ICML 2026 论文列表

Text-to-Audio-Video (T2AV) generation aims to synthesize temporally coherent video and semantically synchronized audio from natural language, yet its evaluation remains fragmented, often relying on unimodal metrics or narrowly scoped benchmarks that fail to capture cross-modal alignment, instruction following, and perceptual realism under complex prompts. To address this limitation, we present T2AV-Compass, a unified benchmark for comprehensive evaluation of T2AV systems, consisting of 500 diverse and complex prompts constructed via a taxonomy-driven pipeline to ensure semantic richness and physical plausibility. Besides, T2AV-Compass introduces a dual-level evaluation framework that integrates objective signal-level metrics for video quality, audio quality, and cross-modal alignment with a subjective MLLM-as-a-Judge protocol for instruction following and realism assessment. Extensive evaluation of 15 representative T2AV systems reveals that even the strongest models fall substantially short of human-level realism and cross-modal consistency, with persistent failures in audio realism, fine-grained synchronization, instruction following, etc. These results indicate significant improvement room for future models and highlight the value of T2AV-Compass as a challenging and diagnostic testbed for advancing text-to-audio-video generation.

Coverless Image Steganography (CIS) hides information without explicitly modifying a cover image, providing strong imperceptibility and inherent robustness to steganalysis. However, existing CIS methods largely lack robust access control, making it difficult to selectively reveal different hidden contents to different authorized users. Such access control is critical for scalable and privacy-sensitive information hiding in multi-user settings. We propose MIDAS (Multi-Image Diffusion-based Access-controlled Steganography), a training-free diffusion-based CIS framework that enables multi-image hiding with user-specific access control via latent-level fusion. MIDAS introduces a Random Basis mechanism to suppress residual structural information, together with a theoretical analysis of information leakage, and a Latent Vector Fusion module that reshapes aggregated latents to better align with the diffusion process. Experimental results demonstrate that MIDAS consistently outperforms existing training-free CIS baselines in access control functionality, stego image quality and diversity, robustness to noise, and resistance to steganalysis, establishing a practical and scalable approach to access-controlled coverless steganography.

Although Federated Learning (FL) offers advantages in privacy-preserving for cross-device collaborative learning, its practical deployment remains severely constrained by heterogeneous hardware resources and non-IID (non-independent and identically distributed) data across devices. Sub-model extraction has emerged as a widely adopted strategy for enabling collaborative training among devices with heterogeneous models. However, existing sub-model extraction methods in FL typically rely on coarse-grained stochastic selection or rigid rule-based neuron selection, which severely limits training performance. Specifically, stochastic strategies lead to severe parameter conflicts under non-IID data distributions, while rule-based approaches lack diversity in neuron selection per device, preventing comprehensive parameter optimization. To address this problem, this paper presents a novel sub-model extraction-based FL framework, named SpineFL, which adopts a backbone-sharing mechanism and an activation-guided pruning strategy for sub-model extraction. Specifically, SpineFL decomposes each global model layer into two portions: i) a mandatory backbone shared by all the sub-models to maintain model generalization, and ii) a dynamic portion for sub-model extraction. SpineFL adopts the activation-guided selection strategy to probabilistically select neurons according to their activation frequency from the dynamic portion to generate sub-model, where neurons exhibiting higher historical activation are more likely to be included, thereby simultaneously addressing parameter conflicts while preserving selection diversity. Experimental results demonstrate that compared with state-of-the-art heterogeneous FL methods, SpineFL can achieve up to 3.28% accuracy improvement.

Scene Graph Generation (SGG) connects visual perception with structured reasoning, but is limited by scarce annotations and the long-tailed distribution of relational predicates. Training-free methods based on vision-language models (VLMs) reduce supervision requirements, yet often rely on flat graph structures and produce hallucinated relations. We present HSGG, a training-free framework for open-world Hierarchical Scene Graph Generation, whose inference proceeds in two steps. First, bidirectional hierarchical entity perception combines top-down object expansion with bottom-up attribute reasoning to construct multi-level scene hierarchies capturing part–whole semantics. Second, geometry-guided relation reasoning infers valid relations from these structured entities: geometry-aware relation filtering first prunes spatially implausible object pairs using 2D proximity, depth cues, and object scale, and geometry-grounded contrastive relation decoding then suppresses hallucinated predicates by contrasting predictions from a visually grounded expert against a hallucination-prone geometric prior, ensuring relations are both geometrically consistent and semantically coherent. Experiments show that HSGG generalizes effectively to unseen objects and predicates without training, substantially reduces relational hallucinations, and consistently improves downstream reasoning performance.

