equalization. Simulation results show that the proposed channel prediction-based pre-equalization framework significantly reduces receiver complexity and pilot overhead while achieving symbol detection performance close to minimum mean square error equalization with perfect CSI under high mobility, as well as substantially improving sensing accuracy.

eess.SP↗

arXiv2025

MixedGaussianAvatar: Realistically and Geometrically Accurate Head Avatar via Mixed 2D-3D Gaussians

Reconstructing high-fidelity 3D head avatars is crucial in various applications such as virtual reality. The pioneering methods reconstruct realistic head avatars with Neural Radiance Fields (NeRF), which have been limited by training and rendering speed. Recent methods based on 3D Gaussian Splatting (3DGS) significantly improve the efficiency of training and rendering. However, the surface inconsistency of 3DGS results in subpar geometric accuracy; later, 2DGS uses 2D surfels to enhance geometric accuracy at the expense of rendering fidelity. To leverage the benefits of both 2DGS and 3DGS, we propose a novel method named MixedGaussianAvatar for realistically and geometrically accurate head avatar reconstruction. Our main idea is to utilize 2D Gaussians to reconstruct the surface of the 3D head, ensuring geometric accuracy. We attach the 2D Gaussians to the triangular mesh of the FLAME model and connect additional 3D Gaussians to those 2D Gaussians where the rendering quality of 2DGS is inadequate, creating a mixed 2D-3D Gaussian representation. These 2D-3D Gaussians can then be animated using FLAME parameters. We further introduce a progressive training strategy that first trains the 2D Gaussians and then fine-tunes the mixed 2D-3D Gaussians. We use a unified mixed Gaussian representation to integrate the two modalities of 2D image and 3D mesh. Furthermore, the comprehensive experiments demonstrate the superiority of MixedGaussianAvatar. The code will be released.

cs.CV↗

arXiv2025

Encoder-Decoder or Decoder-Only? Revisiting Encoder-Decoder Large Language Model

Recent large language model (LLM) research has undergone an architectural shift from encoder-decoder modeling to nowadays the dominant decoder-only modeling. This rapid transition, however, comes without a rigorous comparative analysis especially \textit{from the scaling perspective}, raising concerns that the potential of encoder-decoder models may have been overlooked. To fill this gap, we revisit encoder-decoder LLM (RedLLM), enhancing it with recent recipes from decoder-only LLM (DecLLM). We conduct a comprehensive comparison between RedLLM, pretrained with prefix language modeling (LM), and DecLLM, pretrained with causal LM, at different model scales, ranging from $\sim$150M to $\sim$8B. Using RedPajama V1 (1.6T tokens) for pretraining and FLAN for instruction tuning, our experiments show that RedLLM produces compelling scaling properties and surprisingly strong performance. While DecLLM is overall more compute-optimal during pretraining, RedLLM demonstrates comparable scaling and context length extrapolation capabilities. After instruction tuning, RedLLM achieves comparable and even better results on various downstream tasks while enjoying substantially better inference efficiency. We hope our findings could inspire more efforts on re-examining RedLLM, unlocking its potential for developing powerful and efficient LLMs.

cs.CL↗

arXiv2025

Trajectory Design for UAV-Based Low-Altitude Wireless Networks in Unknown Environments: A Digital Twin-Assisted TD3 Approach

Unmanned aerial vehicles (UAVs) are emerging as key enablers for low-altitude wireless network (LAWN), particularly when terrestrial networks are unavailable. In such scenarios, the environmental topology is typically unknown; hence, designing efficient and safe UAV trajectories is essential yet challenging. To address this, we propose a digital twin (DT)-assisted training and deployment framework. In this framework, the UAV transmits integrated sensing and communication signals to provide communication services to ground users, while simultaneously collecting echoes that are uploaded to the DT server to progressively construct virtual environments (VEs). These VEs accelerate model training and are continuously updated with real-time UAV sensing data during deployment, supporting decision-making and enhancing flight safety. Based on this framework, we further develop a trajectory design scheme that integrates simulated annealing for efficient user scheduling with the twin-delayed deep deterministic policy gradient algorithm for continuous trajectory design, aiming to minimize mission completion time while ensuring obstacle avoidance. Simulation results demonstrate that the proposed approach achieves faster convergence, higher flight safety, and shorter mission completion time compared with baseline methods, providing a robust and efficient solution for LAWN deployment in unknown environments.

