arXiv subjects

Ling Zhou

Publications and source records attributed to Ling Zhou.

At least 19 recordsLinked to original sources

arXiv2026

EAST: Embodied AI Safety Testing via Systematic Fault Injection in Digital Twins

Task success is not a sufficient measure of safety for embodied artificial intelligence (AI), because robots may complete objectives while violating safety requirements. Existing evaluations often emphasize performance and robustness, but lack a systematic route from documented safety hazards to executable, verifiable tests. We present EAST (Embodied AI Safety Testing), a digital-twin framework that converts literature-grounded safety evidence into fault-injection tests across perception, instructions, environments, actions, and timing. EAST combines an expert-reviewed safety-fault taxonomy and eight executable operators with applicability and validity constraints. Its three-stage workflow generates test specifications, verifies fault realization during simulated execution, and independently assesses safety challenges, violations, task outcomes, and behavioral signatures using simulator-state evidence. The framework preserves execution traces and replayable records, supporting attribution and reproducibility across experiments. We evaluate EAST on six RoboTwin manipulation tasks and two policy architectures, ACT and RDT, against alternative testing strategies under matched budgets. Across five comparative campaigns, EAST targets all 74 admissible task-policy-operator combinations and obtains valid evidence for 63, while achieving safety-challenge exposure rates close to Random-Legal and Rule-Based. Moreover, 82 valid EAST executions complete their tasks despite safety violations, revealing risks concealed by success-oriented evaluation. These findings support systematic, validity-aware, reproducible safety testing beyond task-completion metrics.

cs.SE↗

arXiv2026

Text-Anchored Modality Experts with Uncertainty-Guided Routing for Multimodal Sentiment Analysis

Multimodal sentiment analysis (MSA) aims to understand user-generated videos by integrating multiple modalities. Existing methods often treat modalities as symmetric, overlooking the fact that language provides a dominant semantic prior while nonverbal signals are complementary and noise-prone. To address this, we propose a Text-Anchored Modality Experts with Uncertainty-Guided Routing, termed TAME, that reformulates MSA as an uncertainty-guided refinement paradigm, where the language prior is calibrated and enhanced by selectively trusted nonverbal evidence. Specifically, TAME adopts an asymmetric Mixture-of-Modality-Experts framework, in which textual features act as a fixed and always-activated shared expert, while audio-textual and video-textual cross-modal branches serve as dynamically routed auxiliary experts. Furthermore, we introduce lightweight Prophet Heads to quantify the sample-level predictive uncertainty of each individual modality, and this uncertainty metric is further incorporated into the modality router to realize reliability-aware expert aggregation and fusion. Experiments on CMU-MOSI, CMU-MOSEI, and CH-SIMS datasets demonstrate that TAME achieves state-of-the-art performance on the majority of metrics while reducing parameter count by up to 78.7% and FLOPs by up to 41.2% over the mixture-of-expert baseline.

cs.CV↗

arXiv2026

Persistence and Topological Complexity

Topological complexity is a homotopy invariant that measures the minimal number of continuous rules required for motion planning in a space. In this work, we introduce persistent analogs of topological complexity and its cohomological lower bound, the zero-divisor-cup-length, for persistent topological spaces, and establish their stability. For Vietoris--Rips filtrations of metric spaces, we show that the erosion distances between these persistent invariants are bounded above by twice the Gromov--Hausdorff distance. We develop metric models of classical topological spaces based on cube boundaries with the $\ell_\infty$ metric and their products and wedges, allowing explicit computations of persistent invariants and their distances. These examples show that, in certain cases, persistent topological complexity and zero-divisor-cup-length distinguish spaces that persistent homology, persistent LS-category, and persistent cup-length fail to distinguish.

math.AT↗

arXiv2026

Synergistic Learning in Domain Adaptation with a Known Link: A Minimax Understanding

