arXiv subjects

Chunlei Meng

Publications and source records attributed to Chunlei Meng.

35 records · Page 2Linked to original sources

arXiv2026

SpaMEM: Benchmarking Dynamic Spatial Reasoning via Perception-Memory Integration in Embodied Environments

Multimodal large language models (MLLMs) have advanced static visual--spatial reasoning, yet they often fail to preserve long-horizon spatial coherence in embodied settings where beliefs must be continuously revised from egocentric observations under environmental change. We introduce SpaMEM (Spatial Memory from Action Sequences), a large-scale diagnostic benchmark that isolates the mechanics of spatial belief evolution via action-conditioned scene transformations (spawn, place, remove) over long interaction horizons. SpaMEM is built on a physically grounded dataset with 10,601,392 high-fidelity images across four modalities (RGB, depth, instance, semantic segmentation), collected from 25,000+ interaction sequences in 1,000 procedurally generated houses. We formalize embodied spatial reasoning as a three-level hierarchy with 15 diagnostic tasks: Level 1 measures atomic spatial perception from single observations; Level 2 probes temporal reasoning with oracle textual state histories to factor out perceptual noise; and Level 3 requires end-to-end belief maintenance from raw visual streams under the same task dimensions. We further evaluate both short-term (step-wise) updates and long-term (episodic) reconstruction. Benchmarking representative open-source VLM families reveals a consistent stacked bottleneck: coordinate-consistent grounding remains a hard ceiling, and the sharp collapse from Level 2 to Level 3 exposes a pronounced symbolic scaffolding dependency, where models succeed with text-based bookkeeping but struggle to sustain robust visual memory. SpaMEM provides a granular diagnostic standard and motivates explicit mechanisms for state representation, belief revision, and long-horizon episodic integration. A subset of SpaMEM is publicly available at https://huggingface.co/datasets/mill-ct-liao/SpaMEM.

cs.CV↗

arXiv2026

Do LLMs Build World Models From Text? A Multilingual Diagnostic of Spatial Reasoning

Whether large language models (LLMs) construct internal spatial world models from pure-text descriptions remains contested, and whether such capabilities transfer across languages has not been systematically studied. We introduce MentalMap, a multilingual diagnostic benchmark with a six-level capability hierarchy (L0-L5) spanning atomic spatial facts to generative world-graph construction, together with four diagnostic axes probing frame of reference, reading-direction bias, reasoning-effort allocation, and hallucination. MentalMap is built from 100 ProcTHOR household scenes, covers eight typologically diverse languages plus a structured-text control, and contains 39 task families across 1,950 evaluation cells. Evaluating thirteen LLMs across scales and model families, we identify a universal L3 reasoning cliff: no model retains even half of its L0 performance on viewpoint reasoning once baseline atomic accuracy exceeds 40%. The cliff persists across languages, scales, and prompting strategies, while structured-output failures and reasoning patterns vary substantially across models. Human evaluation under the identical pure-text protocol reproduces the same failure pattern, suggesting that the bottleneck arises from text-only working memory constraints rather than being specific to current LLM architectures. Our findings reframe pure-text spatial reasoning as a multi-axis world-modeling problem and motivate multimodal and scratchpad-augmented reasoning as future directions.

cs.AI↗

arXiv2026

Group Cognition Learning: Making Everything Better Through Governed Two-Stage Agents Collaboration

Centralized multimodal learning commonly compresses language, acoustic, and visual signals into a single fused representation for prediction. While effective, this paradigm suffers from two limitations: modality dominance, where optimization gravitates towards the path of least resistance, ignoring weaker but informative modalities, and spurious modality coupling, where models overfit to incidental cross-modal correlations. To address these, we propose Group Cognition Learning (GCL), a governed collaboration paradigm that applies a two-stage protocol after modality-specific encoding. In Stage 1 (Selective Interaction), a Routing Agent proposes directed interaction routes, and an Auditing Agent assigns sample-wise gates to emphasize exchanges that yield positive marginal predictive gain while suppressing redundant coupling. In Stage 2 (Consensus Formation), a Public-Factor Agent maintains an explicit shared factor, and an Aggregation Agent produces the final prediction through contribution-aware weighting while keeping each modality representation as a specialization channel. Extensive experiments on CMU-MOSI, CMU-MOSEI, and MIntRec demonstrate that GCL mitigates dominance and coupling, establishing state-of-the-art results across both regression and classification benchmarks. Analysis experiments further demonstrate the effectiveness of the design.

cs.LG↗

arXiv2026

Disentangling Fact from Sentiment: A Dynamic Conflict-Consensus Framework for Multimodal Fake News Detection

