arXiv subjects

Jian Wang

Publications and source records attributed to Jian Wang.

At least 397 records · Page 22Linked to original sources

arXiv2025

Efficient Tuning of Large Language Models for Knowledge-Grounded Dialogue Generation

Large language models (LLMs) demonstrate remarkable text comprehension and generation capabilities but often lack the ability to utilize up-to-date or domain-specific knowledge not included in their training data. To address this gap, we introduce KEDiT, an efficient method for fine-tuning LLMs for knowledge-grounded dialogue generation. KEDiT operates in two main phases: first, it employs an information bottleneck to compress retrieved knowledge into learnable parameters, retaining essential information while minimizing computational overhead. Second, a lightweight knowledge-aware adapter integrates these compressed knowledge vectors into the LLM during fine-tuning, updating less than 2\% of the model parameters. The experimental results on the Wizard of Wikipedia and a newly constructed PubMed-Dialog dataset demonstrate that KEDiT excels in generating contextually relevant and informative responses, outperforming competitive baselines in automatic, LLM-based, and human evaluations. This approach effectively combines the strengths of pretrained LLMs with the adaptability needed for incorporating dynamic knowledge, presenting a scalable solution for fields such as medicine.

cs.CL↗

arXiv2025

ZFusion: An Effective Fuser of Camera and 4D Radar for 3D Object Perception in Autonomous Driving

Reliable 3D object perception is essential in autonomous driving. Owing to its sensing capabilities in all weather conditions, 4D radar has recently received much attention. However, compared to LiDAR, 4D radar provides much sparser point cloud. In this paper, we propose a 3D object detection method, termed ZFusion, which fuses 4D radar and vision modality. As the core of ZFusion, our proposed FP-DDCA (Feature Pyramid-Double Deformable Cross Attention) fuser complements the (sparse) radar information and (dense) vision information, effectively. Specifically, with a feature-pyramid structure, the FP-DDCA fuser packs Transformer blocks to interactively fuse multi-modal features at different scales, thus enhancing perception accuracy. In addition, we utilize the Depth-Context-Split view transformation module due to the physical properties of 4D radar. Considering that 4D radar has a much lower cost than LiDAR, ZFusion is an attractive alternative to LiDAR-based methods. In typical traffic scenarios like the VoD (View-of-Delft) dataset, experiments show that with reasonable inference speed, ZFusion achieved the state-of-the-art mAP (mean average precision) in the region of interest, while having competitive mAP in the entire area compared to the baseline methods, which demonstrates performance close to LiDAR and greatly outperforms those camera-only methods.

cs.CV↗

arXiv2025

A Survey on Human Interaction Motion Generation

Humans inhabit a world defined by interactions -- with other humans, objects, and environments. These interactive movements not only convey our relationships with our surroundings but also demonstrate how we perceive and communicate with the real world. Therefore, replicating these interaction behaviors in digital systems has emerged as an important topic for applications in robotics, virtual reality, and animation. While recent advances in deep generative models and new datasets have accelerated progress in this field, significant challenges remain in modeling the intricate human dynamics and their interactions with entities in the external world. In this survey, we present, for the first time, a comprehensive overview of the literature in human interaction motion generation. We begin by establishing foundational concepts essential for understanding the research background. We then systematically review existing solutions and datasets across three primary interaction tasks -- human-human, human-object, and human-scene interactions -- followed by evaluation metrics. Finally, we discuss open research directions and future opportunities.

cs.CV↗

arXiv2025

GeoRAG: A Question-Answering Approach from a Geographical Perspective

Geographic Question Answering (GeoQA) addresses natural language queries in geographic domains to fulfill complex user demands and improve information retrieval efficiency. Traditional QA systems, however, suffer from limited comprehension, low retrieval accuracy, weak interactivity, and inadequate handling of complex tasks, hindering precise information acquisition. This study presents GeoRAG, a knowledge-enhanced QA framework integrating domain-specific fine-tuning and prompt engineering with Retrieval-Augmented Generation (RAG) technology to enhance geographic knowledge retrieval accuracy and user interaction. The methodology involves four components: (1) A structured geographic knowledge base constructed from 3267 corpora (research papers, monographs, and technical reports), categorized via a multi-agent approach into seven dimensions: semantic understanding, spatial location, geometric morphology, attribute characteristics, feature relationships, evolutionary processes, and operational mechanisms. This yielded 145234 classified entries and 875432 multi-dimensional QA pairs. (2) A multi-label text classifier based on BERT-Base-Chinese, trained to analyze query types through geographic dimension classification. (3) A retrieval evaluator leveraging QA pair data to assess query-document relevance, optimizing retrieval precision. (4) GeoPrompt templates engineered to dynamically integrate user queries with retrieved information, enhancing response quality through dimension-specific prompting. Comparative experiments demonstrate GeoRAG's superior performance over conventional RAG across multiple base models, validating its generalizability. This work advances geographic AI by proposing a novel paradigm for deploying large language models in domain-specific contexts, with implications for improving GeoQA systems scalability and accuracy in real-world applications.