eess.SP↗

arXiv2025

Quantum Artificial Intelligence for Software Engineering: the Road Ahead

In order to handle the increasing complexity of software systems, Artificial Intelligence (AI) has been applied to various areas of software engineering, including requirements engineering, coding, testing, and debugging. This has led to the emergence of AI for Software Engineering as a distinct research area within the field of software engineering. With the development of quantum computing, the field of Quantum AI (QAI) is arising, enhancing the performance of classical AI and holding significant potential for solving classical software engineering problems. Some initial applications of QAI in software engineering have already emerged, such as test case optimization. However, the path ahead remains open, offering ample opportunities to solve complex software engineering problems cost-effectively with QAI. To this end, this paper presents a roadmap towards the application of QAI in software engineering. Specifically, we consider two of the main categories of QAI, i.e., quantum optimization algorithms and quantum machine learning. For each software engineering phase, we discuss how these QAI approaches can address some of the tasks associated with that phase. Moreover, we provide an overview of some of the possible challenges that need to be addressed to make the application of QAI for software engineering successful.

cs.SE↗

arXiv2025

A Stitch in Time Saves Nine: Proactive Self-Refinement for Language Models

Recent advances in self-refinement have demonstrated significant potential for improving the outputs of large language models (LLMs) through iterative refinement. However, most existing self-refinement methods rely on a reactive process with a fixed number of iterations, making it difficult to determine the optimal timing and content of refinement based on the evolving generation context. Inspired by the way humans dynamically refine their thoughts during execution, we propose ProActive Self-Refinement (PASR), a novel method that enables LLMs to refine their outputs during the generation process. Unlike methods that regenerate entire responses, PASR proactively decides whether, when, and how to refine based on the model's internal state and evolving context. We conduct extensive experiments on a diverse set of 10 tasks to evaluate the effectiveness of PASR. Experimental results show that PASR significantly enhances problem-solving performance. In particular, on Qwen3-8B, PASR reduces average token consumption by 41.6% compared to standard generation, while also achieving an 8.2% improvement in accuracy. Our code and baselines used in the paper are available on GitHub.

cs.CL↗

arXiv2025

Sensing Performance Analysis in Cooperative Air-Ground ISAC Networks for LAE

To support the development of low altitude economy, the air-ground integrated sensing and communication (ISAC) networks need to be constructed to provide reliable and robust communication and sensing services. In this paper, the sensing capabilities in the cooperative air-ground ISAC networks are evaluated in terms of area radar detection coverage probability under a constant false alarm rate, where the distribution of aggregated sensing interferences is analyzed as a key intermediate result. Compared with the analysis based on the strongest interferer approximation, taking the aggregated sensing interference into consideration is better suited for pico-cell scenarios with high base station density. Simulations are conducted to validate the analysis.

cs.IT↗

arXiv2025

Towards Secure ISAC Beamforming: How Many Dedicated Sensing Beams Are Required?

In this paper, sensing-assisted secure communication in a multi-user multi-eavesdropper integrated sensing and communication (ISAC) system is investigated. Confidential communication signals and dedicated sensing signals are jointly transmitted by a base station (BS) to simultaneously serve users and sense aerial eavesdroppers (AEs). A sum rate maximization problem is formulated under AEs' Signal-to-Interference-plus-Noise Ratio (SINR) and sensing Signal-to-Clutter-plus-Noise Ratio (SCNR) constraints. A fractional-programming-based alternating optimization algorithm is developed to solve this problem for fully digital arrays, where successive convex approximation (SCA) and semidefinite relaxation (SDR) are leveraged to handle non-convex constraints. Furthermore, the minimum number of dedicated sensing beams is analyzed via a worst-case rank bound, upon which the proposed beamforming design is further extended to the hybrid analog-digital (HAD) array architecture, where the unit-modulus constraint is addressed by manifold optimization. Simulation results demonstrate that only a small number of sensing beams are sufficient for both sensing and jamming AEs, and the proposed designs consistently outperform strong baselines while also revealing the communication-sensing trade-off.

eess.SP↗

arXiv2025

AI Foundation Model for Time Series with Innovations Representation

This paper introduces an Artificial Intelligence (AI) foundation model for time series in engineering applications, where causal operations are required for real-time monitoring and control. Since engineering time series are governed by physical, rather than linguistic, laws, large-language-model-based AI foundation models may be ineffective or inefficient. Building on the classical innovations representation theory of Wiener, Kallianpur, and Rosenblatt, we propose Time Series GPT (TS-GPT) -- an innovations-representation-based Generative Pre-trained Transformer for engineering monitoring and control. As an example of foundation model adaptation, we consider Probabilistic Generative Forecasting, which produces future time series samples from conditional probability distributions given past realizations. We demonstrate the effectiveness of TS-GPT in forecasting real-time locational marginal prices using historical data from U.S. independent system operators.

stat.ML↗

arXiv2025

Prompt-Driven Agentic Video Editing System: Autonomous Comprehension of Long-Form, Story-Driven Media