Prevalent multimodal fake news detection relies on consistency-based fusion, yet this paradigm fundamentally misinterprets critical cross-modal discrepancies as noise, leading to over-smoothing, which dilutes critical evidence of fabrication. Mainstream consistency-based fusion inherently minimizes feature discrepancies to align modalities, yet this approach fundamentally fails because it inadvertently smoothes out the subtle cross-modal contradictions that serve as the primary evidence of fabrication. To address this, we propose the Dynamic Conflict-Consensus Framework (DCCF), an inconsistency-seeking paradigm designed to amplify rather than suppress contradictions. First, DCCF decouples inputs into independent Fact and Sentiment spaces to distinguish objective mismatches from emotional dissonance. Second, we employ physics-inspired feature dynamics to iteratively polarize these representations, actively extracting maximally informative conflicts. Finally, a conflict-consensus mechanism standardizes these local discrepancies against the global context for robust deliberative judgment.Extensive experiments conducted on three real world datasets demonstrate that DCCF consistently outperforms state-of-the-art baselines, achieving an average accuracy improvement of 3.52\%.

cs.LG↗

arXiv2026

NeuroSymb-MRG: Differentiable Abductive Reasoning with Active Uncertainty Minimization for Radiology Report Generation

Automatic generation of radiology reports seeks to reduce clinician workload while improving documentation consistency. Existing methods that adopt encoder-decoder or retrieval-augmented pipelines achieve progress in fluency but remain vulnerable to visual-linguistic biases, factual inconsistency, and lack of explicit multi-hop clinical reasoning. We present NeuroSymb-MRG, a unified framework that integrates NeuroSymbolic abductive reasoning with active uncertainty minimization to produce structured, clinically grounded reports. The system maps image features to probabilistic clinical concepts, composes differentiable logic-based reasoning chains, decodes those chains into templated clauses, and refines the textual output via retrieval and constrained language-model editing. An active sampling loop driven by rule-level uncertainty and diversity guides clinician-in-the-loop adjudication and promptbook refinement. Experiments on standard benchmarks demonstrate consistent improvements in factual consistency and standard language metrics compared to representative baselines.

cs.CV↗

arXiv2026

CoDA: Exploring Chain-of-Distribution Attacks and Post-Hoc Token-Space Repair for Medical Vision-Language Models

Medical vision--language models (MVLMs) are increasingly used as perceptual backbones in radiology pipelines and as the visual front end of multimodal assistants, yet their reliability under real clinical workflows remains underexplored. Prior robustness evaluations often assume clean, curated inputs or study isolated corruptions, overlooking routine acquisition, reconstruction, display, and delivery operations that preserve clinical readability while shifting image statistics. To address this gap, we propose CoDA, a chain-of-distribution framework that constructs clinically plausible pipeline shifts by composing acquisition-like shading, reconstruction and display remapping, and delivery and export degradations. Under masked structural-similarity constraints, CoDA jointly optimizes stage compositions and parameters to induce failures while preserving visual plausibility. Across brain MRI, chest X-ray, and abdominal CT, CoDA substantially degrades the zero-shot performance of CLIP-style MVLMs, with chained compositions consistently more damaging than any single stage. We also evaluate multimodal large language models (MLLMs) as technical-authenticity auditors of imaging realism and quality rather than pathology. Proprietary multimodal models show degraded auditing reliability and persistent high-confidence errors on CoDA-shifted samples, while the medical-specific MLLMs we test exhibit clear deficiencies in medical image quality auditing. Finally, we introduce a post-hoc repair strategy based on teacher-guided token-space adaptation with patch-level alignment, which improves accuracy on archived CoDA outputs. Overall, our findings characterize a clinically grounded threat surface for MVLM deployment and show that lightweight alignment improves robustness in deployment.

cs.CV↗

arXiv2026

Tri-Subspaces Disentanglement for Multimodal Sentiment Analysis

Multimodal Sentiment Analysis (MSA) integrates language, visual, and acoustic modalities to infer human sentiment. Most existing methods either focus on globally shared representations or modality-specific features, while overlooking signals that are shared only by certain modality pairs. This limits the expressiveness and discriminative power of multimodal representations. To address this limitation, we propose a Tri-Subspace Disentanglement (TSD) framework that explicitly factorizes features into three complementary subspaces: a common subspace capturing global consistency, submodally-shared subspaces modeling pairwise cross-modal synergies, and private subspaces preserving modality-specific cues. To keep these subspaces pure and independent, we introduce a decoupling supervisor together with structured regularization losses. We further design a Subspace-Aware Cross-Attention (SACA) fusion module that adaptively models and integrates information from the three subspaces to obtain richer and more robust representations. Experiments on CMU-MOSI and CMU-MOSEI demonstrate that TSD achieves state-of-the-art performance across all key metrics, reaching 0.691 MAE on CMU-MOSI and 54.9% ACC-7 on CMU-MOSEI, and also transfers well to multimodal intent recognition tasks. Ablation studies confirm that tri-subspace disentanglement and SACA jointly enhance the modeling of multi-granular cross-modal sentiment cues.