cs.IR↗

arXiv2025

SelfMedHPM: Self Pre-training With Hard Patches Mining Masked Autoencoders For Medical Image Segmentation

In recent years, deep learning methods such as convolutional neural network (CNN) and transformers have made significant progress in CT multi-organ segmentation. However, CT multi-organ segmentation methods based on masked image modeling (MIM) are very limited. There are already methods using MAE for CT multi-organ segmentation task, we believe that the existing methods do not identify the most difficult areas to reconstruct. To this end, we propose a MIM self-training framework with hard patches mining masked autoencoders for CT multi-organ segmentation tasks (selfMedHPM). The method performs ViT self-pretraining on the training set of the target data and introduces an auxiliary loss predictor, which first predicts the patch loss and determines the location of the next mask. SelfMedHPM implementation is better than various competitive methods in abdominal CT multi-organ segmentation and body CT multi-organ segmentation. We have validated the performance of our method on the Multi Atlas Labeling Beyond The Cranial Vault (BTCV) dataset for abdomen mult-organ segmentation and the SinoMed Whole Body (SMWB) dataset for body multi-organ segmentation tasks.

cs.CV↗

arXiv2025

Training-Free Text-Guided Image Editing with Visual Autoregressive Model

Text-guided image editing is an essential task that enables users to modify images through natural language descriptions. Recent advances in diffusion models and rectified flows have significantly improved editing quality, primarily relying on inversion techniques to extract structured noise from input images. However, inaccuracies in inversion can propagate errors, leading to unintended modifications and compromising fidelity. Moreover, even with perfect inversion, the entanglement between textual prompts and image features often results in global changes when only local edits are intended. To address these challenges, we propose a novel text-guided image editing framework based on VAR (Visual AutoRegressive modeling), which eliminates the need for explicit inversion while ensuring precise and controlled modifications. Our method introduces a caching mechanism that stores token indices and probability distributions from the original image, capturing the relationship between the source prompt and the image. Using this cache, we design an adaptive fine-grained masking strategy that dynamically identifies and constrains modifications to relevant regions, preventing unintended changes. A token reassembling approach further refines the editing process, enhancing diversity, fidelity, and control. Our framework operates in a training-free manner and achieves high-fidelity editing with faster inference speeds, processing a 1K resolution image in as fast as 1.2 seconds. Extensive experiments demonstrate that our method achieves performance comparable to, or even surpassing, existing diffusion- and rectified flow-based approaches in both quantitative metrics and visual quality. The code will be released.

cs.CV↗

arXiv2025

Style Quantization for Data-Efficient GAN Training

Under limited data setting, GANs often struggle to navigate and effectively exploit the input latent space. Consequently, images generated from adjacent variables in a sparse input latent space may exhibit significant discrepancies in realism, leading to suboptimal consistency regularization (CR) outcomes. To address this, we propose \textit{SQ-GAN}, a novel approach that enhances CR by introducing a style space quantization scheme. This method transforms the sparse, continuous input latent space into a compact, structured discrete proxy space, allowing each element to correspond to a specific real data point, thereby improving CR performance. Instead of direct quantization, we first map the input latent variables into a less entangled ``style'' space and apply quantization using a learnable codebook. This enables each quantized code to control distinct factors of variation. Additionally, we optimize the optimal transport distance to align the codebook codes with features extracted from the training data by a foundation model, embedding external knowledge into the codebook and establishing a semantically rich vocabulary that properly describes the training dataset. Extensive experiments demonstrate significant improvements in both discriminator robustness and generation quality with our method.

cs.CV↗

arXiv2025

Subtraction of the $t\bar{t}$ contribution in $tW\bar{b}$ production at the one-loop level

