Multimodal

Unmanned aerial vehicles (UAVs) increasingly rely on visual grounding capabilities to localize task-relevant targets from diverse instructions in complex aerial scenes. Existing referring expression comprehension (REC) benchmarks and methods, however, are largely built around text-only queries and single-object outputs, which limits their applicability to practical UAV scenarios involving reference images, multimodal instructions, absent targets, and multiple valid target instances. To address this gap, we introduce \emph{Universal Referring}, a generalized UAV referring task that jointly expands the query modality and the output cardinality. We construct \emph{UniRef-UAV}, a multimodal benchmark that supports text-only, image-only, and text+image queries with modality-dependent target cardinality, where text-only and text+image queries admit no-target, single-target, and multi-target grounding while image-only queries focus on existence-aware single-instance grounding. It also provides in-domain and cross-domain evaluation protocols for visual-query generalization. We further present \emph{UAV-URNet}, a detection-style baseline that maps heterogeneous queries into a shared query space and predicts variable-size target sets through set prediction. Extensive experiments show that UAV-URNet provides a stable and reproducible baseline with more consistent no-target discrimination and a more lightweight, reproducible implementation than large general-purpose MLLMs. Additional domain analysis, query-representation analysis, and ablation studies demonstrate that multimodal queries help reduce visual-query ambiguity and promote a more unified query--target alignment space. The annotations, visual query crops/images, train/validation/test splits, evaluation scripts, and baseline code will be made publicly available to facilitate reproducible research.

RGB-Thermal (RGBT) Video Object Detection (VOD) has gained significant attention because of the limitations of conventional RGB-based VOD methods under challenging conditions, such as low light, heavy fog, and adverse weather, etc. However, spatial misalignment commonly exists between RGBT image pairs. To address this, we propose a Dual-Correlation Hypergraph Network (DCHNet) that captures high-dimensional complementary information by explicitly modeling two types of correlations: temporal correlation across consecutive frames and spatial correlation from cross-modal features. Specifically, we first design a Patch-based Spatial Alignment Module (PSAM) to sequentially align the multimodal features at the local region level. Subsequently, we propose a Dual Hypergraph Fusion Module (DHFM), which constructs temporal and multimodal hypergraphs, respectively, to enhance object characteristic through dual-correlation learning. Furthermore, the field currently lacks a large-scale, scene-diverse benchmark dataset for comprehensive evaluation. Therefore, we construct DVT-VOD1000, a large-scale RGBT VOD dataset containing 1,000 video sequences with 103,464 RGBT image pairs. The dataset covers diverse scenarios, including campuses, parks, traffics, rural areas, night scenes, rainy weather, and snowy weather. Comprehensive experiments on VT-VOD50 and our DVT-VOD1000 demonstrate that DCHNet achieves state-of-the-art detection accuracy. The dataset and source code will be made publicly available on https://github.com/tzz-ahu/ to support academic research.

Fetal electrocardiogram (fECG) and Doppler ultrasound provide complementary views of fetal cardiovascular function: fECG captures electrical activity while Doppler reflects mechanical hemodynamics shaped by factors such as placental resistance and vascular compliance. Understanding the recoverable and unrecoverable Doppler components through reconstruction from fECG offers insight into the relative contributions of electrical versus mechanical factors in fetal circulation, thereby informing clinical decisions. In addition, clinical evidence of maternal-fetal cardiac coupling suggests that maternal cardiovascular dynamics may also inform fetal hemodynamics. To computationally model these relationships, we propose a cross-modal generative framework combining dilated convolutions with cross-modal attention to selectively incorporate maternal ECG and self-attention to capture long-range temporal dependencies. Trained on 885 synchronized fetal/maternal ECG and Doppler envelope segments from 39 pregnancies, our model synthesizes Doppler envelopes with power spectral density mean squared error (PSD MSE) of 49.9 +/- 15.8 dB^2 (51% lower than two-channel baseline) and heart-rate error of 4.71 +/- 0.77 bpm (1.5% better than baseline; negligible relative to the 110-160 bpm physiological range). Cross-modal attention yields a 39% PSD MSE reduction over naive dual-channel concatenation, quantifying the contribution of maternal-fetal coupling. Our proposed framework advances computational modeling of the maternal-fetal cardiovascular system by enabling the synthesis of Doppler envelopes from dual-lead ECG. By analysis of both recoverable and residual Doppler components, this approach enables quantification of the purely mechanical contributions to Doppler waveforms -- those not recoverable from electrical recordings -- ultimately facilitating a more comprehensive fetal assessment.

Diffusion Language Models (DLMs) have recently achieved substantial progress in natural language generation tasks. Recent research demonstrates that adaptive token generation ordering can significantly improve performance in mathematical reasoning and code synthesis applications. In this work, we investigate the optimization of generation order for both text-to-image synthesis and multimodal understanding. We first establish that, unlike structured problems in language generation such as Sudoku puzzles, model logits alone are insufficient for determining optimal generation sequences in text-to-image generation and multimodal understanding. To address this challenge, we introduce a learnable control module trained via Group Relative Policy Optimization (GRPO) to determine the generation order. Our results demonstrate that learning this control block substantially improves both text-to-image alignment and multimodal understanding in DLMs. In particular, it enhances the model's ability to capture fine-grained spatial relationships in generated images while also strengthening performance on multimodal reasoning and comprehension tasks. We evaluate our framework on GenEval, an object-focused benchmark for text-to-image alignment, where it achieves 4.08% relative improvements. In addition, experiments on VLMEvalKit confirm 4.85% relative improvements in multimodal understanding, highlighting the broad effectiveness of our approach.

With the growing adoption of VLMs, DMs, LLMs, and AFMs, these multimodal foundation models can inadvertently encode sensitive, copyrighted, biased, or unsafe cross-modal associations that originate from their training data. Retraining after deletion requests or policy updates is often impractical, and targeted forgetting remains difficult because knowledge is distributed across shared representations. Multimodal unlearning addresses this challenge by enabling selective removal across modalities while retaining overall utility. This survey offers a unified, system-oriented view of multimodal unlearning across vision, language, audio, and video, grounded in recent advances, emerging applications, and open problems. Our taxonomy enables systematic comparison across model architectures and modalities, clarifying trade-offs among deletion strength, retention, efficiency, reversibility, and robustness. This survey highlights open problems and practical considerations to support future research and deployment of multimodal unlearning. We release a curated repository: https://smsnobin77.github.io/Awesome-Multimodal-Unlearning/

Medicine is inherently multimodal, requiring clinicians to synthesize information across diverse data streams. Yet the development of multimodal foundation models is constrained by limited access to large-scale, high-quality clinical data. Although PubMed Central (PMC) offers a complementary source of expert-authored image-text data, existing PMC-derived resources remain limited in fidelity, reproducibility, and clinical validation. We introduce MedPMC, an automated, continuously updatable framework that transforms permissively licensed literature into high-fidelity infrastructure for medical multimodal models. Applied to 6.1 million PMC articles, MedPMC curated 11 million medical image-text pairs. Component evaluations showed strong performance for initial screening (F1 = 93.2), multi-panel figure detection (F1 = 96.5), figure separation (mAP = 89.8), caption separation and alignment (F1 = 81.4; ROUGE-L = 85.3), and medical figure classification (F1 = 96.5). Manual review by five annotators, three with medical training, found 95.3% of MedPMC images medically relevant, versus 19.7% in a prior PMC-derived dataset. Across 26 benchmarks spanning 11 specialties, a MedPMC-trained CLIP-style model improved average zero-shot AUC by 7.1 percentage points over the strongest architecture-matched biomedical CLIP baseline despite using fewer than half as many image-text pairs. As the vision encoder in a multimodal large language model, it improved medical visual question-answering by 1.9 and 16.9 percentage points across two benchmarks. In 10,524 Yale New Haven Health System dermatology photographs, it improved morphology-to-image retrieval Recall@5 by 11.7 percentage points. These findings show that high-fidelity literature curation strengthens medical multimodal foundation models across benchmark and clinical settings. We publicly release the framework, corpus, benchmarks, and pretrained models.

Hallucinations in vision language models (VLMs) are commonly treated as semantic errors, yet they often arise from partial or ambiguous visual evidence. Prior work mainly focuses on detecting or suppressing hallucinations at generation time, leaving the subsequent reasoning stage largely unexplored. In this work, we study Post Hallucination Reasoning (PHR), the stage in which hallucinated semantics enter the model's inference context and influence downstream predictions. To systematically investigate PHR, we introduce HIVE, Hallucination Inference and Verification Engine, an evaluation infrastructure that enables controlled comparisons between faithful and hallucinated captions. Across nine tasks and nine models, we observe structured modality dependent patterns: hallucinated captions often improve accuracy on vision language tasks, while text only tasks exhibit limited or unstable effects. Further analyses show that hallucinated cues broaden semantic coverage and reshape reasoning dynamics while preserving stable inference. These findings highlight that hallucinated semantics may influence downstream reasoning once they enter the model's inference context. Understanding this post hallucination stage is important for improving the reliability and interpretability of multimodal reasoning systems. Code is publicly available at https://github.com/hefengcs/HIVE.

