arXiv subjects

Qi Wu

Publications and source records attributed to Qi Wu.

At least 127 records · Page 7Linked to original sources

arXiv2025

Vertical Federated Continual Learning via Evolving Prototype Knowledge

Vertical Federated Learning (VFL) has garnered significant attention as a privacy-preserving machine learning framework for sample-aligned feature federation. However, traditional VFL approaches do not address the challenges of class and feature continual learning, resulting in catastrophic forgetting of knowledge from previous tasks. To address the above challenge, we propose a novel vertical federated continual learning method, named Vertical Federated Continual Learning via Evolving Prototype Knowledge (V-LETO), which primarily facilitates the transfer of knowledge from previous tasks through the evolution of prototypes. Specifically, we propose an evolving prototype knowledge method, enabling the global model to retain both previous and current task knowledge. Furthermore, we introduce a model optimization technique that mitigates the forgetting of previous task knowledge by restricting updates to specific parameters of the local model, thereby enhancing overall performance. Extensive experiments conducted in both CIL and FIL settings demonstrate that our method, V-LETO, outperforms the other state-of-the-art methods. For example, our method outperforms the state-of-the-art method by 10.39% and 35.15% for CIL and FIL tasks, respectively. Our code is available at https://anonymous.4open.science/r/V-LETO-0108/README.md.

cs.LG↗

arXiv2025

Adaptive Prototype Knowledge Transfer for Federated Learning with Mixed Modalities and Heterogeneous Tasks

Multimodal Federated Learning (MFL) with mixed modalities enables unimodal and multimodal clients to collaboratively train models while ensuring clients' privacy. As a representative sample of local data, prototypes offer an approach with low resource consumption and no reliance on prior knowledge for MFL with mixed modalities. However, existing prototype-based MFL methods assume unified labels across clients and identical tasks per client, which is impractical in MFL with mixed modalities. In this work, we propose an Adaptive prototype-based Multimodal Federated Learning (AproMFL) framework for mixed modalities to address the aforementioned issues. Our AproMFL transfers knowledge through adaptively-constructed prototypes without unified labels. Clients adaptively select prototype construction methods in line with labels; server converts client prototypes into unified multimodal prototypes and cluster them to form global prototypes. To address model aggregation issues in task heterogeneity, we develop a client relationship graph-based scheme to dynamically adjust aggregation weights. Furthermore, we propose a global prototype knowledge transfer loss and a global model knowledge transfer loss to enable the transfer of global knowledge to local knowledge. Experimental results show that AproMFL outperforms four baselines on three highly heterogeneous datasets ($α=0.1$) and two heterogeneous tasks, with the optimal results in accuracy and recall being 0.42%~6.09% and 1.6%~3.89% higher than those of FedIoT (FedAvg-based MFL), respectively.

cs.LG↗

arXiv2025

GSCache: Real-Time Radiance Caching for Volume Path Tracing using 3D Gaussian Splatting

Real-time path tracing is rapidly becoming the standard for rendering in entertainment and professional applications. In scientific visualization, volume rendering plays a crucial role in helping researchers analyze and interpret complex 3D data. Recently, photorealistic rendering techniques have gained popularity in scientific visualization, yet they face significant challenges. One of the most prominent issues is slow rendering performance and high pixel variance caused by Monte Carlo integration. In this work, we introduce a novel radiance caching approach for path-traced volume rendering. Our method leverages advances in volumetric scene representation and adapts 3D Gaussian splatting to function as a multi-level, path-space radiance cache. This cache is designed to be trainable on the fly, dynamically adapting to changes in scene parameters such as lighting configurations and transfer functions. By incorporating our cache, we achieve less noisy, higher-quality images without increasing rendering costs. To evaluate our approach, we compare it against a baseline path tracer that supports uniform sampling and next-event estimation and the state-of-the-art for neural radiance caching. Through both quantitative and qualitative analyses, we demonstrate that our path-space radiance cache is a robust solution that is easy to integrate and significantly enhances the rendering quality of volumetric visualization applications while maintaining comparable computational efficiency.

cs.GR↗

arXiv2025

Recursive Visual Imagination and Adaptive Linguistic Grounding for Vision Language Navigation