Creators struggle to edit long-form, narrative-rich videos not because of UI complexity, but due to the cognitive demands of searching, storyboarding, and sequencing hours of footage. Existing transcript- or embedding-based methods fall short for creative workflows, as models struggle to track characters, infer motivations, and connect dispersed events. We present a prompt-driven, modular editing system that helps creators restructure multi-hour content through free-form prompts rather than timelines. At its core is a semantic indexing pipeline that builds a global narrative via temporal segmentation, guided memory compression, and cross-granularity fusion, producing interpretable traces of plot, dialogue, emotion, and context. Users receive cinematic edits while optionally refining transparent intermediate outputs. Evaluated on 400+ videos with expert ratings, QA, and preference studies, our system scales prompt-driven editing, preserves narrative coherence, and balances automation with creator control.

cs.AI↗

arXiv2025

Wireless Powered MEC Systems via Discrete Pinching Antennas: TDMA versus NOMA

Pinching antennas (PAs), a new type of reconfigurable and flexible antenna structures, have recently attracted significant research interest due to their ability to create line-of-sight links and mitigate large-scale path loss. Owing to their potential benefits, integrating PAs into wireless powered mobile edge computing (MEC) systems is regarded as a viable solution to enhance both energy transfer and task offloading efficiency. Unlike prior studies that assume ideal continuous PA placement along waveguides, this paper investigates a practical discrete PA-assisted wireless powered MEC framework, where devices first harvest energy from PA-emitted radio-frequency signals and then adopt a partial offloading mode, allocating part of the harvested energy to local computing and the remainder to uplink offloading. The uplink phase considers both the time-division multiple access (TDMA) and non-orthogonal multiple access (NOMA), each examined under three levels of PA activation flexibility. For each configuration, we formulate a joint optimization problem to maximize the total computational bits and conduct a theoretical performance comparison between the TDMA and NOMA schemes. To address the resulting mixed-integer nonlinear problems, we develop a two-layer algorithm that combines closed-form solutions based on Karush-Kuhn-Tucker (KKT) conditions with a cross-entropy-based learning method. Numerical results validate the superiority of the proposed design in terms of the harvested energy and computation performance, revealing that TDMA and NOMA achieve comparable performance under coarser PA activation levels, whereas finer activation granularity enables TDMA to achieve superior computation performance over NOMA.

eess.SP↗

arXiv2025

From Promise to Practical Reality: Transforming Diffusion MRI Analysis with Fast Deep Learning Enhancement

Fiber orientation distribution (FOD) is an advanced diffusion MRI modeling technique that represents complex white matter fiber configurations, and a key step for subsequent brain tractography and connectome analysis. Its reliability and accuracy, however, heavily rely on the quality of the MRI acquisition and the subsequent estimation of the FODs at each voxel. Generating reliable FODs from widely available clinical protocols with single-shell and low-angular-resolution acquisitions remains challenging but could potentially be addressed with recent advances in deep learning-based enhancement techniques. Despite advancements, existing methods have predominantly been assessed on healthy subjects, which have proved to be a major hurdle for their clinical adoption. In this work, we validate a newly optimized enhancement framework, FastFOD-Net, across healthy controls and six neurological disorders. This accelerated end-to-end deep learning framework enhancing FODs with superior performance and delivering training/inference efficiency for clinical use ($60\times$ faster comparing to its predecessor). With the most comprehensive clinical evaluation to date, our work demonstrates the potential of FastFOD-Net in accelerating clinical neuroscience research, empowering diffusion MRI analysis for disease differentiation, improving interpretability in connectome applications, and reducing measurement errors to lower sample size requirements. Critically, this work will facilitate the more widespread adoption of, and build clinical trust in, deep learning based methods for diffusion MRI enhancement. Specifically, FastFOD-Net enables robust analysis of real-world, clinical diffusion MRI data, comparable to that achievable with high-quality research acquisitions.

cs.CV↗

arXiv2025

PAR-AdvGAN: Improving Adversarial Attack Capability with Progressive Auto-Regression AdvGAN