With the rapid advancement of large multimodal models (LMMs), inference-time overhead has become a key bottleneck for real-world deployment. Existing methods typically prune visual tokens at prefill, assuming the required visual evidence remains static during reasoning. However, we empirically show that visual evidence is strongly step-dependent: only a sparse subset of visual tokens is critical at each decoding step, and the critical set evolves across reasoning. Furthermore, we identify a coupled bottleneck where redundant visual context can steer the model toward query-irrelevant regions, lengthening the reasoning trace. Guided by these insights, we propose **VisionPulse**, a step-wise visual token pruning framework during reasoning. VisionPulse computes a lightweight visual attention mass to estimate the step-wise retention budget by exploiting its strong positive correlation with LMMs' effective visual token usage and retain only the most critical tokens under this budget. By enforcing visual sparsity during reasoning, VisionPulse filters redundant visual context while preserving relevant visual evidence, shortening reasoning traces naturally. Extensive experiments show that VisionPulse only retains 5\% of visual tokens per step with reasoning traces shortened by 11.2\%, while keeping accuracy almost unchanged.

Auditory attention decoding (AAD) based on Electroencephalography (EEG) aims to identify the attended speaker in multi-speaker environments. However, existing methods typically overlook the crucial phase information of EEG signals, which limits their ability to distinguish structured neural patterns from random noise in the frequency domain and hinders robust decoding. To address these issues, this paper proposes a Phase-aware Complex Refinement Network (PCRNet) for AAD, which consists of a Temporal Context Calibration (TCC) module and a Dual-Domain Integration (DDI) module. Specifically, the TCC module captures long-range temporal dependencies through multi-scale temporal attention mechanism, while the DDI module employs a phase-guided spectral filtering strategy to dynamically suppress noise-dominated frequencies and refine the real and imaginary components separately. This design enables effective phase recalibration and enhances the discriminability of target features in the complex domain. Experimental results on three public datasets demonstrate that PCRNet outperforms state-of-the-art (SOTA) methods, particularly under challenging ultra-short 0.1-second windows. Code is available at: https://github.com/SunshineGreeny/PCRNet.

The trustworthiness of evaluation is critical to reliable model comparison and deployment in Video Anomaly Understanding (VAU). However, existing metrics are sensitive to expression styles and normal content, and this field lacks a diagnostic benchmark to validate metric validity and robustness. To bridge this gap, we propose: (1) a Class-Guided Chain-of-Evaluation (CG-CoE) metric, which structures assessment by extracting anomalous events and matching them under a class-specific semantic tolerance boundary, thereby decoupling anomaly semantics from descriptive style; and (2) an anomaly-focused meta-evaluation benchmark with two subsets: Anomalous Event-level Annotations (AEA) for measuring the validity of reflecting VAU models’ anomaly understanding ability and Controlled Variant Pairs (CVP) with fixed anomalies for quantifying robustness to stylistic perturbations. Extensive experiments demonstrate that CG-CoE achieves SOTA validity and robustness.

Charts are a fundamental visualization format for structured data analysis. Enabling end-to-end chart editing according to user intent is of great practical value, yet remains challenging due to the need for both fine-grained control and global structural consistency. Most existing approaches adopt pipeline-based designs, where natural language or code serves as an intermediate representation, limiting their ability to faithfully execute complex edits. We introduce ChartE$^{3}$, an **E**nd-to-**E**nd Chart **E**diting benchmark that directly evaluates models without relying on intermediate natural language programs or code-level supervision. ChartE$^{3}$ focuses on two complementary editing dimensions: local editing, which involves fine-grained appearance changes such as font or color adjustments, and global editing, which requires holistic, data-centric transformations including data filtering and trend line addition. ChartE$^{3}$ contains over 1,200 high-quality samples constructed via an well-designed data pipeline with human curation. Each sample is provided as a triplet of a chart image, its underlying code, and a multimodal editing instruction, enabling evaluation from both visual and structural perspectives. Extensive benchmarking of state-of-the-art multimodal large language models reveals substantial performance gaps, particularly on global editing tasks, highlighting critical limitations in current end-to-end chart editing capabilities.