Vision Language Navigation (VLN) typically requires agents to navigate to specified objects or remote regions in unknown scenes by obeying linguistic commands. Such tasks require organizing historical visual observations for linguistic grounding, which is critical for long-sequence navigational decisions. However, current agents suffer from overly detailed scene representation and ambiguous vision-language alignment, which weaken their comprehension of navigation-friendly high-level scene priors and easily lead to behaviors that violate linguistic commands. To tackle these issues, we propose a navigation policy by recursively summarizing along-the-way visual perceptions, which are adaptively aligned with commands to enhance linguistic grounding. In particular, by structurally modeling historical trajectories as compact neural grids, several Recursive Visual Imagination (RVI) techniques are proposed to motivate agents to focus on the regularity of visual transitions and semantic scene layouts, instead of dealing with misleading geometric details. Then, an Adaptive Linguistic Grounding (ALG) technique is proposed to align the learned situational memories with different linguistic components purposefully. Such fine-grained semantic matching facilitates the accurate anticipation of navigation actions and progress. Our navigation policy outperforms the state-of-the-art methods on the challenging VLN-CE and ObjectNav tasks, showing the superiority of our RVI and ALG techniques for VLN.

cs.CV↗

arXiv2025

Improved bounds on the $H$-rank of a mixed graph in terms of the matching number and fractional matching number

A mixed graph $\widetilde{G}$ is obtained by orienting some edges of a graph $G$, where $G$ is the underlying graph of $\widetilde{G}$. Let $r(\widetilde{G})$ be the $H$-rank of $\widetilde{G}$. Denote by $r(G)$, $κ(G)$, $m(G)$ and $m^{\ast}(G)$ the rank, the number of even cycles, the matching number and the fractional matching number of $G$, respectively. Zhou et al. [Discrete Appl. Math. 313 (2022)] proved that $2m(G)-2κ(G)\leq r(G)\leq 2m(G)+ρ(G)$, where $ρ(G)$ is the largest number of disjoint odd cycles in $G$. We extend their results to the setting of mixed graphs and prove that $2m(G)-2κ(G)\leq r(\widetilde{G}) \leq 2m^{\ast}(G)$ for a mixed graph $\widetilde{G}$. Furthermore, we characterize some classes of mixed graphs with rank $r(\widetilde{G})=2m(G)-2κ(G)$, $r(\widetilde{G})=2m(G)-2κ(G)+1$ and $r(\widetilde{G})=2m^{\ast}(G)$, respectively. Our results also improve those of Chen et al. [Linear Multiliear Algebra. 66 (2018)]. In addition, our results can be applied to signed graphs and oriented graphs in some situations.

math.CO↗

arXiv2025

$B$ meson decays to vector charmonium(like) states and a $K$ meson: the role of final-state interactions

A series of vector charmonium(like) states, accompanied by a $K$ meson, have been observed in the decays of $B$ meson. These processes are color-suppressed at the quark level, as inferred from topological diagram analysis. In this work, we calculate the branching fractions of the decays $B \to ψK$, where $ψ$ denotes the charmonium(like) states $ψ(1S)$, $ψ(2S)$, $ψ(4040)$, $ψ(3770)$, and $ψ(4160)$. Our analysis incorporates both short-distance (naive factorization approach) and long-distance (final-state interactions) contributions. Within reasonable parameters, our results align with experimental data except for the $ ψ(4160)$, suggesting its possible exotic nature. Furthermore, we find that long-distance contributions dominate these decay processes, highlighting the crucial role of final-state interactions in the productions of charmonium(like) states in $B$ decays.

hep-ph↗

arXiv2025

SmartWay: Enhanced Waypoint Prediction and Backtracking for Zero-Shot Vision-and-Language Navigation

Vision-and-Language Navigation (VLN) in continuous environments requires agents to interpret natural language instructions while navigating unconstrained 3D spaces. Existing VLN-CE frameworks rely on a two-stage approach: a waypoint predictor to generate waypoints and a navigator to execute movements. However, current waypoint predictors struggle with spatial awareness, while navigators lack historical reasoning and backtracking capabilities, limiting adaptability. We propose a zero-shot VLN-CE framework integrating an enhanced waypoint predictor with a Multi-modal Large Language Model (MLLM)-based navigator. Our predictor employs a stronger vision encoder, masked cross-attention fusion, and an occupancy-aware loss for better waypoint quality. The navigator incorporates history-aware reasoning and adaptive path planning with backtracking, improving robustness. Experiments on R2R-CE and MP3D benchmarks show our method achieves state-of-the-art (SOTA) performance in zero-shot settings, demonstrating competitive results compared to fully supervised methods. Real-world validation on Turtlebot 4 further highlights its adaptability.