When can combining source and target data yield much better performance than using either alone? From a minimax viewpoint, writing $R(n,n_T)$ for the minimax risk from $n$ source and $n_T$ target observations, we ask when $R(n,n_T)$ decays strictly faster than both $R(n,0)$ and $R(0,n_T)$. We call this the Synergistic Learning Phenomenon (SLP) and quantify it by the Synergistic Advantage Ratio $\mathrm{SAR}=\min(R(n,0),R(0,n_T))/R(n,n_T)$. We study domain adaptation with a known link: source regression $f$, target regression $g=\mathcal{A}f$ for a known operator $\mathcal{A}$, and possibly different designs. Three links---identity, integration, and differentiation---and the density ratio $ρ=h_T/h_S$ determine whether synergy is possible, at what rate, and in which sample-size window. For the identity link (covariate shift) with a Beta$(a,1)$ source and uniform target, a known necessary condition for SLP is $a\geq2+1/(2β)$, where $β$ is the smoothness of $f$ (Result I). An integral link $g(x)=\int_0^xf(t)\,dt$ can yield the joint rate $(nn_T)^{-1/2}$ even for a bounded ratio (Result II), and under a Beta$(a,1)$ source has a phase diagram with cuts at $a=2$ and $a=3$: integral SLP for $1 3$ (Result III). A derivative link $g=\dot f$ synergizes only under a stronger singularity (Result IV), giving a sharp integration--differentiation asymmetry under a bounded ratio. Across all three links, synergy arises from role separation: each dataset resolves a distinct bottleneck associated with the link and designs. We characterize the sample-size windows for synergy and how SAR varies with $n$ and $n_T$. Simulations support the predicted phase structure, and a mirrored integration--differentiation analysis of real vehicle and smartphone signals illustrates the asymmetry.

math.ST↗

arXiv2026

From Mono to Stereo: Accelerating Binocular Gaussian Splatting via Reprojection and Selective Patching

Binocular rendering requires two nearby views of the same scene and therefore repeats substantial visibility and shading work. We present a 2D Gaussian Splatting (2DGS) pipeline that fully renders a dominant-eye RGB image and an alpha-weighted depth proxy, reprojects that image to the affiliated eye, and repairs uncovered pixels. Small interior gaps are interpolated, whereas larger disoccluded regions are identified as regions of interest (ROIs) and selectively re-rendered. The depth proxy reuses the alpha-blending weights computed during dominant-eye rasterization, avoiding a separate depth-rendering pass. An adaptive ROI generator localizes the required updates using reprojected image boundaries and optional connected center-hole detection. On DTU, Tanks and Temples, and MipNeRF-360, the method reduces the measured time of a sequential two-pass binocular reference by 15.5\% to 28.8\% and peak GPU memory by 6\% to 11\%. The corresponding affiliated-eye quality degradation is at most 1.3 dB PSNR, 0.02 SSIM, and 0.02 LPIPS, representing a measurable trade-off that requires application-specific perceptual validation. These results establish a practical efficiency-quality trade-off for controlled static-scene stereo rendering and motivate future evaluation under continuous motion and on physical VR hardware.

cs.CV↗

arXiv2026

GAPS: Generative Active Pseudo-view Selection for Sparse-View 3D Gaussian Splatting

Novel view synthesis from sparse observations is severely under-constrained. Although 3D Gaussian Splatting (3DGS) enables real-time rendering, it produces floaters, broken geometry, and washed-out backgrounds when trained with few views. We propose an alternating optimization framework that uses a pre-trained image diffusion model to generate geometrically consistent pseudo-views for additional 3DGS supervision. Generation is constrained by depth-conditioned ControlNet, IP-Adapter style transfer, LoRA scene adaptation, and img2img structural anchoring. We introduce Generative Active Pseudo-view Selection (GAPS) to balance reconstruction informativeness and generative reliability when choosing target views. Its annealing schedule shifts from conservative interpolation early in training to exploratory extrapolation later, gradually covering unobserved regions. A dual-criterion admission gate and uncertainty-weighted losses reject unreliable generations, while density-adaptive DropGaussian reduces overfitting in complex scenes. On LLFF with 3/6/9 views, our method improves average PSNR over vanilla 3DGS by 0.40/0.89/0.70 dB. On Mip-NeRF 360 with 12/24 views, the gains are 1.18/0.80 dB. SSIM improves and LPIPS decreases in every setting. Ablations show that active selection and density-adaptive regularization are both necessary; only the full method reduces LPIPS below the no-pseudo-view baseline on unbounded 360-degree scenes.