Adaptive volumetric finite element meshing is a critical step in computer-aided engineering and analysis that dictates the computational budget of a given problem. It traditionally requires iterative PDE solvers or heavily supervised, data-driven surrogates trained on large-scale simulation data. While Multimodal Large Language Models (MLLMs) excel in 2D visual tasks, their zero-shot capability to semantically ground regions based on geometric understanding and physics remains an open question. Overall, this study explores a significant question: can the high-level semantic understanding of off-the-shelf MLLMs serve as a viable, zero-shot geometric proxy for finite element mesh refinement? To investigate this, we introduce GReFEM (Geometric Reasoning Enhanced Multimodal LLMs for Finite Element Meshing), a framework that utilizes MLLMs to visually localize stress-critical regions based on physics-guided textual prompts. To bridge the gap between 2D MLLM pre-training and 3D geometries, we introduce orthoViews, a view-selection module that maximizes the observability of key geometric features. We conduct an in-depth empirical evaluation across diverse CAD geometries, loading cases, and SOTA MLLMs, comparing them against a tuned geometric heuristic under a strict, matched refinement budget. Our findings reveal that MLLMs demonstrate robust zero-shot capacity to accurately follow complex spatial-physical instructions, isolating stress-relevant features with higher precision than blind heuristics. By mapping both the successes and current limitations of MLLMs in physical grounding, this study defines the frontier of foundation models as semantic assistants in automated simulation workflows.

Action Quality Assessment (AQA) aims to evaluate how well a person performs a movement, which is essential in applications such as sports scoring, skill assessment, and healthcare. However, unimodal approaches often struggle to capture subtle cues of movement quality in real-world settings. Although multi-modal inputs provide complementary information, existing methods still face two major challenges: heterogeneous modalities often lead to cross-modal misalignment and unstable fusion, and reliable multi-modal annotation is costly, resulting in limited dataset diversity. To address these challenges, we propose DualAlign, a two-stage multi-modal fusion framework with adaptive alignment. The framework first constructs a coherent visual representation by maximizing shared structural information across RGB video, optical flow, and skeleton modalities. Textual semantics are then incorporated after visual stabilization, allowing high-level descriptions to complement rather than distort the underlying visual manifold. To evaluate the framework under realistic multi-modal conditions, we introduce MM--JDM, a movement-quality assessment dataset integrating RGB videos, optical flow, skeleton sequences, and structured text. MM--JDM naturally exhibits modality noise, class imbalance, and label scarcity, making it a challenging benchmark for studying multi-modal fusion and alignment. Extensive experiments show that DualAlign improves average correlation on MM--JDM by 21.16% over the state-of-the-art methods and achieves gains of 3.53% and 5.95% on the RG and Fis-V benchmarks, respectively. DualAlign also remains robust under missing-modality and label-scarce conditions.

Current Vision-Language Models (VLMs) often struggle to handle complex visual tasks that require consistent and fine-grained reasoning. Recent methods aim to train models to facilitate self-reflective reasoning, i.e., reviewing and improving the generated reasoning. However, they require large volumes of annotated data and lack explicit reflective behavior during test time. By contrast, humans perform explicit and efficient self-reflection through mechanisms such as backward prediction, i.e., predicting which current states are likely to precede a given future state. Inspired by neuroscience, this work proposes a novel solution to address these challenges. We first observe and investigate the phenomenon that mainstream VLMs can perform backward prediction, similar to the human brain. A label-free training framework named Brain-inspired Unsupervised Self-reflection (BUS) is proposed to leverage and exploit backward prediction capability to enhance reflective reasoning in complex visual tasks. BUS enables self-verification of reflective reasoning based on backward prediction, providing explicit learning signals under unsupervised conditions. In this way, BUS eliminates reliance on annotated data while improving reasoning performance. Designed as a model-agnostic plug-in, our framework is compatible with popular fine-tuning methods, such as Supervised Fine-Tuning (SFT) and Reinforcement Learning (RL). Initialized from Qwen3-VL-8B, it improves HR-Bench-8K (+8.0%), HR-Bench-4K (+7.7%), V* Bench (+6.3%), and MME-RealWorld-Lite (+5.8%), proving backward prediction is key to advancing reflective reasoning.

In text-to-image in-context learning (T2I-ICL), a model has to infer a latent compositional pattern from fewshot demonstrations for generating a query image. Recent studies show that state-of-the-art multimodal large language models struggle with this setting, particularly due to limited compositional reasoning and sensitivity to prompt construction. In this work, we propose a Tree-of-Thoughts (ToT) reasoning framework for T2I-ICL that introduces a multi-stage reasoning and selection layer that generates, evaluates, and selects among multiple candidate hypotheses before constructing the final prompt for image synthesis. By exploring alternative reasoning branches and selecting a coherent interpretation, the proposed approach mitigates prompt ambiguity and compositional errors. We implement the proposed approach in a complete ToT-T2IICL inference pipeline and evaluate it on the CoBSAT benchmark. Both qualitative and quantitative results show that structured multi-branch reasoning leads to more consistent and semantically aligned image generation compared to baseline and Chain-of-Thought prompting strategies, without any additional training or fine-tuning.

Accurate forecasting of cellular network traffic is essential for network planning, resource allocation, and quality-of-service assurance in modern mobile communication systems. Real-world traffic often exhibits bursty endogenous dynamics and disturbances triggered by external urban events, which makes reliable prediction highly challenging. Most existing spatiotemporal traffic forecasting methods primarily focus on intrinsic traffic patterns or structural relationships within a single modality, and rarely model burst behavior together with exogenous contextual signals. To address this issue, we propose \textbf{MSPF-Net}, a multimodal cellular traffic forecasting framework that integrates external contextual information. Specifically, MSPF-Net consists of a Spatiotemporal-Frequency Traffic Encoder for capturing temporal, spatial, and spectral traffic patterns, a Peak Enhancement Module for extracting burst-aware representations of sudden spikes, a News Context Representation Module for encoding urban news streams into exogenous contextual embeddings, and a Dynamic Fusion Prediction Module for adaptively integrating these heterogeneous signals to generate forecasts. Experiments on the Milano, Trento, and LTE traffic datasets demonstrate that jointly modeling traffic dynamics, burst patterns, and news contextual signals can effectively improve forecasting performance.

Anticipating human motion from an egocentric perspective is fundamental for proactive assistance in AR/VR, human-robot collaboration, and embodied AI. While recent works incorporate language as a semantic prior to reduce the ill-posed nature of egocentric forecasting, they largely neglect the 3D spatial and semantic context that governs how motion unfolds, and treat pose and language prediction as separate inference streams. We introduce Ego3DLM, built on two core principles: accurate motion forecasting requires explicit spatial and semantic understanding of the 3D environment, and pose and language must be predicted holistically in a single pass, since motion is inherently tied to the semantic interpretation of actions being performed. Given three-point tracking, 3D scene features, and egocentric video, Ego3DLM simultaneously decodes past pose, future pose, past narration, and future narration in a single autoregressive pass, grounding predicted poses and descriptions in one another to enforce cross-modal and temporal consistency. We adopt a three-stage training scheme: (1) spatial-semantic scene awareness pretraining; (2) holistic instruction tuning over all four outputs in a single pass; and (3) GRPO-based reinforcement finetuning with intra- and inter-modal rewards that directly optimize pose-language fidelity. Experiments on the Nymeria benchmark demonstrate that Ego3DLM achieves state-of-the-art performance across future motion prediction, past motion tracking, and motion description, showing that 3D scene grounding and holistic cross-modal prediction yield physically plausible and semantically coherent motion forecasts. The project page is available at https://jaewoo97.github.io/Ego3DLM/.

We introduce a new context-enriched, multimodal time series forecasting benchmark, TimesX. TimesX contains a wide selection of high-quality real-world time series with diverse domains and textual contexts obtained from an automated data generation pipeline, which helps address three main issues of existing multimodal forecasting benchmarks: (1) poor generalization due to the small scale and synthetic nature of benchmark data, (2) very limited types of textual contexts in the benchmarks, and (3) an inability to mitigate data leakage in evaluation. We conduct a thorough empirical study of zero-shot multimodal forecasting approaches on TimesX. Our results suggest that many approaches that perform well on existing benchmarks may fail on TimesX. In contrast, simple ensemble methods that leverage rich textual context accompanying time-series can outperform strong baselines on TimesX.