The $tW\bar{b}$ production contributes to the real corrections to the $tW$ cross section. It would interfere with the top quark pair production, causing difficulties in a clear definition of the $tW{\bar b}$ events. The subtraction of the $t\bar{t}$ contributions has been performed in the diagram removal or diagram subtraction schemes for the tree-level processes. However, these schemes rely on the ability to identify the double resonant diagrams and thus can not be extended to loop diagrams. We propose a new scheme to subtract the $t\bar{t}$ contributions by power expansion of the squared amplitude in the resonant region. In order to cancel the infra-red divergences of the loop amplitudes, a widely used method is to introduce the dipole counter-terms, an ingredient in calculations of the full next-to-leading order QCD corrections. In our scheme, these counter-terms are also power-expanded. As a proof of principle, we calculate the one-loop correction to the $d\bar{d}\to \bar{b}Wt$ process, and present the invariant mass distribution of the $W\bar{b}$ system.

hep-ph↗

arXiv2025

FRAME: Floor-aligned Representation for Avatar Motion from Egocentric Video

Egocentric motion capture with a head-mounted body-facing stereo camera is crucial for VR and AR applications but presents significant challenges such as heavy occlusions and limited annotated real-world data. Existing methods rely on synthetic pretraining and struggle to generate smooth and accurate predictions in real-world settings, particularly for lower limbs. Our work addresses these limitations by introducing a lightweight VR-based data collection setup with on-board, real-time 6D pose tracking. Using this setup, we collected the most extensive real-world dataset for ego-facing ego-mounted cameras to date in size and motion variability. Effectively integrating this multimodal input -- device pose and camera feeds -- is challenging due to the differing characteristics of each data source. To address this, we propose FRAME, a simple yet effective architecture that combines device pose and camera feeds for state-of-the-art body pose prediction through geometrically sound multimodal integration and can run at 300 FPS on modern hardware. Lastly, we showcase a novel training strategy to enhance the model's generalization capabilities. Our approach exploits the problem's geometric properties, yielding high-quality motion capture free from common artifacts in prior works. Qualitative and quantitative evaluations, along with extensive comparisons, demonstrate the effectiveness of our method. Data, code, and CAD designs will be available at https://vcai.mpi-inf.mpg.de/projects/FRAME/

cs.CV↗

arXiv2025

Higgs boson pair production and decay at NLO in QCD: the $b\bar{b}γγ$ final state

The Higgs boson pair production at the LHC provides a probe to the Higgs boson self-coupling. The higher-order QCD corrections in this process are sizable and must be taken into account in comparison with data. Due to the small cross section, it is necessary to consider at least one of the Higgs bosons decaying to bottom quarks. The QCD corrections to the decay processes would also be important in such cases. We present a full calculation of the total and differential cross sections for the $b\bar{b}γγ$ final state with next-to-leading order (NLO) QCD corrections. After applying typical kinematic cuts in the final state, we find that QCD NLO corrections in the decay decrease the LO result by $19\%$ and reduce the scale uncertainties by a factor of two. The QCD corrections to the invariant mass $m_{jjγγ}$ distribution, the transverse momentum spectra of the leading bottom quark jet and photon are significant and can not be approximated by a constant factor.

hep-ph↗

arXiv2025

Investigating the Limits of Hard X-ray Coherence Length Measurement Employing Young's Double Slit Experiment

Young's double slit experiment has been the most explored technique to gauge the coherence properties of a given system. The limits of this technique in characterizing spatial coherence properties of high emittance, hard x-ray synchrotron sources have been performed at the BXDS-IVU beamline, Canadian Light Source (CLS). High emittance synchrotron sources have always been assumed to possess sub-optimal coherence properties, especially in the hard X-ray regime. While this is largely true, it is very important to understand the limits of coherence for these existing high emittance sources. We have demonstrated that the Young's double slit experiment has harsher limits than what is normally expected in the form of inherent ambiguity. We present data obtained at multiple energies in both the horizontal and vertical direction leading to a thorough understanding of the fundamental limitations of employing Young's double slit experiment to characterize inherently low coherence length systems. We have put forth a novel numerical technique to estimate the source size directly from the Young's double slit interference patterns. With these results, we have demonstrated that CLS has functional coherent beam properties in the hard X-ray regime with Spatial coherence lengths ranging from 5.37$μ\text{m}$ to 17.61$μ\text{m}$ in the horizontal direction. The same in the vertical direction was at least 3 times larger. Finally, we present theoretical calculations showcasing the limits of Young's double slit experiment in characterizing diffraction limited sources.