cs.CV↗

arXiv2026

A$^2$Safe: Counterfactual Evidence-Aligned Adaptive Agent Collaboration for Safe and Effective Visual Question Answering

Visual Question Answering (VQA) with Multimodal Large Language Models (MLLMs) requires not only producing safe and effective responses, but also grounding safety decisions in the multimodal evidence that determines risk. Recent safety-alignment methods improve refusal behavior and contextual risk awareness, yet correct safety outcomes may still rely on superficial textual or visual correlations, particularly when risk emerges from interactions between individually benign image and question content. To address this issue, we propose A$^2$Safe, a counterfactual evidence-aligned adaptive agent collaboration framework for safe and effective VQA. A$^2$Safe organizes localized visual observations, textual intent, and cross-modal risk relations through a Grounded Safety Evidence Board, making the basis of safety decisions explicit. Counterfactual safety evidence alignment enforces invariance to safety-irrelevant changes while requiring appropriate safety-state and response-mode transitions when risk-critical evidence is minimally altered. The resulting evidence state further supports adaptive collaboration, enabling direct answering when grounded evidence is sufficient and invoking policy critique and response revision when evidence is risky, uncertain, or conflicting. Under complementary safety-critical and general VQA protocols, A$^2$Safe achieves a 95.72 SIUO safety score, reduces the benign refusal rate on MOSSBench to 14.67%, and maintains an average general VQA score of 78.34 with 27.8% token overhead. These results support counterfactual evidence-aligned adaptive collaboration for safe and effective multimodal question answering.

cs.CV↗

arXiv2026

FINSKILLOPS: A Self-Evolving Multi-Agent System for SEC Filing QA

Financial QA systems are typically improved before deployment through better retrieval, prompting, or agent coordination, leaving their reliability behavior fixed thereafter. In practice, new SEC-filing questions repeatedly expose heterogeneous errors in period, entity, evidence use, and calculation. Existing self-improvement methods can turn failures into new behaviors, but offer limited control over where a correction should apply or which previously correct answers it may break. We therefore frame post-deployment improvement as controlled behavioral maintenance: recurring failures should become scoped skill patches, and each patch should earn deployment with- out introducing regressions. We instantiate this view in FINSKILLOPS, a multi-agent system for SEC filing QA. FINSKILLOPS derives reusable skills from evidence-grounded, typed failure diagnoses and governs them through targeted validation, protected-case regression checks, negative controls, and versioned replacement or retirement. Across six financial QA benchmarks, a single frozen skill registry achieves the highest verdict-weighted correctness and reference consistency among the evaluated systems. Evolved skills raise correctness from 3.70 to 4.55 on our enhanced benchmark. In a separate 12-round operational study, only six of 33 proposed skills are promoted, while the monitoring non-correct rate falls from 20.0% to 12.5%. These results establish controlled skill scope, admission, and lifecycle management as the foundation for reliable self-improvement.

cs.AI↗

arXiv2026

Persistent cohomology operations and Gromov-Hausdorff estimates

We establish the foundations of the theory of persistent cohomology operations, derive decomposition formulas for wedge sums and products, and prove their Gromov-Hausdorff stability. We use these results to construct pairs of Riemannian pseudomanifolds for which the Gromov-Hausdorff estimates derived from persistent cohomology operations are strictly sharper than those obtained using persistent homology.

math.AT↗

arXiv2026

QIRL: Optimized Question-Image Relation Learning for Bias-Robust Visual Question Answering

Existing bias mitigation methods for Visual Question Answering (VQA), a typical Artificial intelligence application, endure two main limitations. First, they fail to capture the optimal relation between images and texts, as prevailing learning frameworks lack the capacity to extract deep correlations from highly contrasting samples. Second, they overlook assessing Question-Image (QI) relevance during inference, since prior work has not examined the degree of input relevance in debiasing studies. To address these issues, we propose a novel neural network framework termed Optimized Question-Image Relation Learning (QIRL), which provides a reliable implementation of artificial intelligence for VQA tasks and improves the robustness of conventional VQA models through a generation-driven self-supervised learning strategy. Specifically, two modules are introduced. The Negative Image Generation (NIG) module automatically produces highly irrelevant QI pairs during training to strengthen relational learning. In contrast, the Irrelevant Sample Identification (ISI) module enhances model robustness by detecting and filtering out irrelevant inputs, thereby reducing prediction errors. Moreover, to verify the effectiveness of filtering out unrelated QI pairs in mitigating output errors, we propose a specialized metric to evaluate the ISI module's performance. Notably, our approach is model-agnostic and can be seamlessly integrated with various VQA architectures. Extensive experiments on VQA-CPv2 and VQA-v2 datasets demonstrate the effectiveness and generalization ability of our method. Our approach achieves state-of-the-art performance.