Visual navigation policies built on large pretrained models have so far followed a common recipe: a dedicated visual encoder, a bespoke action head, and training on thousands of hours of cross-embodiment datasets. We ask whether this recipe is necessary. In this paper, we introduce GemNav, a visual robot navigation policy that adapts a frozen Multimodal Large Language Model (MLLM) for short-to-medium horizon waypoint navigation using Low-Rank Adaptation (LoRA) on the language tower alone, with no auxiliary visual encoder and no continuous regression head. Waypoints and categorical navigation signals share a single discrete token vocabulary generated by the language-model head, and a soft-decoded auxiliary loss recovers the metric structure that pure cross-entropy training discards. On a single 8.7-hour open corpus, roughly three orders of magnitude smaller than competing training sets, the policy transfers zero-shot to four physically distinct unseen environments and stops within 0.25-0.42m of the goal across 20 real-world trials covering an open carpark, an obstacle carpark, a long outdoor chemical yard, and an indoor warehouse. Conditioning on short image histories improves offline metrics but yields no robot benefit, pointing to a ceiling on what temporal context adds once pretrained vision features are in place. These results indicate that discrete-token adaptation of frozen MLLMs can provide a data-efficient, deployable alternative for foundation model robot navigation.

The increasing accessibility of artificial intelligence has led to a rapid rise in AI-generated videos, making it more difficult to distinguish between real and manipulated content. Many existing detection methods rely on a single model and often struggle to generalize across different types of deepfakes. In this work, we developed a multimodal deepfake detection system that combines both audio and visual analysis using an ensemble of models. The system includes AASIST for audio-based detection, and EfficientNet, XceptionNet, and MesoNet for analyzing visual features in video frames. The pipeline takes a video as input, separates the audio, and extracts face frames using MTCNN. Each model produces a score indicating the likelihood of the input being fake. These scores are then combined using ensemble strategies, including mean averaging and stacking. Mean fusion provides a simple and stable baseline, while stacking uses a trained meta-model to learn how to combine predictions more effectively. Results show that while individual models perform well on the datasets they were trained on, their performance drops when tested on more diverse datasets. The ensemble approach helps improve overall robustness by combining predictions from multiple models, leading to more consistent performance across different types of deepfakes. This suggests that using both audio and visual information together is a more reliable approach for deepfake detection. Our results highlight generalization to unseen manipulations as the central open challenge, with average accuracy around 70%.

Human memory is reconstructive, not a faithful recording. Current multimodal LLMs (MLLMs) lack this capability: they process images through a frozen visual encoder, produce a one-shot text output, and discard internal representations. We present DoYouRemember, a three-stage architecture introducing reconstructive memory into MLLMs: (1) a VQ-VAE compresses images into discrete visual tokens, (2) a LoRA-fine-tuned LLM jointly attends to visual and text tokens, and (3) a Diffusion Decoder reconstructs images from the LLM's hidden states. On 1,000 3D facial skin texture maps and 99,000 unlabeled facial images, we find that LLM hidden states contain approximately zero recoverable visual information -- the same Decoder producing clear reconstructions from VQ-VAE tokens (pre-LLM) produces pure noise from LLM hidden states (post-LLM), demonstrating that the LLM understands images but does not remember them. Training a shared memory matrix M under backpropagation systematically fails due to gradient cancellation (O(1/sqrt(N)) attenuation). We identify three root causes and show that local EMA updating resolves all three: each image updates only its top-8 slots out of 64, preserving inter-slot diversity. The resulting M (229K parameters, 16x compressed) approaches the VQ upper bound on unseen test images. Scaling to 1,024 slots surpasses it (LPIPS 0.056 vs. 0.071), as M's continuous representation avoids VQ quantization error. We unify these findings under an information-theoretic framework: memory is lossy compression, recall is decompression, and hallucination is an inherent property of lossy decompression rather than a defect.

We formulate computer vision as unified multimodal generation, where heterogeneous visual tasks are expressed in the native text and image generation spaces of a unified multimodal model, without task-specific architectures. Under this formulation, SenseNova-Vision uses natural-language instructions and optional visual prompts to specify tasks, target regions or views, and decoding conventions, and generates responses as text for symbolic outputs, images for dense spatial predictions, or mixed text-and-image outputs for compositional tasks. To support large-scale training, we convert diverse computer vision annotations into instruction-response examples compatible with these generation spaces, resulting in the SenseNova-Vision Corpus, a computer-vision instruction-response corpus spanning text, image, and mixed targets. Starting from an off-the-shelf pretrained unified multimodal model, SenseNova-Vision is trained primarily on this corpus, with auxiliary multimodal data used as a capability-preserving mixture, and requires no task-specific prediction heads or architectural modifications. The resulting model covers a broad range of vision tasks, including detection, OCR, keypoint estimation, segmentation, depth estimation, surface normal prediction, point maps, and camera pose estimation, while supporting language-defined variants that combine category, color, region, and other visual cues. Experiments show that a single unified model can match leading task-specialized systems across structured visual understanding, dense geometric prediction, segmentation, and multi-view visual geometry. These results suggest unified multimodal generation as a scalable route for integrating computer vision capabilities into general-purpose foundation models. The model and corpus are publicly available.

Developing seamless, high-performance, native intelligent full-duplex Spoken Language Models (SLMs) remains a critical challenge and long-standing goal for the speech and NLP community. Despite notable progress, recent endeavors are fundamentally constrained by severe modality interference, which causes substantial knowledge degradation and compromises semantic integrity -- ultimately making full-duplex SLMs feel unnatural and unintelligent. In this paper, through an exhaustive fine-grained analysis of model optimization dynamics, we uncover the root cause of such performance degradation, revealing that modality interference arises from inherent gradient conflicts between acoustic and semantic modeling when the two modalities are forced to share a deep parameter space. Guided by this key insight, we introduce Lychee-FD, a native end-to-end full-duplex framework designed to mitigate modality interference. Importantly, we propose a hierarchical parameter separation strategy that decouples conflicting modalities in deep layers while preserving cross-modality coherence via a dedicated semantic alignment channel. Extensive experiments on multiple full-duplex benchmarks demonstrate that our method significantly advances the state of the art, yielding substantial improvements in both speech intelligence (+7.4% on Spoken QA) and full-duplex interaction fluidity (+28.5% on FullDuplexBench 1.5) without compromising inference efficiency. To the best of our knowledge, this work is the first to achieve two key advances: 1) uncovering and elucidating the root cause of modality interference in full-duplex SLMs, and 2) designing an elegant hierarchical model together with a practical solution for seamless, high-performance, native intelligent full-duplex SLMs.

Multimodal Large Language Models (MLLMs) have demonstrated impressive performance on cross-modal tasks by jointly training on large-scale textual and visual data, where privacy-sensitive examples could be unintentionally encoded, raising concerns about privacy or copyright violation. To this end, Multi-modality Machine Unlearning (MMU) was proposed as a mitigation that can effectively force MLLMs to forget private information. However, the robustness of such unlearning methods is not fully exploited when the model is published and accessible to malicious users. In this paper, we propose a novel adversarial strategy, namely Prompt-Optimized Parameter Shaking (POPS), aiming to recover the supposedly unlearned multi-modality knowledge from the MLLMs. Our method elicits the victim MLLMs to generate potential private examples via prompt-suffix optimization, and then exploits these synthesized outputs to fine-tune the models so they disclose the true private information. The experiments on the different MMU benchmarks reveal substantial weaknesses in the existing MMU algorithms. Our POPS can even achieve a near-complete recovery of supposedly erased sensitive information on the unlearned MLLMs, exposing fundamental vulnerabilities that challenge the foundational robustness of representative MMU-based privacy protections.

Visual counting is a fundamental pillar of multimodal intelligence, requiring a seamless integration of fine-grained grounding and spatial reasoning. While Multimodal Large Language Models (MLLMs) have achieved remarkable success in qualitative scene understanding, their quantitative precision remains a significant bottleneck, often characterized by persistent numerical hallucinations. Existing counting benchmarks primarily focus on basic perception in simplified contexts, failing to capture the complex failure modes that emerge under logical constraints or adversarial conditions. To address these limitations, we introduce HoloCount, a holistic and diagnostically rich benchmark structured around a three-level hierarchical taxonomy. HoloCount evaluates MLLMs across: (1) Semantic Counting, focusing on atomic and property-based enumeration; (2) Analytical Counting, assessing logical composition through spatial and set-based reasoning; and (3) Robustness Testing, probing model integrity against adverse scenarios and grounded counter-priors, such as high-density scenes and linguistic biases. Through an exhaustive evaluation of over 20 state-of-the-art MLLMs, we reveal a critical performance gap: even top-tier models degrade significantly as tasks transition from perception to complex analytical reasoning and adverse scenarios. Our findings provide a systematic landscape of current MLLM counting capabilities and offer a roadmap for developing more grounded and reliable multimodal systems. The dataset is available at https://mm-mvr.github.io/HoloCount/.