cs.MM↗

arXiv2026

CLCR: Cross-Level Semantic Collaborative Representation for Multimodal Learning

Multimodal learning aims to capture both shared and private information from multiple modalities. However, existing methods that project all modalities into a single latent space for fusion often overlook the asynchronous, multi-level semantic structure of multimodal data. This oversight induces semantic misalignment and error propagation, thereby degrading representation quality. To address this issue, we propose Cross-Level Co-Representation (CLCR), which explicitly organizes each modality's features into a three-level semantic hierarchy and specifies level-wise constraints for cross-modal interactions. First, a semantic hierarchy encoder aligns shallow, mid, and deep features across modalities, establishing a common basis for interaction. And then, at each level, an Intra-Level Co-Exchange Domain (IntraCED) factorizes features into shared and private subspaces and restricts cross-modal attention to the shared subspace via a learnable token budget. This design ensures that only shared semantics are exchanged and prevents leakage from private channels. To integrate information across levels, the Inter-Level Co-Aggregation Domain (InterCAD) synchronizes semantic scales using learned anchors, selectively fuses the shared representations, and gates private cues to form a compact task representation. We further introduce regularization terms to enforce separation of shared and private features and to minimize cross-level interference. Experiments on six benchmarks spanning emotion recognition, event localization, sentiment analysis, and action recognition show that CLCR achieves strong performance and generalizes well across tasks.

cs.CV↗

arXiv2026

ADS-POI: Agentic Spatiotemporal State Decomposition for Next Point-of-Interest Recommendation

Next point-of-interest (POI) recommendation requires modeling user mobility as a spatiotemporal sequence, where different behavioral factors may evolve at different temporal and spatial scales. Most existing methods compress a user's history into a single latent representation, which tends to entangle heterogeneous signals such as routine mobility patterns, short-term intent, and temporal regularities. This entanglement limits the flexibility of state evolution and reduces the model's ability to adapt to diverse decision contexts. We propose ADS-POI, a spatiotemporal state decomposition framework for next POI recommendation. ADS-POI represents a user with multiple parallel evolving latent sub-states, each governed by its own spatiotemporal transition dynamics. These sub-states are selectively aggregated through a context-conditioned mechanism to form the decision state used for prediction. This design enables different behavioral components to evolve at different rates while remaining coordinated under the current spatiotemporal context. Extensive experiments on three real-world benchmark datasets from Foursquare and Gowalla demonstrate that ADS-POI consistently outperforms strong state-of-the-art baselines under a full-ranking evaluation protocol. The results show that decomposing user behavior into multiple spatiotemporally aware states leads to more effective and robust next POI recommendation. Our code is available at https://github.com/YuZhenyuLindy/ADS-POI.git.

cs.IR↗

arXiv2026

Temporal-Spatial Decouple before Act: Disentangled Representation Learning for Multimodal Sentiment Analysis

Multimodal Sentiment Analysis integrates Linguistic, Visual, and Acoustic. Mainstream approaches based on modality-invariant and modality-specific factorization or on complex fusion still rely on spatiotemporal mixed modeling. This ignores spatiotemporal heterogeneity, leading to spatiotemporal information asymmetry and thus limited performance. Hence, we propose TSDA, Temporal-Spatial Decouple before Act, which explicitly decouples each modality into temporal dynamics and spatial structural context before any interaction. For every modality, a temporal encoder and a spatial encoder project signals into separate temporal and spatial body. Factor-Consistent Cross-Modal Alignment then aligns temporal features only with their temporal counterparts across modalities, and spatial features only with their spatial counterparts. Factor specific supervision and decorrelation regularization reduce cross factor leakage while preserving complementarity. A Gated Recouple module subsequently recouples the aligned streams for task. Extensive experiments show that TSDA outperforms baselines. Ablation analysis studies confirm the necessity and interpretability of the design.

cs.CL↗

arXiv2026

DIVER: Dynamic Iterative Visual Evidence Reasoning for Multimodal Fake News Detection