cs.CV↗

arXiv2026

TempJail: Temporal Jailbreak Attack against Large Vision-Language Models via Subtitle Scheduling

Large vision-language models (LVLMs) have achieved remarkable progress in video understanding and reasoning. Despite extensive studies on text- and image-based jailbreaks, video jailbreaks against LVLMs remain largely unexplored. Existing video jailbreak methods mainly manipulate textual content embedded in videos, while overlooking how such information is organized over time. Our analysis reveals that jailbreak effectiveness depends not only on the semantics of textual information but also on its temporal presentation, including duration and timing-slot allocation. Motivated by this finding, we use subtitles, which are common in real-world videos and allow semantic content to be presented under precise temporal control without appearing visually intrusive, as a natural attack medium. Based on this insight, we propose TempJail, a black-box video-based jailbreak framework that constructs query-aligned dialogue-style subtitle sequences and optimizes their temporal scheduling to exploit temporal vulnerabilities in LVLMs and elicit responses that satisfy the harmful intent of the source query. Extensive experiments on four representative LVLMs and two datasets demonstrate that TempJail achieves the highest attack success rate across all evaluated model--dataset settings, outperforming the strongest baseline by 53 and 18 percentage points in dataset-averaged ASR on GPT-5 and Gemini 3.5-Flash, respectively.

cs.CV↗

arXiv2026

Vietoris--Rips Contractibility for Dense Subsets of Finite-Dimensional Normed Spaces

Let $Y$ be a $δ$-dense subset of a finite-dimensional normed space $X$, so every point of $X$ is at distance at most $δ$ from some point of $Y$. We prove that the open Vietoris--Rips complex $\operatorname{VR}_{<}(Y;r)$ is contractible whenever $r>δ/(1-J_{\mathrm{fin}}(X))$, where $J_{\mathrm{fin}}(X)$ is the normalized finite Jung constant: the supremum of circumradius divided by diameter over finite subsets of positive diameter. We use the canonical isometric embedding of $X$ into a hyperconvex normed space, where $\operatorname{VR}_{<}(Y;r)$ is the nerve of an equal-radius open-ball cover. The finite Jung estimate restricts the cover to a convex tube about $X$ without changing its nerve, while $δ$-density makes the restricted balls cover the tube. Thus the complex is homotopy equivalent to the tube and hence contractible. Applying this result to lattices gives explicit contractibility bounds. In $(\mathbb{Z}^n,\ell_1)$, gaps in the distance spectrum sharpen the bound: $\operatorname{VR}_{<}(\mathbb{Z}^n,\ell_1;r)$ is contractible above $\max\{n,n(n-1)/2\}$, and $\operatorname{VR}_{\le}(\mathbb{Z}^n,\ell_1;r)$ is contractible at or above the same value. This asymptotically halves the leading term of Zaremsky's previous bound. We also give explicit bounds for $(\mathbb{Z}^n,\ell_p)$ for all $1\le p\le\infty$, recovering the optimal bound $r=1$ for $p=\infty$; for fixed $n$, the bounds converge as $p\to\infty$. For a general locally finite $δ$-dense subset $Y\subset X$ whose distance spectrum may not be discrete, a separate closed-cover argument proves that $\operatorname{VR}_{\le}(Y;r)$ is contractible whenever $r\geδ/(1-J_{\mathrm{fin}}(X))$.

math.AT↗

arXiv2026

Ribonucleic-Acid Protein Interaction Prediction Based on Deep Learning: A Comprehensive Survey