Video-to-audio (V2A) generation aims to synthesize realistic audio that is both semantically consistent with and temporally synchronized to a silent video. Despite recent progress, many methods still rely on multi-stage training, resulting in high computational costs and long runtimes, or transform visual input into text to leverage pretrained text-to-audio models, sacrificing fine-grained temporal cues. To overcome these limitations, we propose Flowley, an end-to-end, single-stage training architecture that produces soundtracks by combining visual features with textual prompts. Crucially, we introduce Progressive Soft-masked Cross-Attention, which embeds audio-visual synchronization directly within its attention mechanism, adding zero additional computational cost compared to standard attention layers. We further observe that existing V2A benchmarks lack sound-oriented descriptive captions, which can potentially degrade the quality of the synthesized audio. To remedy this, we propose SoundCap, a plug-and-play pipeline for creating detailed, sound-aware captions that guide the model. Remarkably, without integrating any pretrained audio-visual alignment modules, Flowley achieves state-of-the-art performance on VGGSound across multiple metrics. Moreover, by incorporating SoundCap, we further exceed the performance of the strongest existing close-sourced methods in terms of audio quality in the zero-shot setting.

Current AI safety evaluation and benchmarking frameworks predominantly rely on Western-centric culture-agnostic defaults that mask critical regional laws, socio-linguistic nuances, and cultural taboos, leaving Vision-Language Models (VLMs) vulnerable in global deployments. We introduce Pluralis v0.1: a novel multimodal, multi-regional, and multilingual dataset built from a culture-first perspective. Spanning 6,448 prompts across six Asia-Pacific countries (Bangladesh, India, Korea, Pakistan, Singapore, Taiwan) and eight languages, Pluralis diverges from prior work by natively sourcing localized safety hazards rather than adapting Western datasets. Crucially, it introduces a multimodal evaluation paradigm: user text (e.g., "Should I gift this?") and an image referring to "this" (e.g., a clock) - both innocuous in isolation, but synergistically triggering specific legal or cultural violations. Pluralis disentangles universal safety violations from localized cultural appropriateness, establishing the latter as a first-class evaluation axis. To operationalize this, we present Judge-Pluralis, an agreement-gated LLM-as-a-Judge ensemble trained on examples classified in an empirically derived cultural taxonomy. Observing VLM behavior on a subset of the Pluralis surfaces recurring, locale-specific failure modes such as image misidentifications with downstream harm, missed item-context-locale interactions, and inadequate refusals. These failure modes vary systematically across locales and languages, exposing blind spots that globally averaged metrics conceal. Ultimately, Pluralis is not presented as a solved evaluation framework for cultural alignment, but rather as a first step and catalyst for future innovation. We call upon the research community to utilize this foundation to advance the science of multilingual, multicultural evaluation to better support AI cultural alignment globally.

Following the CMIVQA, MMI-VQA, and M4IVQA challenges in NLPCC 2023--2025, we introduce the Difficulty-Aware Medical Instructional Video Question Answering (DA-MIVQA) shared task for NLPCC 2026. DA-MIVQA extends previous multilingual and multimodal medical video benchmarks by explicitly distinguishing questions according to the type and complexity of evidence required for answering. Specifically, simple questions can often be answered from subtitle-based textual cues, whereas complex questions require visual grounding, procedural understanding, and cross-modal evidence integration. The challenge contains three tracks: Difficulty-Aware Temporal Answer Grounding in Single Video (DA-TAGSV), Difficulty-Aware Video Corpus Retrieval (DA-VCR), and Difficulty-Aware Temporal Answer Grounding in Video Corpus (DA-TAGVC). The dataset is collected from public medical instructional channels, covers diverse scenarios such as first aid, emergency response, rehabilitation, nursing, and general medical education, and is manually verified with difficulty annotations. This paper presents the task motivation, dataset construction, evaluation protocol, participation overview, competition results, and representative systems of DA-MIVQA. DA-MIVQA provides a practical benchmark for evaluating medical instructional video question answering systems under varying textual, visual, temporal, and procedural reasoning requirements.

Music represents a cornerstone of human culture, existing digitally across diverse modalities, including audio, symbolic encodings (e.g., MIDI, MusicXML), and sheet music. Despite the advancement of Multimodal Large Language Models (MLLMs), current music benchmarks face three major limitations. First, large static benchmarks are resource-intensive to evaluate, and it remains unclear how their results transfer to diverse kinds of music beyond those included in the benchmark. Second, benchmarks claiming to measure "music understanding" often fail to require music perception. Third, they do not support systematic performance comparisons across musical modalities. To overcome these issues, we introduce the Music I Care About Meta-Benchmark (MusICA-MetaBench), a framework that automatically derives on-demand benchmarks directly from user-provided data. By leveraging structured symbolic representations (e.g., MusicXML) and our pre-defined question templates, we build multiple-choice question-answer pairs that probe music perception competencies, aligned with music pedagogy, across audio, music notation images, and symbolic files. We demonstrate our framework with the ChoraleBricks dataset, and experimentally determine benchmark sizes that ensure statistically reliable model comparisons for this setup. By comparing against text-only and white-noise baselines, we show our questions do measure music perception. Ultimately, MusICA-MetaBench represents a significant advancement in the cross-modal assessment of music perception for MLLMs. By proposing a dataset-specific benchmarking paradigm, it enables efficient on-demand evaluation of music perception capabilities.

Multimodal large language models can emit localized predictions, bounding boxes for objects and temporal windows for video and audio events, but they hallucinate these regions prolifically. The model's own token log-probabilities are nearly uninformative: they conflate grounding quality with input ambiguity, and coordinate tokens become near-deterministic once the model commits. We propose Multi-Token Localized Attention (MTLA): a training-free, post-hoc score that measures how strongly a prediction's tokens attend to the region they claim. Prior attention-based detectors, which sum attention over the entire input modality and read a single response token, are weaker special cases; we show that summing only within the claimed region and aggregating across all prediction tokens recovers a stronger grounding signal. The same recipe applies almost trivially to other modalities and tasks: object detection in images and temporal localization in video and audio. Across multiple MLLM families and three modalities, MTLA improves hallucination AUROC by +7 to +38 over the best prior training-free baseline. Used as a confidence score for re-ranking, it nearly doubles the zero-shot COCO detection AP of an open-source 8B generalist (from 20.4 to 37.0), narrowing the gap to supervised detectors without any task-specific training.

Training multimodal search agents to perform multi-hop reasoning remains challenging due to a fundamental structural disconnect: existing pipelines construct training data, search environments, and reward signals independently, causing synthesized structural metadata to be discarded, environments to rely on irreproducible external engines, and RL rewards to remain sparse at the trajectory level. We present \textbf{SearchEyes}, which uses a typed knowledge graph as the backbone of a \emph{simulated search world} that unifies all three components. We propose \textbf{Perception-Knowledge Chains (PKC)} to sample constrained multi-hop paths over the visual-knowledge intersection of Wikidata5M, retaining hop-level entity metadata that simultaneously defines a self-contained search world and step-level reward anchors. We further propose \textbf{Hop-Anchored Policy Optimization (HaPO)}, which reuses these anchors for step-level credit assignment without a separately trained process reward model. Experiments on six multimodal knowledge-intensive benchmarks show that SearchEyes achieves state-of-the-art performance among open-source multimodal search agents, with SearchEyes-27B improving over the strongest open-source baseline by 6.2 points on average.%

Automatically recognizing the sentiment, positive or negative, from speech is a challenging task, requiring both the analysis of vocal inflections and the interpretation of uttered words. Recent solutions rely on audio foundation models to solve the task, but it remains unclear if such models can take all aspects into account. To this end, we propose a multimodal solution that integrates audio and text information via cross-modal transformers, where text transcripts are automatically generated via an automatic speech recognition (ASR) tool. Moreover, we create multiple text modalities by automatically translating the transcripts into multiple languages via machine translation tools. Audio and multilingual text features are combined via a cascaded architecture comprising cross-modal transformer blocks that integrate modalities one by one. We further distill knowledge from the multimodal model, called teacher, into a unimodal (audio only) model, called student. We conduct experiments on a large-scale dataset, demonstrating that the automatically generated textual information can bring significant performance boosts in multimodal sentiment polarity classification. Our ablation study confirms that both automatic transcripts and automatic translations are helpful. Moreover, we show that the audio-only model can be enhanced via distillation, boosting performance without any computational overhead during inference. To reproduce the reported results, we publicly release our code at https://github.com/andreidurdun/cross-modal-audio-sentiment.

Recent advancements in Multimodal Large Language Models (MLLMs) have evolved from static perception to interleaved visual-language reasoning, often referred to as ``thinking with images''. A basic operation in this reasoning process is to zoom in on regions of interest (often represented with bounding boxes) to acquire finer visual details. In this paper, we propose \textbf{Seg}mentation before \textbf{Answer}ing (SegAnswer), which shifts the unit of zoom-in from the popular bounding box to pixel-level segmentation mask. By employing fine-grained masks to isolate the target area from cluttered environments, segmented visual input yields a more precise region of interest, effectively filtering out redundant background and interfering objects. Furthermore, the discrete patches of segmented visual input align more seamlessly with how MLLMs structure visual tokens via positional embeddings. In experiments, we evaluate SegAnswer across diverse benchmarks, including high-resolution perception, general perception, and hallucination. It achieves consistent improvements and also exhibits considerable performance on segmentation tasks, validating its capability for reliable pixel grounding.