cs.RO↗

arXiv2025

ContentV: Efficient Training of Video Generation Models with Limited Compute

Recent advances in video generation demand increasingly efficient training recipes to mitigate escalating computational costs. In this report, we present ContentV, an 8B-parameter text-to-video model that achieves state-of-the-art performance (85.14 on VBench) after training on 256 x 64GB Neural Processing Units (NPUs) for merely four weeks. ContentV generates diverse, high-quality videos across multiple resolutions and durations from text prompts, enabled by three key innovations: (1) A minimalist architecture that maximizes reuse of pre-trained image generation models for video generation; (2) A systematic multi-stage training strategy leveraging flow matching for enhanced efficiency; and (3) A cost-effective reinforcement learning with human feedback framework that improves generation quality without requiring additional human annotations. All the code and models are available at: https://contentv.github.io.

cs.CV↗

arXiv2025

$Z_{cs}^+$ production in the $B^+$ decays process

In the present work, we studied the $Z_{cs}^+$ exotic state production in the $B^+$ meson decays through the meson loop mechanism, where the $B^+$ meson decays into an anti-charmed meson and a charmed-strange meson pair, and consequently couples to $Z_{cs}$ and a light meson ($ϕ,ρ^0,ω,η,η^\prime$, and $π^0$) via exchanging a proper charmed or charmed-strange meson. Using the effective Lagrangian approach, we estimated the branching fractions of different channels find them to be on the order of $10^{-4}$ and their ratios are almost independent of the model parameter. The fit fractions of $Z_{cs}^+$ in different $B^+$ decay processes are also estimated, and we propose searching for $Z_{cs}^+$ in the $B^+ \to J/ψK^+η^\prime$ process, which should be accessible to both Belle II and LHCb Collaborations.

hep-ph↗

arXiv2025

Navigating Motion Agents in Dynamic and Cluttered Environments through LLM Reasoning

This paper advances motion agents empowered by large language models (LLMs) toward autonomous navigation in dynamic and cluttered environments, significantly surpassing first and recent seminal but limited studies on LLM's spatial reasoning, where movements are restricted in four directions in simple, static environments in the presence of only single agents much less multiple agents. Specifically, we investigate LLMs as spatial reasoners to overcome these limitations by uniformly encoding environments (e.g., real indoor floorplans), agents which can be dynamic obstacles and their paths as discrete tokens akin to language tokens. Our training-free framework supports multi-agent coordination, closed-loop replanning, and dynamic obstacle avoidance without retraining or fine-tuning. We show that LLMs can generalize across agents, tasks, and environments using only text-based interactions, opening new possibilities for semantically grounded, interactive navigation in both simulation and embodied systems.

cs.AI↗

arXiv2025

Evaluating and Advancing Multimodal Large Language Models in Perception Ability Lens

As multimodal large language models (MLLMs) advance rapidly, rigorous evaluation has become essential, providing further guidance for their development. In this work, we focus on a unified and robust evaluation of \textbf{vision perception} abilities, the foundational skill of MLLMs. We find that existing perception benchmarks, each focusing on different question types, domains, and evaluation metrics, introduce significant evaluation variance, complicating comprehensive assessments of perception abilities when relying on any single benchmark. To address this, we introduce \textbf{AbilityLens}, a unified benchmark designed to evaluate MLLMs in six key perception abilities (ranging from counting, OCR, to understanding structural data), focusing on both accuracy and stability, with each ability encompassing diverse types of questions, domains, and metrics. With the assistance of AbilityLens, we: (1) identify the strengths and weaknesses of current main-stream MLLMs, highlighting stability patterns and revealing a notable performance gap between state-of-the-art open-source and closed-source models; (2) uncover interesting ability conflict and early convergence phenomena during MLLM training; (3) reveal the primary reason of ability conflict is data mixing ratio and LLM model size; and (4) discuss the effectiveness of some straightforward strategies \eg, fine-tuning and model merging, to solve the ability conflict. The benchmark and online leaderboard is released in https://github.com/Chenfeng1271/AbilityLens.