The interaction between Ribonucleic Acids (RNAs) and proteins, also called RNA Protein Interaction (RPI), governs biological processes, including gene regulation and disease pathogenesis. This comprehensive survey examines Artificial Intelligence (AI) applications in Deep Learning-based RPI Prediction (DL-based RPIP) through eight Research Questions (RQs), analyzing 179 studies (2014--2023). The key findings include: sustained technical evolution through embryonic (2014--2017), accelerated (2018--2022), and expansion phases (2023) (RQ1); hybrid models integrating Graph Neural Networks (GNNs) (for topological interface modeling) and Transformers (for long-range dependencies) achieve state-of-the-art performance (RQ4); pretrained language models enhance small-sample learning, but the cross-species generalization declines sharply with evolutionary distance (RQ5). Critical challenges persist, including data heterogeneity across databases, the scarcity of standardized benchmarks (RQ2), and balancing the trade-off between feature encoding and information preservation (RQ3). Future advancements require biologically informed DL architectures, multi-feature fusion, and rigorous cross-validation to bridge the generalization-interpretability gap (RQ8): This would accelerate the clinical translation of predictive tools (RQ6/RQ7). As the first comprehensive analysis spanning feature encoding, modeling, evaluation, applications, and tools, this work fills a critical gap in the DL-based RPIP literature.

q-bio.QM↗

arXiv2026

EGRL: Edge generation-guided relation-aware learning for RNA-protein interaction prediction

RNA-Protein Interactions (RPIs) are critical for regulating cellular functions. While traditional wet-lab experiments for RPI detection are costly and time-consuming, Deep Learning (DL) methods provide an efficient computational alternative for RPI Prediction (RPIP). In particular, Graph Neural Networks (GNNs) are promising, as they naturally model RPI networks. However, existing GNN-based methods often rely on homogeneous graphs or predefined meta-paths, which limit their ability to handle data sparsity and to generalize to cold-start scenarios involving unknown molecules. To address these limitations, we propose Edge Generation-guided Relation-aware Learning (EGRL), a novel framework with several key components: implicit meta-path learning to capture relational semantics without handcrafted paths; a multi-relation-aware attention mechanism for adaptive fusion of interaction patterns; a graph generator that predicts potential ("soft") edges to support cold-start nodes; and a multi-feature fusion predictor for final interaction scoring. EGRL is jointly trained with a primary task loss and an auxiliary generator loss. Comprehensive evaluations on four benchmark datasets demonstrate that EGRL achieves competitive overall performance. More importantly, it exhibits superior generalization in cold-start settings, achieving an Area Under the Receiver Operating Characteristic curve (AUROC) of 0.867 and an Area Under the Precision-Recall curve (AUPR) of 0.861 on unknown molecules, corresponding to improvements of 8.6% in AUROC and 5.0% in AUPR over prior state-of-the-art methods. The code will be released soon.

cs.LG↗

arXiv2026

FinSAgent: Corpus-Aligned Multi-Agent RAG Framework for Evidence-Grounded SEC Filing Question Answering

Financial question answering over U.S. Securities and Exchange Commission (SEC) filings requires retrieving and synthesizing heterogeneous evidence dispersed across long, standardized, and highly redundant disclosures. Existing retrieval-augmented and multi-agent systems typically derive retrieval queries directly from the user's question and rank candidates by semantic similarity. Together, these choices create prior-corpus misalignment: a mismatch between model priors and the target filings' structure, terminology, and evidence standards. As a result, query generation misses corpus-specific evidence, while semantic reranking favors topically similar but evidentially invalid false-positive chunks. We propose FinSAgent, an evidence-grounded multi-agent framework that reframes SEC filing QA as corpus-aligned retrieval planning and corrects both ends with a single principle: inject corpus-side conditioning wherever model priors would otherwise dominate. FinSAgent combines (1) role-specialized agents anchored to the mandated 10-K item structure, (2) database-aware query decomposition that conditions each agent's sub-queries on a lightweight, summary-level view of the local corpus, and (3) multi-path retrieval with a learned feature-gated reranker that separates evidential validity from semantic similarity. Across five offline financial QA benchmarks, FinSAgent improves retrieval coverage and answer correctness over strong single-agent and multi-agent baselines; in a three-arm randomized online experiment with 1,000 anonymous user ratings, it also receives higher scores than baselines.