Autonomous vehicles (AVs) face increasing threats from vandalism-induced occlusion attacks (VOAs) that compromise camera-based perception. While detection frameworks can identify vandalized images, restoring camera-stream utility after physical occlusion remains underexplored. This paper presents present the Recovery and Enhancement of Vandalized Images for Vision Excellence (REVIVE) framework, a vandalism recovery pipeline integrating: (1) binary VOA detection, (2) multi-class VOA pattern identification, (3) EfficientNet-based U-Net segmentation, and (4) type-aware recovery using Bootstrapping Language-Image Pre-training (BLIP)-guided Stable Diffusion inpainting, direct pixel replacement, or adaptive median filtering. Stable Diffusion shows variable reconstruction performance (per-pattern SSIM 0.667-0.867, PSNR 15.4-26.7dB) across VOA patterns, while aligned direct pixel replacement achieves near-identical reconstruction under the aligned-reference condition. On 500 tracked clean/vandalized image pairs, unrecovered VOAs reduce YOLOv8l object-detection recall to 0.588, while direct pixel replacement restores recall to 0.967 and F1-score to 0.970 under that aligned-reference condition. LaMa, Telea, and Navier-Stokes baselines improve image similarity but provide more limited downstream detection recovery, and Stable Diffusion is treated as an asynchronous recovery branch subject to a quality gate rather than a blocking real-time perception step. We evaluate a reference-available quality gate that filters recovered candidates before downstream use: without it, type-aware routing degrades per-image recall to 0.304, whereas with it, recall returns to 0.608, at or above the unrecovered baseline, ensuring the forwarded stream is never worse than the unrecovered frame. REVIVE therefore, provides a structured recovery framework from VOAs in AVs.

Wound monitoring is a critical yet underserved clinical challenge, where timely identification of severe adverse events (SAEs) such as infection, tissue deterioration, and delayed healing can significantly impact patient outcomes. While vision-language models (VLMs) show strong multimodal reasoning, they often lack domain-specific grounding to integrate wound imagery with heterogeneous clinical information, and provide limited mechanisms for detecting cases that diverge from the training distribution. We present a multimodal framework for automated wound monitoring and SAE detection. Our approach leverages paired clinical notes and wound descriptions capturing visual characteristics such as appearance, surrounding skin condition, color changes, and signs of inflammation or healing progression, encoded through a dual-stream Low-Rank Adaptation (LoRA) framework built on a frozen BiomedCLIP backbone. We introduce a cross-contextual LoRA fusion mechanism enabling information exchange between clinical semantics and visual wound descriptors, producing context-aware multimodal representations without full model fine-tuning. To identify personalized SAEs, we propose a wound-specific out-of-distribution (OOD) detection framework combining semantic matching, visual typicality, caption-text alignment, and caption-visual alignment into a unified SAE (OOD) score. To capture healing dynamics, we incorporate covariate consistency and temporal drift penalties that leverage changes in wound characteristics across visits. Experiments on a longitudinal wound dataset collected through clinical visits show promising performance on both wound healing assessment and SAE detection, highlighting the potential of semantically enriched, temporally aware vision-language systems for clinical wound monitoring and early risk identification.

Recommendation systems have undergone significant transformations in the past years. The transition from traditional feature interaction modules to generative next-action prediction has pushed the boundaries of personalized content. Developments have largely evolved along two separate tracks. Sequence modeling approaches on the one hand and feature interaction methods on the other. In this paper, we introduce Bumblebee, a recommendation architecture that addresses the lack of interaction between the two directions through an interleaved, stackable block design. Each block implements a micro-pipeline of layers combining sequence personalization, attention-based encoding, and feature crossing into a self-contained unit. Every block produces a joint representation of both feature modalities which is consumed by the next block in the sequence. This mechanism encourages early and repeated mixture of modalities and enriches downstream features with additional contextual information. Residual connections between blocks create cross-modal information pathways and yield additional predictive performance without adding additional parameters. Blocks can be specialized by selectively dropping components, enabling flexible trade-offs between quality and throughput. We evaluate our approach on large-scale industrial data and show consistent improvements over comparable baseline models across several classification and regression tasks. Furthermore, we conduct ablation studies to confirm that the interleaved composition itself is the primary driver of these improvements. Our results suggest that interleaving heterogeneous functional units, rather than composing deep stacks, is a promising paradigm for future-generation recommendation architectures.

Agentic video understanding equips models with long-term memory to autonomously process and respond to continuous, long-horizon multimodal streams. However, advanced video agents often rely on ``detective-style'' iterative reasoning for action control (e.g., $\mathtt{search}$) and evidence aggregation, incurring prohibitive costs and latency. We argue that such heavy reasoning primarily compensates for the lack of global context and semantic misalignment in retrieval. This paper introduces Light-Omni, a multimodal agent framework for reflexive and lightweight video understanding. It achieves this through dual contextual states that instantly build the required context in a single forward pass. First, we maintain a global state, a finite-sized multimodal script continuously consolidated from episodic memory, serving as the global context for Light-Omni. Through hierarchical merging, it preserves recent details while summarizing past events. Second, conditioned on this global context, we generate a parametric latent state that directly drives autonomous actions and produces retrieval embeddings, with minimal latency. Benefiting from this coupled design, Light-Omni achieves semantically aligned retrieval and reflexive responses while avoiding iterative reasoning. Extensive experiments validate the effectiveness of Light-Omni across multiple video benchmarks. Notably, it outperforms M3-Agent with an average 2.4% accuracy gain, a 12.1$\times$ speedup, and a 2.6$\times$ improvement in GPU memory efficiency. Furthermore, it serves as a memory system to enhance both the performance and efficiency of existing MLLMs. Project page: https://clare-nie.github.io/Light-Omni.

Model editing promises a fast, targeted way to correct post-deployment mistakes in medical vision-language models (VLMs) without costly retraining. However, existing multimodal model editing benchmarks focus on general-purpose tasks and do not reflect realistic clinical domain requirements and variability. To address this, we introduce M3Bench, a clinically grounded benchmark for multimodal model editing that evaluates whether an edit remains reliable, precise, and generalizable under the challenges of image and text variation, modality and protocol shifts, clinical knowledge composition, and temporal progression. M3Bench contains 16,276 questions spanning diverse anatomy, modalities, and specialties, and supports both single and sequential edits. By evaluating 4 representative editors across 6 medical and general VLMs, we find that no method excels across all criteria. Gradient-based editors achieve strong transfer but suffer from catastrophic locality violations, whereas memory-based methods preserve locality but lack compositional generality and exhibit high backbone-dependent hyperparameter sensitivity. We further attribute these failures to the latent space geometry of VLMs and how different editing methods shift its landscape. Overall, M3Bench establishes a rigorous clinical stress test for multimodal model editing and offers actionable guidance for safer post-deployment adaptation. The benchmark is publicly available at https://github.com/BioMed-AI-Lab-U-Michgan/M3Bench .

Charts and images appear together throughout scientific publications, yet most computational work does not characterize their coherence. We argue that a chart, its accompanying image, and the caption that links them form a multimodal unit, and that the inferential work required to read it varies systematically. To capture this variation, we develop a typology of reasoning gaps, R1 through R5, that characterizes how chart, image, and text jointly convey a scientific claim, and the interpretive work this demands of the reader. Some pairs restate the same data, while in other pairs, charts are used to quantify a structure the image localizes, project image content onto an external variable, audit an image-based claim, or jointly construct a frame that neither panel can establish alone. The typology is anchored in the grounding theory of communication and was derived bottom-up, with a neuroscience expert, from a corpus of 79 traumatic brain injury papers and 32 chart-image pairs. Crucially, the levels provide a systematic mechanism for identifying where grounding succeeds or breaks down, rather than leaving it to subjective inference. We show this in a study in which a domain expert and three non-experts judge vision-language model (VLM) descriptions of 25 pairs: the level predicts where their judgments align and where they diverge, isolating the points at which contextual knowledge, not the figure, carries coherence. This typology thus offers figure designers a systematic way to balance text against chart-image pairs, bridging the expert-to-non-expert divide in reading a scientific takeaway.

Video-text temporal localization requires precise alignment between natural language queries and corresponding video segments, a fundamental challenge in multimodal understanding. We present a novel framework that addresses two critical limitations of existing methods: inadequate modeling of hierarchical temporal structure and inability to handle complex many-to-many correspondences between modalities. Our approach introduces a multi-scale temporal convolutional encoder that captures motion patterns across different temporal granularities - from instantaneous frame transitions to extended action sequences. We further propose a capsule-based dynamic routing mechanism that iteratively refines segment-query associations through structured agreement updates, enabling flexible modeling of non-monotonic alignments. These components are unified through a multi-task learning objective that jointly optimizes temporal boundary regression, cross-modal semantic alignment, and capsule diversity. Extensive experiments on ActivityNet Captions demonstrate significant improvements, achieving 42.9% Recall@0.5 and 41.1% mean IoU, surpassing strong transformer-based baselines while maintaining computational efficiency. Our results validate that combining hierarchical temporal modeling with structured semantic routing provides an effective solution for fine-grained video-language understanding.