Multimodal fake news detection is crucial for mitigating adversarial misinformation. Existing methods, relying on static fusion or LLMs, face computational redundancy and hallucination risks due to weak visual foundations. To address this, we propose DIVER (Dynamic Iterative Visual Evidence Reasoning), a framework grounded in a progressive, evidence-driven reasoning paradigm. DIVER first establishes a strong text-based baseline through language analysis, leveraging intra-modal consistency to filter unreliable or hallucinated claims. Only when textual evidence is insufficient does the framework introduce visual information, where inter-modal alignment verification adaptively determines whether deeper visual inspection is necessary. For samples exhibiting significant cross-modal semantic discrepancies, DIVER selectively invokes fine-grained visual tools (e.g., OCR and dense captioning) to extract task-relevant evidence, which is iteratively aggregated via uncertainty-aware fusion to refine multimodal reasoning. Experiments on Weibo, Weibo21, and GossipCop demonstrate that DIVER outperforms state-of-the-art baselines by an average of 2.72\%, while optimizing inference efficiency with a reduced latency of 4.12 s.

cs.CV↗

arXiv2026

Pheromone-Focused Ant Colony Optimization algorithm for path planning

Ant Colony Optimization (ACO) is a prominent swarm intelligence algorithm extensively applied to path planning. However, traditional ACO methods often exhibit shortcomings, such as blind search behavior and slow convergence within complex environments. To address these challenges, this paper proposes the Pheromone-Focused Ant Colony Optimization (PFACO) algorithm, which introduces three key strategies to enhance the problem-solving ability of the ant colony. First, the initial pheromone distribution is concentrated in more promising regions based on the Euclidean distances of nodes to the start and end points, balancing the trade-off between exploration and exploitation. Second, promising solutions are reinforced during colony iterations to intensify pheromone deposition along high-quality paths, accelerating convergence while maintaining solution diversity. Third, a forward-looking mechanism is implemented to penalize redundant path turns, promoting smoother and more efficient solutions. These strategies collectively produce the focused pheromones to guide the ant colony's search, which enhances the global optimization capabilities of the PFACO algorithm, significantly improving convergence speed and solution quality across diverse optimization problems. The experimental results demonstrate that PFACO consistently outperforms comparative ACO algorithms in terms of convergence speed and solution quality.

cs.NE↗

arXiv2025

Adversarial Robustness for Unified Multi-Modal Encoders via Efficient Calibration

Recent unified multi-modal encoders align a wide range of modalities into a shared representation space, enabling diverse cross-modal tasks. Despite their impressive capabilities, the robustness of these models under adversarial perturbations remains underexplored, which is a critical concern for safety-sensitive applications. In this work, we present the first comprehensive study of adversarial vulnerability in unified multi-modal encoders. We find that even mild adversarial perturbations lead to substantial performance drops across all modalities. Non-visual inputs, such as audio and point clouds, are especially fragile, while visual inputs like images and videos also degrade significantly. To address this, we propose an efficient adversarial calibration framework that improves robustness across modalities without modifying pretrained encoders or semantic centers, ensuring compatibility with existing foundation models. Our method introduces modality-specific projection heads trained solely on adversarial examples, while keeping the backbone and embeddings frozen. We explore three training objectives: fixed-center cross-entropy, clean-to-adversarial L2 alignment, and clean-adversarial InfoNCE, and we introduce a regularization strategy to ensure modality-consistent alignment under attack. Experiments on six modalities and three Bind-style models show that our method improves adversarial robustness by up to 47.3 percent at epsilon = 4/255, while preserving or even improving clean zero-shot and retrieval performance with less than 1 percent trainable parameters.

cs.CV↗

arXiv2025

ViG3D-UNet: Volumetric Vascular Connectivity-Aware Segmentation via 3D Vision Graph Representation

Accurate vascular segmentation is essential for coronary visualization and the diagnosis of coronary heart disease. This task involves the extraction of sparse tree-like vascular branches from the volumetric space. However, existing methods have faced significant challenges due to discontinuous vascular segmentation and missing endpoints. To address this issue, a 3D vision graph neural network framework, named ViG3D-UNet, was introduced. This method integrates 3D graph representation and aggregation within a U-shaped architecture to facilitate continuous vascular segmentation. The ViG3D module captures volumetric vascular connectivity and topology, while the convolutional module extracts fine vascular details. These two branches are combined through channel attention to form the encoder feature. Subsequently, a paperclip-shaped offset decoder minimizes redundant computations in the sparse feature space and restores the feature map size to match the original input dimensions. To evaluate the effectiveness of the proposed approach for continuous vascular segmentation, evaluations were performed on two public datasets, ASOCA and ImageCAS. The segmentation results show that the ViG3D-UNet surpassed competing methods in maintaining vascular segmentation connectivity while achieving high segmentation accuracy. Our code will be available soon.