cs.CV↗

arXiv2025

Boosting Bot Detection via Heterophily-Aware Representation Learning and Prototype-Guided Cluster Discovery

Detecting social media bots is essential for maintaining the security and trustworthiness of social networks. While contemporary graph-based detection methods demonstrate promising results, their practical application is limited by label reliance and poor generalization capability across diverse communities. Generative Graph Self-Supervised Learning (GSL) presents a promising paradigm to overcome these limitations, yet existing approaches predominantly follow the homophily assumption and fail to capture the global patterns in the graph, which potentially diminishes their effectiveness when facing the challenges of interaction camouflage and distributed deployment in bot detection scenarios. To this end, we propose BotHP, a generative GSL framework tailored to boost graph-based bot detectors through heterophily-aware representation learning and prototype-guided cluster discovery. Specifically, BotHP leverages a dual-encoder architecture, consisting of a graph-aware encoder to capture node commonality and a graph-agnostic encoder to preserve node uniqueness. This enables the simultaneous modeling of both homophily and heterophily, effectively countering the interaction camouflage issue. Additionally, BotHP incorporates a prototype-guided cluster discovery pretext task to model the latent global consistency of bot clusters and identify spatially dispersed yet semantically aligned bot collectives. Extensive experiments on two real-world bot detection benchmarks demonstrate that BotHP consistently boosts graph-based bot detectors, improving detection performance, alleviating label reliance, and enhancing generalization capability.

cs.AI↗

arXiv2025

NavBench: Probing Multimodal Large Language Models for Embodied Navigation

Multimodal Large Language Models (MLLMs) have demonstrated strong generalization in vision-language tasks, yet their ability to understand and act within embodied environments remains underexplored. We present NavBench, a benchmark to evaluate the embodied navigation capabilities of MLLMs under zero-shot settings. NavBench consists of two components: (1) navigation comprehension, assessed through three cognitively grounded tasks including global instruction alignment, temporal progress estimation, and local observation-action reasoning, covering 3,200 question-answer pairs; and (2) step-by-step execution in 432 episodes across 72 indoor scenes, stratified by spatial, cognitive, and execution complexity. To support real-world deployment, we introduce a pipeline that converts MLLMs' outputs into robotic actions. We evaluate both proprietary and open-source models, finding that GPT-4o performs well across tasks, while lighter open-source models succeed in simpler cases. Results also show that models with higher comprehension scores tend to achieve better execution performance. Providing map-based context improves decision accuracy, especially in medium-difficulty scenarios. However, most models struggle with temporal understanding, particularly in estimating progress during navigation, which may pose a key challenge.

cs.CV↗

arXiv2025

LiDARDustX: A LiDAR Dataset for Dusty Unstructured Road Environments

Autonomous driving datasets are essential for validating the progress of intelligent vehicle algorithms, which include localization, perception, and prediction. However, existing datasets are predominantly focused on structured urban environments, which limits the exploration of unstructured and specialized scenarios, particularly those characterized by significant dust levels. This paper introduces the LiDARDustX dataset, which is specifically designed for perception tasks under high-dust conditions, such as those encountered in mining areas. The LiDARDustX dataset consists of 30,000 LiDAR frames captured by six different LiDAR sensors, each accompanied by 3D bounding box annotations and point cloud semantic segmentation. Notably, over 80% of the dataset comprises dust-affected scenes. By utilizing this dataset, we have established a benchmark for evaluating the performance of state-of-the-art 3D detection and segmentation algorithms. Additionally, we have analyzed the impact of dust on perception accuracy and delved into the causes of these effects. The data and further information can be accessed at: https://github.com/vincentweikey/LiDARDustX.

cs.CV↗

arXiv2025

Many Heads Are Better Than One: Improved Scientific Idea Generation by A LLM-Based Multi-Agent System