In multimodal classification, late-fusion approaches classify concatenated modality-specific features extracted by unimodal neural networks. When modality imbalance is pronounced, various regularization techniques have been proposed to balance the learning process and overcome the inferior performance of late-fusion networks. In contrast, this work demonstrates that multimodal data can be effectively classified without any explicit modality fusion, using deep ensembles of unimodal networks. We systematically compare deep ensembles to late-fusion networks at equal parameter count and show that ensembles consistently outperform state-of-the-art late-fusion methods designed to address modality imbalance. This advantage also holds over intermediate-fusion techniques we evaluated and over hybrid methods that combine unimodal and multimodal predictions. We propose and empirically validate a method for selecting the number of models per modality in an ensemble, avoiding computationally expensive exhaustive search. Under extreme modality imbalance and small ensemble sizes, the heuristic indicates that ensembles of unimodal models trained solely on the stronger modality are preferable; as the ensemble scales up, incorporating models from the weaker modality becomes beneficial. Both predictions align with our empirical findings. To systematically explore the challenges of optimizing multimodal models, we propose a synthetic multimodal framework that allows control over both the number of modalities and their predictive strength; our findings are consistent across synthetic and real-world datasets. Finally, by fitting scaling laws to bimodal datasets, we estimate the asymptotic performance of ensembles.

Supervised fine-tuning (SFT) is the standard approach for adapting pretrained language models to downstream domains, yet it often improves target-domain behavior at the cost of degrading pre-existing capabilities. Standard cross-entropy fine-tuning promotes only the observed label token and leaves unconstrained how probability mass is redistributed over other plausible alternatives, potentially distorting the rich local preference structure learned during pretraining. We first analyze next-token predictions using Shannon and Renyi entropies, revealing that pretrained models exhibit a regular multimodal entropy structure. These entropy peaks correspond to varying numbers of plausible alternatives, indicating that the base model intrinsically encodes rich distributional knowledge beyond the single supervised token. Motivated by this observation, we propose LP-SFT, a Local-Preserving Supervised Fine-Tuning objective designed to explicitly protect this inherent entropy structure. At each step, LP-SFT constructs an adaptive support of alternative tokens and applies a locally normalized preservation loss to maintain the base model's relative structure among them, while standard cross-entropy independently optimizes the supervised token. Across mixed-domain and single-domain fine-tuning experiments, LP-SFT improves overall performance over vanilla SFT and recent SFT-enhancement baselines, achieving the best balance between pass@1 accuracy and pass@k performance. These results suggest that local preservation helps mitigate capability degradation without collapsing sampling-accessible diversity.

With the proliferation of AI-generated content, sophisticated multimedia manipulation has raised critical concerns about malicious applications such as opinion manipulation and evidence fabrication, making Audio-Visual Temporal Forgery Localization (AV-TFL) an urgent research frontier. Existing TFL methods have progressed along two main paradigms: Transformer-based temporal modeling and channel-wise multimodal fusion. While these approaches capture temporal dependencies and cross-modal correlations, they process all frequency components indiscriminately, leading to overfitting on high-frequency noise and limited robustness under real-world data degradation. Through systematic frequency domain analysis, we find that forgery-discriminative patterns concentrate in the low/mid-frequency range (normalized frequency 0-0.15), while high-frequency components primarily introduce noise, removing them even improves detection performance by +1.4%. Based on this phenomenon, we propose UniSkip-Mamba, a frequency-aware State Space Model framework that incorporates Unified Multimodal Sequence Fusion to preserve cross-modal phase relationships, and Skip-Scanning Mamba Blocks that implement frequency-aware regularization through a novel Group-Scan-Merge mechanism, naturally biasing learning toward discriminative low/mid-frequency patterns (0-0.15) while maintaining representational completeness. We achieve state-of-the-art (SOTA) performance: 63.4% AP@0.95 on LAV-DF (+9.8% improvement) and 63.58% mAP on AV-Deepfake1M (+14.32% improvement), with 6x faster inference. Our frequency-domain analysis provides theoretical justification from a signal processing perspective for why skip-scanning inherently improves both accuracy and robustness.

The rapid proliferation of artificial intelligence-generated content necessitates reliable multimodal forensics. Beyond video-level binary classification, precisely localizing sparsely distributed forged segments in long-form videos remains a critical challenge. This task is particularly difficult when manipulations are subtly embedded and cross-modal signals are weak and temporally diffuse. To address these challenges, we propose EVAS, an end-to-end multimodal framework for temporal forgery localization. At its core, a Multi-Stage Audio-Visual Synergy mechanism facilitates progressive cross-modal interaction to learn deep multimodal forensic representations and capture high-order semantic traces of sparse manipulations. Furthermore, we introduce a Boundary-Aware Refinement strategy to achieve steered boundary calibration. By incorporating invalid-frame masking, this strategy suppresses ambiguous regions and sharpens transition predictions. We adopt a decoupled training paradigm with auxiliary heads to disentangle representation learning from inference objectives, enhancing model generalization and stability. Additionally, a lightweight HourglassFFN is incorporated to reduce computational overhead. Extensive experiments demonstrate that EVAS achieves state-of-the-art average localization accuracy and average recall across three benchmark datasets, validating its effectiveness for fine-grained temporal forgery localization.

Unified Multimodal Models (UMMs) integrate image understanding and generation within a single architecture, yet how the two tasks interact remains understudied. We investigate $\boldsymbol{\mathsf{transferability}}$ in UMMs: whether training a capability on one task improves the same capability on the other without explicit supervision. Through controlled experiments, we empirically find that transferability depends on architecture-models with fully shared transformer backbone and a unified visual encoder exhibit consistent cross-task transfer, while loosely coupled designs show little or none. Leveraging this transferability, we propose a practical training strategy. The most straightforward way to improve a target generative capability (e.g., counting) is to fine-tune generation directly, but this can degrade visual quality due to distribution shift. Instead, we train the corresponding understanding task and let it transfer into generation, which improves capability-specific generative performance while minimizing distribution shift. We validate this across three capabilities-counting, spatial relation, and text recognition/generation-showing that cross-task transferability can be systematically exploited in UMMs.

Large generative models across text-to-text, text-to-image, and image-to-text modalities have been shown to pose significant privacy risks. One fundamental threat is membership inference attacks (MIA), which aim to determine whether a given data point was used in a model's training set. Although prior work has investigated MIAs against these three classes of generative models, existing approaches treat them in isolation and are not cross-applicable, thereby limiting their real-world utility. To address this limitation, we present the first comprehensive study of a unified membership inference framework that applies across text-to-text, text-to-image, and image-to-text modalities. Our approach is grounded in a key modality-agnostic observation: the output distribution of a generative model can approximate its training data distribution. Leveraging this property, we model the distributions of model-generated outputs and auxiliary non-member samples in a shared embedding space, and perform membership inference via likelihood ratio testing. We conduct extensive experiments in a strict black-box setting under both partial-knowledge and zero-knowledge threat models, and evaluate membership inference against both fine-tuning and pre-training data. Experimental results demonstrate our approach's superior performance in comparison to existing state-of-the-art methods, which are typically optimized for a single model class.

Recent Multi-modal Large Language Models (MLLMs) have demonstrated remarkable performance on 2D question answering tasks. However, extending these models to the 3D question answering remains challenging, as they typically require multiple views of the scene, which incurs substantial computational cost at inference. To mitigate this issue, existing solutions rely on strategic frame selection or token-merging algorithms that require preprocessing in advance all frames of the scene, i.e., an offline fashion. In contrast, we propose the first online token-pruning method that can be integrated seamlessly with current MLLM models for 3D question answering tasks, without additional training and with lower memory usage.Our key insight is to project each input frame into a shared voxel space using depth information and camera pose, identifying spatially-overlapped regions across frames and selectively pruning redundant image tokens before they enter the language model. Our method enables efficient online processing while reducing up to 50% of token usage. We apply this approach to Qwen2.5-VL-7B and Qwen3-VL-8B, demonstrating improved performance on the ScanQA, SQA3D, and OpenEQA-HM3D benchmarks.

Modern language models generally represent text as sequences of discrete token embeddings, an assumption deeply rooted in current practice but rarely questioned. We challenge this representation, especially for Chinese, by replacing index-based token embeddings entirely with a single rasterized image of the character sequence, processed by a vision encoder composed of a shared ResNet and a shallow Vision Transformer. To isolate the role of input representation, we construct a dual-branch controlled framework in which both a Vision-based model and an index-based baseline share an identical decoder backbone, training objective, optimizer, and data curriculum. Any performance difference is therefore attributable to the input modality only. Across all tested decoder backbones, the Vision-based model consistently outperforms the baseline, reaching a peak accuracy of 0.429 versus 0.355 for the index-based baseline,that is, a 21% relative improvement, while converging in about half the number of training epochs. The advantage emerges especially within the first five epochs (under 21% of total data) and persists under moderate character corruption: the corrupted Vision model matches the clean index-based baseline. Ablation studies reveal that the advantage requires both spatially coherent input and a ViT encoder with 2D positional encodings. A cross-script comparison on English shows the advantage does not transfer directly to alphabetic writing systems, suggesting that the uniform visual density and radical structure of Chinese characters are enabling conditions. These findings suggest that transformers are more modality-agnostic than commonly assumed, and that discrete tokenization is not a fundamental requirement for Chinese language modeling.