physics.optics↗

arXiv2025

Phase Stability Analysis of Volume-preserving Algorithms for Accurate Single Particle Orbit Simulations in Tokamak Plasmas

Second-order Volume-preserving algorithms (VPAs) for simulating charged particle motion in electromagnetic fields have been generalized to a rotating angle formulation by using the matrix decomposition methods. Based on this method, the phase stability of this class of VPAs has been analyzed by using the Discrete Fourier Transformations (DFT) technique. It is found that two prominent VPAs, namely the $G_h^2$ and the Boris algorithm, exhibit optimal phase precision for high-frequency (gyro motion) and low-frequency dynamics (transit/bounce motion), respectively. These findings have been empirically verified through numerical experiments. The insights gained from this study enable the selection of an appropriate VPA for practical simulations based on the characteristic frequencies of specific physics problems, which can substantially enhance numerical accuracy and improve computational efficiency for long-term simulations.

physics.plasm-ph↗

arXiv2025

Analytic decay width of the Higgs boson to massive bottom quarks at order $α_s^3$

The Higgs boson decay into bottom quarks is the dominant decay channel contributing to its total decay width, which can be used to measure the bottom quark Yukawa coupling and mass. This decay width has been computed up to $\mathcal{O}(α_s^4)$ for the process induced by the bottom quark Yukawa coupling, assuming massless final states, and the corresponding corrections beyond $\mathcal{O}(α_s^2)$ are found to be less than $0.2\%$. We present an analytical result for the decay into massive bottom quarks at $\mathcal{O}(α_s^3)$ that includes the contribution from the top quark Yukawa coupling induced process. We have made use of the optical theorem, canonical differential equations and the regular basis in the calculation and expressed the result in terms of multiple polylogarithms and elliptic functions. We propose a systematic and unified procedure to derive the $ε$-factorized differential equation for the three-loop kite integral family, which includes the three-loop banana integrals as a sub-sector. We find that the $\mathcal{O}(α_s^3)$ corrections increase the decay width, relative to the result up to $\mathcal{O}(α_s^2)$, by $1\%$ due to the large logarithms $\log^i (m_H^2/m_b^2)$ with $ 1\le i \le 4 $ in the small bottom quark mass limit. The coefficient of the double logarithms is proportional to $C_A-C_F$, which is the typical color structure in the resummation of soft quark contributions at subleading power.

hep-ph↗

arXiv2025

TVDiag: A Task-oriented and View-invariant Failure Diagnosis Framework with Multimodal Data

Microservice-based systems often suffer from reliability issues due to their intricate interactions and expanding scale. With the rapid growth of observability techniques, various methods have been proposed to achieve failure diagnosis, including root cause localization and failure type identification, by leveraging diverse monitoring data such as logs, metrics, or traces. However, traditional failure diagnosis methods that use single-modal data can hardly cover all failure scenarios due to the restricted information. Several failure diagnosis methods have been recently proposed to integrate multimodal data based on deep learning. These methods, however, tend to combine modalities indiscriminately and treat them equally in failure diagnosis, ignoring the relationship between specific modalities and different diagnostic tasks. This oversight hinders the effective utilization of the unique advantages offered by each modality. To address the limitation, we propose \textit{TVDiag}, a multimodal failure diagnosis framework for locating culprit microservice instances and identifying their failure types (e.g., Net-packets Corruption) in microservice-based systems. \textit{TVDiag} employs task-oriented learning to enhance the potential advantages of each modality and establishes cross-modal associations based on contrastive learning to extract view-invariant failure information. Furthermore, we develop a graph-level data augmentation strategy that randomly inactivates the observability of some normal microservice instances during training to mitigate the shortage of training data. Experimental results show that \textit{TVDiag} outperforms state-of-the-art methods in multimodal failure diagnosis, achieving at least a 55.94\% higher $HR@1$ accuracy and over a 4.08\% increase in F1-score across two datasets.