People with limb differences often face significant challenges in accessing inclusive AI services, largely due to the lack of structured, high-quality resources centered on disability contexts. In this work, we introduce a limb-deficiency aware body-centric learning and evaluation paradigm that involves (i) a large-scale limb-aware vision–language dataset and evaluation benchmark for multimodal reasoning, and (ii) a model adaptation strategy for Vision-Language Models (VLM) in limb-difference contexts. Specifically, we first collect limb-difference data covering all eight limb-deficiency types across diverse real-world scenarios. The data are systematically organized into 96 limb-affected human action categories and 68 medical-functional classes defined by the World Health Organization (WHO). Then, we curate an expert-annotated vision–language dataset for limb-aware multimodal understanding, named Inclusive VQA for Limb Deficiency (**IVQA-LD**). IVQA-LD comprises 80K VQA pairs spanning eight core tasks including visual grounding, quantitative reasoning, functional semantic classification, and instructional text generation. We benchmark state-of-the-art VLMs on IVQA-LD and find that they consistently struggle across all tasks, exposing substantial deficiencies in limb-aware perception and reasoning. To address this, we further propose a Body-centric Structure-aware Initialization (BSI) strategy that aligns model representations with limb-specific semantics. With BSI, VLMs fine-tuned on IVQA-LD achieve significant performance improvements across all the tasks. We will publicly release the dataset to support future research.

As digital environments (data distribution) are in flux, with new GUI data arriving over time-introducing new domains or resolutions-agents trained on static environments deteriorate in performance. In this work, we introduce Continual GUI Agents, a new task that requires GUI agents to perform continual learning under shifted domains and resolutions. We find existing methods fail to maintain stable grounding as GUI distributions shift over time, due to the diversity of UI interaction points and regions in fluxing scenarios. To address this, we introduce GUI-Anchoring in Flux (GUI-AiF), a new reinforcement fine-tuning framework that stabilizes continual learning through two novel rewards: Anchoring Point Reward in Flux (APR-iF) and Anchoring Region Reward in Flux (ARR-iF). These rewards guide the agents to align with shifting interaction points and regions, mitigating the tendency of existing reward strategies to over-adapt to static grounding cues (e.g., fixed coordinates or element scales). Extensive experiments show GUI-AiF surpasses state-of-the-art baselines. Our work establishes the first continual learning framework for GUI agents, revealing the untapped potential of reinforcement fine-tuning for continual GUI Agents.

This paper presents OvisOCR, a lightweight and strictly end-to-end Multimodal Large Language Model (MLLM) tailored for document parsing. Unlike current methods that rely on complex "Crop-OCR-Merge'" cascades to handle high-resolution inputs, OvisOCR directly maps full-page visual signals to structured Markdown without localized slicing or layout detection dependencies. Through extensive evaluations on the OmniDocBench, OvisOCR achieves top-tier performance among all compared methods in all aspects, demonstrating that a compact E2E model can effectively ``digest'' the capabilities of intricate pipelines and surpass specialized and general methods. Technically, OvisOCR unifies fine-grained recognition and semantic correction in a single model by leveraging supervision bootstrapped from strong OCR engines and refined via general-model-based data cleaning. To balance the performance across diverse document constituents, we design category-specific reward mechanisms for distinct element types, such as dense text, complex tables, and formulas, and ensure the model enhances its formatting strengths for each modality concurrently. This approach effectively resolves the optimization conflict, guaranteeing that improvements in structural layout parsing do not come at the expense of omitting fine-grained textual details. Empirical results confirm that OvisOCR eliminates the error propagation inherent in split-and-merge architectures, offering a streamlined path for future document intelligence. Our model can be available at \url{https://huggingface.co/AIDC-AI/OvisOCR}.