Diabetic retinopathy (DR) is a leading cause of vision impairment worldwide, highlighting the need for accurate and accessible screening tools. Optical Coherence Tomography (OCT) provides high-resolution structural information of the retina, whereas OCT angiography (OCTA) offers complementary vascular information that is highly relevant for DR diagnosis. In this study, we propose a cross-modal fusion of OCT B-scans with single-channel en face OCTA using a bidirectional cross-modal attention network for automated DR classification. Two independent datasets, OCT500 and UIC, comprising 730 subjects in total, were utilized to evaluate performance under within-dataset, combined-dataset, and cross-dataset generalization settings. A ConvNeXt V2 model trained solely on OCT images served as the unimodal baseline. In addition to ground-truth (GT) OCTA, we explored the use of translated (TR) OCTA generated from OCT scans, eliminating the requirement for dedicated OCTA hardware. Experimental results demonstrate that cross-modal fusion consistently outperforms unimodal OCT classification across all evaluation scenarios. Fusion with GT OCTA improved classification accuracy and discriminative performance, while TR OCTA achieved comparable or superior results in most settings. Furthermore, TR OCTA improved sensitivity and cross-dataset generalization, indicating enhanced robustness to domain shifts. These findings demonstrate that attention-based OCT-OCTA en face fusion provides clinically meaningful improvements for DR detection and suggest that computationally generated OCTA can serve as a practical, low-cost alternative to hardware-acquired OCTA, enabling broader deployment of high-performance retinal screening systems in resource-limited clinical environments.

Audio-Visual Question Answering (AVQA) extends classical VQA by requiring joint reasoning over video and synchronized audio. However, many AVQA systems rely on deeply stacked layers of self- and cross attention across text, video, and audio. Such sequential stacking may incur loss of information such as subtle inter-modal cues over the layers, causing errors to accumulate across sequential attention layers during the fusion. We introduce Q-TriM which performs multi-modal fusion in a shallow and parallel manner instead of a deep and sequential manner. For Q-TriM, we propose a novel framework for attention operation incorporating video and audio conditioned on text. As a result, we obtain not only standard cross attention outputs but also Tri-Modal Attention representations in which Query, Key, and Value come from distinct modalities. These attention representations are combined in parallel at a single stage, thus avoiding the multi-modal fusion with deep stacks in order to mitigate error accumulation and depth-induced issues. Q-TriM achieves state-of-the-art performance on three AVQA benchmarks, including substantial gains on MUSIC-AVQA-R, which demonstrates its robustness and out-of-distribution generalization. Code is available at https://github.com/Sunghun95/Q-TriM

Unified multi-modal models (UMMs) have shown promising interleaved text-image reasoning capabilities, yet effectively optimizing such multi-turn generation via reinforcement learning (RL) remains an open challenge. Existing approaches apply RL exclusively to text steps, relegating image generation to supervised surrogates, preventing policy gradients from propagating through the full interleaved trajectory across heterogeneous modalities. This leaves the potential of RL for UMMs largely untapped. In the paper, we introduce \textbf{BRAID} (\textbf{B}ridging inte\textbf{R}le\textbf{A}ved mult\textbf{I}-modal reasoning as a unified \textbf{D}ecision process), a simple framework that casts multi-turn text-image-text reasoning as a unified Markov decision process (MDP), enabling joint optimization of textual and visual generation via a single, principled RL objective. BRAID computes a shared trajectory-level advantage and propagates it coherently into both text tokens and image denoising paths, each optimized through its modality-native policy gradient mechanism. To further address long-horizon credit assignment, BRAID employs a vision-language model (VLM) judge that scores each intermediate image on its reasoning utility, supplying dense turn-level feedback to sharpen learning at critical visual branches. Experiments on spatial reasoning and visual perception benchmarks show that BRAID consistently outperforms various baselines, confirming that a unified MDP formulation with vision-thinking guidance is essential for effective multi-modal reasoning.

Microbial keratitis requires rapid pathogen identification to guide treatment, but culture- and PCR-based diagnostics are slow and resource-intensive. We developed a triple-phase multimodal framework for bacterial-versus-fungal keratitis classification using slit-lamp photographs acquired under blue-light, sclerotic-scatter, and white-light illumination, together with clinical metadata. The model combines cross-modality contrastive learning, modality-specific fine-tuning, and feature-level multimodal ensemble learning for patient-level prediction. We evaluated the framework on a multicenter dataset of 1,645 patients and 17,158 images from India and the United States. The model achieved 85.84% accuracy, 84.46% average F1-score, and 0.885 AUC. Site-specific evaluation showed that pooled results were overly optimistic, whereas resampling- and balance-based re-evaluation provided a more realistic assessment of cross-site generalization. Under all settings, our framework remained the top-performing approach. The code is available at https://github.com/yqwang01/TPMKA and dataset access will be provided subject to University of Michigan data-sharing clearance.

Multimodal large language models (MLLMs) generate responses autoregressively, integrating visual and linguistic information in an evolving context. Prior work on interpretability has focused on individual layers and circuits (where), leaving the token-level dynamics of multimodal computation during generation (when) underexplored. We address this gap and study attention shifts as per semantic role; tracking model attention to image, text, instruction, and previously generated tokens, One Token at a Time (OTaT). We introduce multimodal tasks that require explicit switching between visual and textual context within a single response. Across two mainstream model families and four open-weight MLLMs of varying sizes, we establish consistent patterns: attention to image peaks at tokens requiring image-derived information, instruction tokens are revisited during task transitions, and attention to previously generated tokens increases as the generation progresses. Causal attention blocking interventions validate the functional role of these trends. We profile model behavior under disrupted attention and observe responses falling back to language priors, or exhibiting cross-modal leakage, denial, or recovery. Finally, informed of the attention dynamics through our novel analysis, we propose a simple test-time intervention to boost attention to the relevant modality at the right time, significantly improving multimodal task performance.

Point cloud-based understanding has become an important enabler for facility operation and maintenance involving indoor building components. However, existing methods output only discrete labels without explaining component functions or natural language interactions. This paper proposes Building-MLLM, a point cloud-centered multimodal large language model (MLLM) for indoor components, which models point clouds and instructions to generate responses across Simple Recognition, Complex Captioning, and Multi-Engineering Question Answering tasks. Building-MLLM addresses semantic concentration through four domain-specific mechanisms: Point Information Enhancer for task-relevant semantics, Geometry-Preserving Regularization preventing geometric erosion, fixed textual prefix for domain stabilization, and multi-dimensional LoRA balancing recognition with reasoning. A multi-constraint progressive instruction-generation engine is developed to compile a synthetic point cloud-text dataset with 4198 objects, 37,782 instruction-following pairs, and 47 categories. Experiments show that Building-MLLM achieves 88.00%, 65.10%, and 68.14% on the three task types, respectively, demonstrating superior indoor component language understanding and providing initial generalizability in transfer inference on other real-world datasets.

Gloss-free Sign Language Translation (SLT) translates sign language videos into spoken-language sentences without gloss annotations, avoiding costly labeling but requiring fine-grained modeling of hands, body, and facial cues. Existing methods often use single-modality or weakly fused features, limiting performance. We propose ViPo-MLLM, a framework that integrates spatio-temporal RGB and human pose features. Dedicated encoders model intra-modal dynamics and cross-modal attention captures long-range dependencies. The fused representation is conditioned with a structured prompt and processed by an LLM trained with contrastive and language modeling objectives. The proposed model was evaluated on the PHOENIX14T and CSL-Daily datasets and achieved new state-of-the-art results on both datasets. Moreover, the ViPo-MLLM model attained competitive performance compared to gloss-based recognition approaches, confirming the effectiveness of the proposed pose cues and cross-modal attention mechanisms.

Decades of orbital missions have produced multi-modal remote sensing data for the Moon, spanning optical imagery, spectroscopy, thermal emission, radar, gravity, and elemental composition. Yet these datasets remain fragmented across archives, and no benchmark exists for evaluating machine learning on lunar data. We introduce Moonstone, the first multi-modal foundation model benchmark for lunar remote sensing. Our contributions are: (1) a 28-channel, 128 pixels-per-degree (~237 m) global lunar pretraining dataset from seven instrument families across five missions, (2) MG-MAE, a modality-grouped masked autoencoder with per-group convolutional tokenizers, a shared Vision Transformer encoder, attention masking for missing modalities, coverage-adaptive masking for heterogeneous spatial coverage, and spectral continuity regularization for physically plausible reconstructions, and (3) a benchmark of six downstream tasks covering classification, regression, and segmentation. MG-MAE pretrained features outperform scratch baselines on all tasks and surpass both ImageNet-pretrained and vanilla MAE baselines by large margins. Data and code are available at https://huggingface.co/datasets/ayushprd/Moonstone and https://github.com/ayushprd/Moonstone .