cs.IR↗

arXiv2026

Liquid Fusion of Heterogeneous Representations Towards General Salient Object Detection

General Salient Object Detection (SOD) aims to identify and segment visually interesting objects from uni-modality or multi-modality scenes, recently advanced by cutting-edge State Space Models (SSMs). However, a critical limitation of current approaches is their neglect of the inherent spectral biases exhibited by different neural network paradigms. By digging to the dataset-level spectral analysis of Convolutional Neural Networks (CNNs) and SSMs, their semantic representations are inherently complementary based on their complementary frequency preferences. Inspired by this, we harmonize heterogeneous representations from SSMs and CNNs to bridge their spectral biases for general salient object detection. To this end, inspired by the dynamic information propagation of Liquid Neural Networks (LNNs), we introduce a liquid fusion to dynamically integrates features from two backbones, including VMamba and ConvNeXt, referred to Liquid Fusion Network (LFNet). Concretely, by treating the continuous VMamba features and ConvNeXt features as evolving states and exogenous stimulus, respectively, LFNet employs a dynamic gating mechanism for content-aware feature aggregation. Crucially, this state-stimulus paradigm enables to scale to multi-modal cues, resulting in flexibility in general SOD. Besides, a Saliency-Guided Upsampling (SGU) operator to propagate the features to the shallow layer, which leverages a spectral-spatial co-design to suppress upsampling artifacts while preserving semantics. Extensive experiments across five diverse tasks (RGB, RGB-D, RGB-T, VSOD, and VDT) demonstrate that LFNet achieves state-of-the-art performance, offering a superior trade-off between detection accuracy and model efficiency. Code has been released at https://github.com/cke520/LFNet.

cs.CV↗

arXiv2026

MetaRA: Metamorphic Robustness Assessment for Multimodal Large Language Model-based Visual Question Answering Systems

Visual Question Answering (VQA), as the representative multimodal task, serves as a key benchmark for evaluating the reasoning capabilities of Multimodal Large Language Models (MLLMs). However, existing evaluations largely rely on static datasets and accuracy-based metrics, which fail to capture robustness, consistency, and generalization. Inspired by Metamorphic Testing (MT), we propose Metamorphic Robustness Assessment (MetaRA), a testing framework that employs Metamorphic Relations (MRs) to systematically probe vulnerabilities in MLLM-based VQA systems. MetaRA generates controlled variations of image-question inputs based on specific MRs and evaluates models across diverse conditions. Applying MetaRA to multiple MLLM-based VQA models across different tasks reveals nuanced failure patterns, including sensitivity to linguistic perturbations, over-reliance on superficial visual cues, and deeper weaknesses in multimodal reasoning. Experimental results demonstrate that MetaRA provides richer diagnostic insights than conventional accuracy metrics, exposing failure modes that remain hidden under standard benchmarks. Overall, this work highlights the need for systematic robustness evaluation in VQA and positions metamorphic assessment as a scalable, model-agnostic approach toward trustworthy multimodal AI.

cs.CV↗

arXiv2026

Enhancing Visual Question Answering with Multimodal LLMs via Chain-of-Question Guided Retrieval-Augmented Generation

With advances in multimodal research and deep learning, Multimodal Large Language Models (MLLMs) have emerged as a powerful paradigm for a wide range of multimodal tasks. As a core problem in vision-language research, Visual Question Answering (VQA) has increasingly employed MLLMs to improve performance, particularly in open-domain settings where external knowledge is essential. In this work, we aim to further enhance retrieval-based VQA by more effectively integrating MLLMs with structured reasoning and knowledge acquisition. We introduce a logical prompting strategy that fuses Chain-of-Thought (CoT) reasoning with Visual Question Decomposition (VQD), termed CoVQD, to guide retrieval toward more accurate and relevant knowledge for MLLM inference. Building on this idea, we propose a new framework, CoVQD-guided RAG (CgRAG), which enables MLLMs to access more comprehensive and coherent external knowledge while benefiting from structured visual-text reasoning guidance, thereby improving generalization and reliability in complex cross-domain VQA scenarios. Extensive experiments on E-VQA, InfoSeek, and OKVQA benchmarks demonstrate the effectiveness of the proposed method.

cs.CV↗