The rapid advancement of scientific progress requires innovative tools that can accelerate knowledge discovery. Although recent AI methods, particularly large language models (LLMs), have shown promise in tasks such as hypothesis generation and experimental design, they fall short of replicating the collaborative nature of real-world scientific practices, where diverse experts work together in teams to tackle complex problems. To address the limitations, we propose an LLM-based multi-agent system, i.e., Virtual Scientists (VirSci), designed to mimic the teamwork inherent in scientific research. VirSci organizes a team of agents to collaboratively generate, evaluate, and refine research ideas. Through comprehensive experiments, we demonstrate that this multi-agent approach outperforms the state-of-the-art method in producing novel scientific ideas. We further investigate the collaboration mechanisms that contribute to its tendency to produce ideas with higher novelty, offering valuable insights to guide future research and illuminating pathways toward building a robust system for autonomous scientific discovery. The code is available at https://github.com/open-sciencelab/Virtual-Scientists.

cs.AI↗

arXiv2025

H2VU-Benchmark: A Comprehensive Benchmark for Hierarchical Holistic Video Understanding

With the rapid development of multimodal models, the demand for assessing video understanding capabilities has been steadily increasing. However, existing benchmarks for evaluating video understanding exhibit significant limitations in coverage, task diversity, and scene adaptability. These shortcomings hinder the accurate assessment of models' comprehensive video understanding capabilities. To tackle this challenge, we propose a hierarchical and holistic video understanding (H2VU) benchmark designed to evaluate both general video and online streaming video comprehension. This benchmark contributes three key features: Extended video duration: Spanning videos from brief 3-second clips to comprehensive 1.5-hour recordings, thereby bridging the temporal gaps found in current benchmarks. Comprehensive assessment tasks: Beyond traditional perceptual and reasoning tasks, we have introduced modules for countercommonsense comprehension and trajectory state tracking. These additions test the models' deep understanding capabilities beyond mere prior knowledge. Enriched video data: To keep pace with the rapid evolution of current AI agents, we have expanded first-person streaming video datasets. This expansion allows for the exploration of multimodal models' performance in understanding streaming videos from a first-person perspective. Extensive results from H2VU reveal that existing multimodal large language models (MLLMs) possess substantial potential for improvement in our newly proposed evaluation tasks. We expect that H2VU will facilitate advancements in video understanding research by offering a comprehensive and in-depth analysis of MLLMs.

cs.CV↗

arXiv2025

From Cluster to Desktop: A Cache-Accelerated INR framework for Interactive Visualization of Tera-Scale Data

Machine learning has enabled the use of implicit neural representations (INRs) to efficiently compress and reconstruct massive scientific datasets. However, despite advances in fast INR rendering algorithms, INR-based rendering remains computationally expensive, as computing data values from an INR is significantly slower than reading them from GPU memory. This bottleneck currently restricts interactive INR visualization to professional workstations. To address this challenge, we introduce an INR rendering framework accelerated by a scalable, multi-resolution GPU cache capable of efficiently representing tera-scale datasets. By minimizing redundant data queries and prioritizing novel volume regions, our method reduces the number of INR computations per frame, achieving an average 5x speedup over the state-of-the-art INR rendering method while still maintaining high visualization quality. Coupled with existing hardware-accelerated INR compressors, our framework enables scientists to generate and compress massive datasets in situ on high-performance computing platforms and then interactively explore them on consumer-grade hardware post hoc.

cs.GR↗

arXiv2025

BadNAVer: Exploring Jailbreak Attacks On Vision-and-Language Navigation

Multimodal large language models (MLLMs) have recently gained attention for their generalization and reasoning capabilities in Vision-and-Language Navigation (VLN) tasks, leading to the rise of MLLM-driven navigators. However, MLLMs are vulnerable to jailbreak attacks, where crafted prompts bypass safety mechanisms and trigger undesired outputs. In embodied scenarios, such vulnerabilities pose greater risks: unlike plain text models that generate toxic content, embodied agents may interpret malicious instructions as executable commands, potentially leading to real-world harm. In this paper, we present the first systematic jailbreak attack paradigm targeting MLLM-driven navigator. We propose a three-tiered attack framework and construct malicious queries across four intent categories, concatenated with standard navigation instructions. In the Matterport3D simulator, we evaluate navigation agents powered by five MLLMs and report an average attack success rate over 90%. To test real-world feasibility, we replicate the attack on a physical robot. Our results show that even well-crafted prompts can induce harmful actions and intents in MLLMs, posing risks beyond toxic output and potentially leading to physical harm.

cs.RO↗