This paper challenges the assumption that vision-language models (VLMs) require fixed patch-based 2D vision features. Analyzing fine-tuned vision encoders, we find that representations become increasingly abstract and less spatially coherent during VLM training. Notably, models trained with image-text alignment (such as SigLIP2) develop a small number of specialized tokens that effectively summarize global image content. Building on this, we introduce RADIO1D, which compresses images into a compact, variable-length 1D token sequence using multi-teacher knowledge distillation and an autoencoder design. The resulting representations exhibit strong hierarchical summarization, enabling accurate scene understanding - even with a single token - and support improved composition-aware image retrieval. In VLMs, RADIO1D provides flexible accuracy-efficiency tradeoffs through adjustable token counts, delivering competitive performance on diverse multimodal benchmarks with lower computational overhead and better accuracy.

Recent studies on multimodal traffic accident understanding have mainly relied on infrastructure-camera footage, satellite imagery, or structured crash records. However, such data sources are costly to deploy and maintain at large scale, and they cannot objectively capture what the driver was actually able to observe before the accident. In contrast, ego-view accident videos directly represent the driver's visual perspective, making them suitable for reasoning about avoidability and driver responsibility. In this paper, we introduce responsibility distribution estimation for ego-view traffic accident videos, a new task in which a model predicts the percentage of responsibility assigned to each involved agent. We construct an LLM-assisted responsibility annotation pipeline and fine-tune multimodal large language models under multiple input settings, including raw frames, segmentation-enhanced input, and textual descriptions. Experimental results establish a strong initial benchmark, demonstrating that multimodal LLMs can effectively perform this nuanced, constraint-based reasoning task. Our findings suggest that ego-centric accident videos provide a promising foundation for socially and legally meaningful multimodal reasoning beyond conventional accident classification and explanation tasks.

World models aim to capture environment dynamics in ways that support perception, reasoning, and action, and have recently become a central direction in Vision-Language-Action-World (VLAW) modeling. Meanwhile, unified vision-language models have demonstrated strong multimodal generation capabilities, yet their potential as world models remains underexplored. In this work, we introduce \texttt{WorldBagel}, a unified VLAW framework built on BAGEL, a modern multimodal unified model, and use it to systematically investigate the role of unification in world modeling. Across multi-task robotic manipulation and cross-domain experiments, \texttt{WorldBagel} consistently outperforms task-specific alternatives and learns action representations that are more structured and semantically aligned with visual and linguistic context. Experiments on LIBERO, Language Table, and Franka show that unification is not only an architectural convenience, but also a key factor in learning effective VLAW models, leading to consistent empirical gains and deeper insights into multimodal world modeling. Code and model checkpoints will be released upon acceptance.

Multimodal retrieval-augmented generation (RAG) grounds a generator in evidence drawn from heterogeneous modalities -- text, tables, and images. The dominant deployment choice is binary and made before the model has tried to answer: either run a cheap text(+table) pipeline, or pay for an expensive vision-language model (VLM) over every image. Recent adaptive systems improve on this by selecting the modality or fidelity pre-retrieval, from a question-conditioned predictor of which modality will be needed. We show that this is the wrong decision point. Through an oracle headroom analysis on MultiModalQA, we find that the relevance of a modality to a question is a weak predictor of whether that modality is actually needed to answer correctly: a large fraction of questions whose gold support includes an image are nonetheless answerable from text and tables alone, and a pre-retrieval router that escalates on apparent visual relevance over-escalates substantially relative to an oracle. We propose \textbf{post-hoc selective modality escalation}: answer cheaply from text and tables, run a verifier on the (query, draft answer, evidence) tuple that localizes which modality is missing, and pay for VLM evidence only there. A calibrated value-of-escalation router then decides whether the expected accuracy gain justifies the visual cost. On MultiModalQA, our router recovers the accuracy of an always-on VLM pipeline while issuing far fewer visual calls, and closes most of the gap to the oracle escalation rate. The result extends a routing-signal hierarchy established for retrieval depth and reasoning hops to a third axis -- modality -- under a single cost-aware selective-escalation view.

The accuracy of consumer-grade inertial navigation is bottlenecked by the stochastic noise of Micro-Electro-Mechanical Systems (MEMS). Traditional deterministic neural architectures often succumb to ``estimation jittering,'' sacrificing high-frequency kinematic fidelity for numerical stability. We propose PedestrianDiffusion, a multimodal spectral-domain generative framework reformulating dense 6D state estimation as a continuous conditional denoising process. By operating in the frequency domain, our formulation bounds the spectral covariance, acting as a mathematical preconditioner to stabilize the reverse diffusion trajectory. Furthermore, we introduce a zero-shot semantic conditioning mechanism leveraging vision-language embeddings as categorical priors to generalize across heterogeneous sensor noise profiles. To address the computational intractability of generative tracking, we deploy a single-step deterministic probability flow ODE solver ($T=1$). This yields high-capacity asynchronous batch trajectory refinement, establishing the viability of generative architectures for asynchronous batch trajectory refinement on edge hardware. Extensive evaluations on the OxIOD, RIDI, RoNIN, and TLIO benchmarks demonstrate that PedestrianDiffusion achieves state-of-the-art performance, exhibiting unprecedented robustness to impulse perturbations and coupled 6D kinematic drift. This work provides a rigorous algorithmic blueprint for next-generation Neural Inertial Measurement Units (N-IMUs).

Recent large language models (LLMs) have demonstrated remarkable progress in 3D spatial reasoning, spatial grounding, and fine-grained geometric understanding. However, their ability to reason about densely packed object placement under strict spatial and functional constraints remains largely unexplored, despite being a fundamental challenge in practical electronic design automation (EDA) workflows. To bridge this gap, we introduce OmniLayout, the first benchmark designed to evaluate LLMs on printed-circuit-board (PCB) layout placement reasoning under real-world geometric, routing, and connectivity constraints. OmniLayout contains 1,681 industrial-grade schematic-coupled PCB layouts and includes four tasks: (1) geometric reasoning for IC physical placement, with 77.24K placement instances constrained within PCB board boundaries; (2) routability-aware placement reasoning, generating physically valid component placements; (3) electrical functionality, preserving schematic-specified connectivity and electronic functional correctness; and (4) tool-augmented agentic reasoning for invoking external tools to accomplish tasks (1)-(3). Our results reveal substantial limitations of current LLMs in PCB layout placement, including weak geometric reasoning, poor routability optimization, and inconsistent preservation of electrical functionality.

We present OpenGlass, an open-source, privacy-oriented, local-first system for low-latency multimodal visual assistance, with a primary focus on blind and low-vision users. Cloud MLLM assistants offer strong visual understanding, but often require uploading first-person visual data and can suffer multi-second network delays; wearable glasses are ideal for sensing, but cannot host large models under tight compute and power budgets. OpenGlass addresses this gap with a sensing-computing split: an ESP32-based glasses-side unit captures visual context, while a nearby consumer-grade device performs local MLLM inference and local speech output, reducing cloud reliance and keeping raw egocentric visual data on user-controlled devices by default. We evaluate response quality, query-ready-to-audio latency, safety-aware abstention, and auditable logs. Under real ESP32 Wi-Fi capture, OpenGlass reaches 993 ms median user-to-audio latency with resized payloads and 1625 ms with raw 1280 x 720 payloads; 97.5% and 93.3% of trials fall below 2 s, respectively. OpenGlass is a user-initiated visual-assistance reference platform for obstacle/hazard awareness, sign/object queries, and image-quality self-checking, rather than a certified navigation aid. We release source code, hardware instructions, prompts, evaluation data, and logs.

While Multimodal Large Language Models (MLLMs) demonstrate impressive general capabilities, they struggle with fine-grained perception in ultra-high-resolution (UHR) images, particularly for tiny objects in cluttered scenes. Existing methods face a dilemma: they either rely on inefficient prior-free scanning, or depend on static prior-driven heuristics that lack posterior correction to rectify initial model biases. To address this, we propose BVS (Bayesian Visual Search), a framework that formulates perception as a global optimization problem over a continuous spatial-scale manifold. Specifically, BVS bridges prior guidance with posterior correction: it utilizes an early-stop attention rollout of MLLM to construct reasoning-aware priors, while employing a scale-aware non-stationary kernel and GP-UCB to dynamically rectify noise and recover missing information in the prior through iterative local observations. We provide theoretical guarantees via sub-linear regret bounds, and extensive experiments demonstrate that BVS significantly outperforms state-of-the-art baselines with a superior trade-off between accuracy and efficiency.