cs.SE↗

arXiv2025

Convex waves grazing convex obstacles to high order

In a recent paper [WW23] we studied the transport of oscillations in solutions to linear and some semilinear second-order hyperbolic boundary problems along rays that graze a convex obstacle to any order. We showed that high frequency exact solutions are well approximated in $H^1$ by much simpler approximate solutions constructed from explicit solutions to profile equations. That result depends on two geometric assumptions, referred to here as the grazing set (GS) and reflected flow map (RFM) assumptions, that are both difficult to verify in general. The GS assumption states that the grazing set, that is, the set of points on the spacetime boundary at which incoming characteristics meet the boundary tangentially, is a codimension two, $C^1$ submanifold of spacetime. The second is that the reflected flow map, which sends points on the spacetime boundary forward in time to points on reflected and grazing rays, is injective and has appropriate regularity properties near the grazing set. In this paper we analyze these assumptions for incoming plane, spherical, and more general ``convex waves" when the governing hyperbolic operator is the wave operator $\Box:=Δ-\partial_t^2$. We prove general results describing when the assumptions hold, and provide explicit examples where the GS assumption fails.

math.AP↗

arXiv2025

Towards Harmless Multimodal Assistants with Blind Preference Optimization

Multimodal Large Language Models (MLLMs) have demonstrated impressive capabilities in multimodal understanding, reasoning, and interaction. Given the extensive applications of MLLMs, the associated safety issues have become increasingly critical. Due to the effectiveness of preference optimization in aligning MLLMs with human preferences, there is an urgent need for safety-related preference data for MLLMs. To address this, we construct the MMSafe-PO preference dataset towards harmless multimodal assistants, featuring multimodal instructions, the conversational format, and ranked paired responses from human feedback. We also identify two insightful observations: modality co-defense and modality cheating, which illustrate that MLLMs possess a certain level of inherent defense while still presenting unique safety challenges. Based on these observations, we propose the Blind Preference Optimization (BPO) approach. Comprehensive experiments on three benchmarks show that BPO effectively enhances the safety capabilities of MLLMs. Notably, BPO significantly improves the safety rate of the base MLLM by 45.0%, outperforming the DPO approach. Additionally, applying BPO to the MMSafe-PO dataset greatly reduces the base MLLM's unsafe rate on other safety benchmarks (14.5% on MM-SafetyBench and 82.9% on HarmEval, demonstrating the effectiveness and robustness of both the dataset and the approach. We release code and data at https://lu-yang666.github.io/MMsafe-PO-Web/.

cs.CL↗

arXiv2025

Minute-cadence observations on Galactic plane with Wide Field Survey Telescope (WFST): Overview, methodology and early results

As the time-domain survey telescope of the highest survey power in the northern hemisphere currently, Wide Field Survey Telescope (WFST) is scheduled to hourly/daily/semi-weekly scan northern sky up to ~23 mag in four optical (ugri) bands. Unlike the observation cadences in the forthcoming regular survey missions, WFST performed "staring" observations toward Galactic plane in a cadence of $\approx$1 minute for a total on-source time of about 13 hours, during the commissioning and pilot observation phases. Such an observation cadence is well applied in producing densely sampling light curves and hunting for stars exhibiting fast stellar variabilities. Here we introduce the primary methodologies in detecting variability, periodicity, and stellar flares among a half million sources from the minute-cadence observations, and present the WFST g-/r-band light curves generated from periodic variable stars and flaring stars. Benefit from high photometric precisions and deep detection limits of WFST, the observations have captured several rare variable stars, such as a variable hot white dwarf (WD) and an ellipsoidal WD binary candidate. By surveying the almost unexplored parameter spaces for variables, WFST will lead to new opportunities in discovering unique variable stars in the northern sky.

astro-ph.SR↗

arXiv2025

Integrating Chain-of-Thought for Multimodal Alignment: A Study on 3D Vision-Language Learning

Chain-of-Thought (CoT) reasoning has proven effective in natural language tasks but remains underexplored in multimodal alignment. This study investigates its integration into 3D vision-language learning by embedding structured reasoning into alignment training. We introduce the 3D-CoT Benchmark, a dataset with hierarchical CoT annotations covering shape recognition, functional inference, and causal reasoning. Through controlled experiments, we compare CoT-structured and standard textual annotations across large reasoning models (LRMs) and large language models (LLMs). Our evaluation employs a dual-layer framework assessing both intermediate reasoning and final inference quality. Extensive experiments demonstrate that CoT significantly improves 3D semantic grounding, with LRMs leveraging CoT more effectively than LLMs. Furthermore, we highlight that annotation structure influences performance-explicit reasoning markers aid LLMs, while unmarked CoT better aligns with LRM inference patterns. Our analyses suggest that CoT is crucial for enhancing multimodal reasoning, with implications beyond 3D tasks. The dataset will be publicly available at https://huggingface.co/datasets/Battam/3D-CoT

cs.CL↗