The online construction of vectorized high-definition (HD) maps is a cornerstone of modern autonomous driving systems. State-of-the-art approaches, particularly those based on the DETR framework, formulate this as an instance detection problem. However, their reliance on independent, learnable object queries results in a predominantly local query perspective, neglecting the inherent global representation within HD maps. In this work, we propose \textbf{MapGR} (\textbf{G}lobal \textbf{R}epresentation learning for HD \textbf{Map} construction), an architecture designed to learn and utilize global representations from queries. Our method introduces two synergistic modules: a Global Representation Learning (GRL) module, which encourages the distribution of all queries to better align with the global map through a carefully designed holistic segmentation task, and a Global Representation Guidance (GRG) module, which endows each individual query with explicit, global-level contextual information to facilitate its optimization. Evaluations on the nuScenes and Argoverse2 datasets validate the efficacy of our approach, demonstrating substantial improvements in mean Average Precision (mAP) compared to leading baselines. Code is available at \href{https://github.com/skyshoumeng/MapGR}{\texttt{github.com/skyshoumeng/MapGR}}.

We present $\textit{Video-in-the-Loop}$ (ViTL), a two-stage long-video QA framework that preserves a fixed token budget by first $\textit{localizing}$ question-relevant interval(s) with a low-fps skim and then $\textit{answering}$ via span-aware reallocation of visual tokens at higher effective frame rate, emitting an interleaved output with both spans and the final option for direct attribution. We also introduce $\textit{VGrounding-QA}$, which converts description based event graphs into $\textit{span-grounded}$ multiple-choice QA by pairing each question with $\textit{ground-truth}$ time span(s) and related reasoning. ViTL is trained end-to-end with an interleaved group-relative objective that couples temporal IoU for localization with answer correctness, allowing credit to flow from answers back to spans without increasing compute. Under fixed token budgets, ViTL attains up to 8.6\% with 50\% less frame input on long-video QA and temporal grounding (e.g., Charades-STA, ActivityNet-Captions) and ablations show that span-aware token reallocation consistently surpasses uniform sampling. Together, $\textit{VGrounding-QA}$ and ViTL provide an interpretable, compute-efficient recipe for scalable long-video QA.

Driven by rapid advances in large-scale generative models, synthetic data has emerged as a promising solution for visual understanding. While modern diffusion models achieve remarkable photorealistic image synthesis, their potential in complex visual segmentation tasks remains underexplored. In this work, we conduct a systematic analysis of synthetic images from state-of-the-art diffusion models to uncover the factors governing their utility. In particular, synthetic images characterized by dense scene composition and fine instance fidelity demonstrate distinctive benefits, yielding significantly more discriminative spatial representations. Building on these insights, we propose SENSE, a unified framework that leverages flexible and scalable synthetic data to substantially enhance segmentation performance. Notably, SENSE is model-agnostic, compatible with diverse architectures (e.g., DPT and Mask2Former), and scales effectively across models with varying parameter capacities. Extensive experiments on Cityscapes, COCO, and ADE20K validate the effectiveness and generalization capability of our approach. Code is available at https://github.com/zhang0jhon/SENSE.

Recent layout-to-image models have achieved remarkable progress in spatial controllability. However, they still struggle with inter-object occlusion. When bounding boxes overlap, most existing methods lack explicit occlusion information, which makes the generation in intersection regions inherently ambiguous and hinders the determination of complex occlusion relationships. As a result, they often produce entangled textures or physically inconsistent layering in the overlapped areas. To address this issue, we first construct ***SA-Z***, a large-scale dataset enriched with explicit occlusion ordering and pixel-level annotations. Building upon our proposed dataset, we introduce ***OcclusionFormer***, a novel occlusion-aware Diffusion Transformer framework that explicitly models Z-order priority by decoupling instances and compositing them via volume rendering. Furthermore, to ensure fine-grained spatial precision, we introduce a queried alignment loss that explicitly supervises individual instances and enhances semantic consistency. The proposed method effectively reduces ambiguity in overlapping regions, enforces correct occlusion dependencies, and preserves structural integrity, leading to substantial accuracy gains across diverse scenes.