eess.IV↗

arXiv2024

CTA-Net: A CNN-Transformer Aggregation Network for Improving Multi-Scale Feature Extraction

Convolutional neural networks (CNNs) and vision transformers (ViTs) have become essential in computer vision for local and global feature extraction. However, aggregating these architectures in existing methods often results in inefficiencies. To address this, the CNN-Transformer Aggregation Network (CTA-Net) was developed. CTA-Net combines CNNs and ViTs, with transformers capturing long-range dependencies and CNNs extracting localized features. This integration enables efficient processing of detailed local and broader contextual information. CTA-Net introduces the Light Weight Multi-Scale Feature Fusion Multi-Head Self-Attention (LMF-MHSA) module for effective multi-scale feature integration with reduced parameters. Additionally, the Reverse Reconstruction CNN-Variants (RRCV) module enhances the embedding of CNNs within the transformer architecture. Extensive experiments on small-scale datasets with fewer than 100,000 samples show that CTA-Net achieves superior performance (TOP-1 Acc 86.76\%), fewer parameters (20.32M), and greater efficiency (FLOPs 2.83B), making it a highly efficient and lightweight solution for visual tasks on small-scale datasets (fewer than 100,000).

cs.CV↗

arXiv2023

A physics-constrained machine learning method for mapping gapless land surface temperature

More accurate, spatio-temporally, and physically consistent LST estimation has been a main interest in Earth system research. Developing physics-driven mechanism models and data-driven machine learning (ML) models are two major paradigms for gapless LST estimation, which have their respective advantages and disadvantages. In this paper, a physics-constrained ML model, which combines the strengths in the mechanism model and ML model, is proposed to generate gapless LST with physical meanings and high accuracy. The hybrid model employs ML as the primary architecture, under which the input variable physical constraints are incorporated to enhance the interpretability and extrapolation ability of the model. Specifically, the light gradient-boosting machine (LGBM) model, which uses only remote sensing data as input, serves as the pure ML model. Physical constraints (PCs) are coupled by further incorporating key Community Land Model (CLM) forcing data (cause) and CLM simulation data (effect) as inputs into the LGBM model. This integration forms the PC-LGBM model, which incorporates surface energy balance (SEB) constraints underlying the data in CLM-LST modeling within a biophysical framework. Compared with a pure physical method and pure ML methods, the PC-LGBM model improves the prediction accuracy and physical interpretability of LST. It also demonstrates a good extrapolation ability for the responses to extreme weather cases, suggesting that the PC-LGBM model enables not only empirical learning from data but also rationally derived from theory. The proposed method represents an innovative way to map accurate and physically interpretable gapless LST, and could provide insights to accelerate knowledge discovery in land surface processes and data mining in geographical parameter estimation.

physics.ao-ph↗

arXiv2021

Generating gapless land surface temperature with a high spatio-temporal resolution by fusing multi-source satellite-observed and model-simulated data

Land surface temperature (LST) is a key parameter when monitoring land surface processes. However, cloud contamination and the tradeoff between the spatial and temporal resolutions greatly impede the access to high-quality thermal infrared (TIR) remote sensing data. Despite the massive efforts made to solve these dilemmas, it is still difficult to generate LST estimates with concurrent spatial completeness and a high spatio-temporal resolution. Land surface models (LSMs) can be used to simulate gapless LST with a high temporal resolution, but this usually comes with a low spatial resolution. In this paper, we present an integrated temperature fusion framework for satellite-observed and LSM-simulated LST data to map gapless LST at a 60-m spatial resolution and half-hourly temporal resolution. The global linear model (GloLM) model and the diurnal land surface temperature cycle (DTC) model are respectively performed as preprocessing steps for sensor and temporal normalization between the different LST data. The Landsat LST, Moderate Resolution Imaging Spectroradiometer (MODIS) LST, and Community Land Model Version 5.0 (CLM 5.0)-simulated LST are then fused using a filter-based spatio-temporal integrated fusion model. Evaluations were implemented in an urban-dominated region (the city of Wuhan in China) and a natural-dominated region (the Heihe River Basin in China), in terms of accuracy, spatial variability, and diurnal temporal dynamics. Results indicate that the fused LST is highly consistent with actual Landsat LST data (in situ LST measurements), in terms of a Pearson correlation coefficient of 0.94 (0.97-0.99), a mean absolute error of 0.71-0.98 K (0.82-3.17 K), and a root-mean-square error of 0.97-1.26 K (1.09-3.97 K).

eess.SP↗