Deep neural networks have demonstrated remarkable performance across various domains. However, they are vulnerable to adversarial examples, which can lead to erroneous predictions. Generative Adversarial Networks (GANs) can leverage the generators and discriminators model to quickly produce high-quality adversarial examples. Since both modules train in a competitive and simultaneous manner, GAN-based algorithms like AdvGAN can generate adversarial examples with better transferability compared to traditional methods. However, the generation of perturbations is usually limited to a single iteration, preventing these examples from fully exploiting the potential of the methods. To tackle this issue, we introduce a novel approach named Progressive Auto-Regression AdvGAN (PAR-AdvGAN). It incorporates an auto-regressive iteration mechanism within a progressive generation network to craft adversarial examples with enhanced attack capability. We thoroughly evaluate our PAR-AdvGAN method with a large-scale experiment, demonstrating its superior performance over various state-of-the-art black-box adversarial attacks, as well as the original AdvGAN.Moreover, PAR-AdvGAN significantly accelerates the adversarial example generation, i.e., achieving the speeds of up to 335.5 frames per second on Inception-v3 model, outperforming the gradient-based transferable attack algorithms. Our code is available at: https://github.com/LMBTough/PAR

cs.LG↗

arXiv2025

A Novel Symbol Level Precoding based AFDM Transmission Framework: Offloading Equalization Burden to Transmitter Side

Affine Frequency Division Multiplexing (AFDM) has attracted considerable attention for its robustness to Doppler effects. However, its high receiver-side computational complexity remains a major barrier to practical deployment. To address this, we propose a novel symbol-level precoding (SLP)-based AFDM transmission framework, which shifts the signal processing burden in downlink communications from user side to the base station (BS), enabling direct symbol detection without requiring channel estimation or equalization at the receiver. Specifically, in the uplink phase, we propose a Sparse Bayesian Learning (SBL) based channel estimation algorithm by exploiting the inherent sparsity of affine frequency (AF) domain channels. In particular, the sparse prior is modeled via a hierarchical Laplace distribution, and parameters are iteratively updated using the Expectation-Maximization (EM) algorithm. We also derive the Bayesian Cramer-Rao Bound (BCRB) to characterize the theoretical performance limit. In the downlink phase, the BS employs the SLP technology to design the transmitted waveform based on the estimated uplink channel state information (CSI) and channel reciprocity. The resulting optimization problem is formulated as a second-order cone programming (SOCP) problem, and its dual problem is investigated by Lagrangian function and Karush-Kuhn-Tucker conditions. Simulation results demonstrate that the proposed SBL estimator outperforms traditional orthogonal matching pursuit (OMP) in accuracy and robustness to off-grid effects, while the SLP-based waveform design scheme achieves performance comparable to conventional AFDM receivers while significantly reducing the computational complexity at receiver, validating the practicality of our approach.

eess.SP↗

arXiv2025

Mind the Generation Process: Fine-Grained Confidence Estimation During LLM Generation

While large language models (LLMs) have demonstrated remarkable performance across diverse tasks, they fundamentally lack self-awareness and frequently exhibit overconfidence, assigning high confidence scores to incorrect predictions. Accurate confidence estimation is therefore critical for enhancing the trustworthiness and reliability of LLM-generated outputs. However, existing approaches suffer from coarse-grained scoring mechanisms that fail to provide fine-grained, continuous confidence estimates throughout the generation process. To address these limitations, we introduce FineCE, a novel confidence estimation method that delivers accurate, fine-grained confidence scores during text generation. Specifically, we first develop a comprehensive pipeline for constructing training data that effectively captures the underlying probabilistic distribution of LLM responses, and then train a model to predict confidence scores for arbitrary text sequences in a supervised manner. Furthermore, we propose a Backward Confidence Integration (BCI) strategy that leverages information from the subsequent text to enhance confidence estimation for the current sequence during inference. We also introduce three strategies for identifying optimal positions to perform confidence estimation within the generation process. Extensive experiments on multiple benchmark datasets demonstrate that FineCE consistently outperforms existing classical confidence estimation methods. Our code and all baselines used in the paper are available on GitHub.

cs.CL↗

arXiv2025

DIVA-VQA: Detecting Inter-frame Variations in UGC Video Quality

The rapid growth of user-generated (video) content (UGC) has driven increased demand for research on no-reference (NR) perceptual video quality assessment (VQA). NR-VQA is a key component for large-scale video quality monitoring in social media and streaming applications where a pristine reference is not available. This paper proposes a novel NR-VQA model based on spatio-temporal fragmentation driven by inter-frame variations. By leveraging these inter-frame differences, the model progressively analyses quality-sensitive regions at multiple levels: frames, patches, and fragmented frames. It integrates frames, fragmented residuals, and fragmented frames aligned with residuals to effectively capture global and local information. The model extracts both 2D and 3D features in order to characterize these spatio-temporal variations. Experiments conducted on five UGC datasets and against state-of-the-art models ranked our proposed method among the top 2 in terms of average rank correlation (DIVA-VQA-L: 0.898 and DIVA-VQA-B: 0.886). The improved performance is offered at a low runtime complexity, with DIVA-VQA-B ranked top and DIVA-VQA-L third on average compared to the fastest existing NR-VQA method. Code and models are publicly available at: https://github.com/xinyiW915/DIVA-VQA.

eess.IV↗