Recent approaches for video generation with camera control often create anchor videos (i.e., rendered videos that approximate desired camera motions) to guide diffusion models as a structured prior, by rendering from estimated point clouds following camera trajectories. However, errors in point cloud and camera trajectory estimation often lead to inaccurate anchor videos with higher training cost and low efficiency, as the model is forced to compensate for rendering misalignments. To address these limitations, we introduce EPiC, an efficient and precise camera control learning framework that constructs well-aligned training anchor videos without the need for camera pose or point cloud estimation. Concretely, we create highly precise anchor videos by masking source videos based on first-frame visibility, which ensures strong alignment, eliminates the need for camera/point cloud estimation, and thus can be readily applied to any in-the-wild video. Furthermore, we introduce Anchor-ControlNet, a lightweight module that integrates anchor video guidance in visible regions to pretrained video diffusion models, with less than 1\% of additional parameters. EPiC achieves efficient training with substantially fewer parameters, training steps, and less data, and generalizes robustly to anchor videos made with point clouds at test time, enabling precise 3D-informed camera control. EPiC achieves SoTA performance on RealEstate10K and MiraData for I2V camera control task. Notably, EPiC also exhibits strong zero-shot generalization to video-to-video (V2V) scenarios.

Text-conditioned image generation has made rapid progress, yet rendering images with long-form text remains challenging due to the limitations of existing datasets, which predominantly focus on short and simple text. We introduce TextAtlas5M, a large-scale dataset designed to evaluate long-text rendering, where “long text” encompasses not only textual length but also layout complexity and semantic richness. TextAtlas5M contains 5 million generated and collected images across diverse data types, enabling comprehensive evaluation of large-scale generative models. We further curate 4,000 human-improved test cases (TextAtlasEval) spanning four domains, forming one of the most extensive benchmarks for text rendering. Evaluations show that TextAtlas5M poses substantial challenges even for state-of-the-art proprietary models (e.g., GPT-4o), with significantly larger gaps observed for open-source models. Training on TextAtlas5M consistently improves text rendering for both diffusion-based and autoregressive models, demonstrating its effectiveness for advancing text-rich image generation.

Generative priors in Image Super-Resolution (SR) often compromise faithful restoration, we attribute this limitation to a fundamental spectral misalignment between isotropic objectives and the intrinsic natural image manifold. While Direct Preference Optimization offers a path to alignment, its reliance on spectrally flat Gaussian noise fails to distinguish authentic high-frequency details from hallucinations. To bridge this geometric gap, we propose ASASR, a theoretically grounded framework that recasts the generative flow into a Sobolev-induced Riemannian geometry by explicitly coloring the noise transition kernel to mirror natural spectral decay. Driving this geometric alignment, we integrate a parametric adversary grounded in the Riesz Representation Theorem, which synthesizes targeted negative samples equivalent to worst-case Sobolev gradients to direct optimization along the tangent space of plausible structural failures. Extensive evaluations demonstrate that ASASR outperforms leading generative baselines, particularly in preserving spectral consistency and structural fidelity, offering a robust solution that effectively mitigates artifacts.

Multimodal large language models (MLLMs) exhibit strong visual-language reasoning, yet remain confined to their native modalities and cannot directly process structured, non-visual data such as human skeletons. Existing methods either compress skeleton dynamics into lossy feature vectors for text alignment, or quantize motion into discrete tokens that generalize poorly across heterogeneous skeleton formats. We present SkeletonLLM, which achieves universal skeleton understanding by translating arbitrary skeleton sequences into the MLLM's native visual modality. At its core is DrAction, a differentiable, format-agnostic renderer that converts skeletal kinematics into compact image sequences. Because the pipeline is end-to-end differentiable, MLLM gradients can directly guide the rendering to produce task-informative visual tokens. To further enhance reasoning capabilities, we introduce a cooperative training strategy: Causal Reasoning Distillation transfers structured, step-by-step reasoning from a teacher model, while Discriminative Finetuning sharpens decision boundaries between confusable actions. SkeletonLLM demonstrates strong generalization on diverse tasks including recognition, captioning, reasoning, and cross-format transfer---suggesting a viable path for applying MLLMs to non-native modalities. Code is available in the Supplementary Material.