CVPR 2026 论文列表

Despite advances in the application of MLLMs for various video tasks, video event prediction (VEP) remains relatively underexplored. VEP requires the model to perform fine-grained temporal modeling of videos and establish logical relationships between videos and future events, which current MLLMs still struggle with. In this work, we first present a comprehensive evaluation of current leading MLLMs on the VEP task, revealing the reasons behind their inaccurate predictions, including lack of logical reasoning ability for future events prediction and insufficient utilization of visual information. To address these challenges, we propose **C**hain **o**f **E**vents (**CoE**) paradigm, which constructs temporal event chains to implicitly enforce MLLM focusing on the visual content and the logical connections between videos and future events, incentivizing model's reasoning capability with multiple training protocols. Experimental results on public benchmarks demonstrate that our method outperforms both leading open-source and commercial MLLMs, establishing a new state-of-the-art on the VEP task.

Test-Time Adaptation (TTA) is essential for enabling deep learning models to handle real-world data distribution shifts. However, current approaches face significant limitations: backpropagation-based methods are not suitable for low-end deployment devices, due to their high computation and memory requirements, as well as their tendency to modify model weights during adaptation; while traditional backpropagation-free techniques exhibit constrained adaptation capabilities. In this work, we propose Forward-Only Zeroth-Order Optimization (FOZO), a novel and practical backpropagation-free paradigm for TTA. FOZO leverages a memory-efficient zeroth-order prompt optimization, which is led by objectives optimizing both intermediate feature statistics and prediction entropy. To ensure efficient and stable adaptation over the out-of-distribution data stream, we introduce a dynamically decaying perturbation scale during zeroth-order gradient estimation and theoretically prove its convergence under the TTA data stream assumption. Extensive continual adaptation experiments on ImageNet-C, ImageNet-R, and ImageNet-Sketch demonstrate FOZO's superior performance, achieving 59.52% Top-1 accuracy on ImageNet-C (5K, level 5) and outperforming main gradient-based methods and SOTA forward-only FOA (58.13%). Furthermore, FOZO exhibits strong generalization on quantized (INT8) models. These findings demonstrate that FOZO is a highly competitive solution for TTA deployment in resource-limited scenarios. Code: https://github.com/eVI-group-SCU/FOZO

Recent visual autoregressive (AR) models have shown promising capabilities in text-to-image generation, operating in a manner similar to large language models. While test-time computation scaling yields higher-quality outputs on challenging natural language tasks, its adaptation to visual AR models remains unexplored and poses unique challenges. Naively applying test-time scaling strategies such as Best-of-N can be suboptimal: they consume full-length computation on erroneous generation trajectories, while the raster-scan decoding scheme lacks a blueprint of the entire canvas, limiting scaling benefits as only a few prompt-aligned candidates are generated. To address these, we introduce GridAR, a test-time scaling framework designed to elicit the best possible results from visual AR models. GridAR employs a grid-partitioned progressive generation scheme in which multiple partial candidates for the same position are generated within a canvas, infeasible ones are pruned early, and viable ones are fixed as anchors to guide subsequent decoding. Coupled with this, we present a layout-specified prompt reformulation strategy that inspects partial views to infer a feasible layout for satisfying the prompt. The reformulated prompt guides subsequent image generation to mitigate the blueprint deficiency. GridAR achieves higher-quality results under limited test-time scaling: with N=4, it even outperforms Best-of-N (N=8) by 14.4% on T2I-CompBench++ while reducing cost by 25.6%. It also generalizes to autoregressive image editing, showing a 14.5% gain in semantic preservation on PIE-Bench over larger-N baselines.

Estimating continuous optical flow is a fundamental yet challenging problem in dynamic visual perception. Event-based cameras, with microsecond latency and high dynamic range, capture brightness changes asynchronously, offering a unique opportunity to model motion with fine temporal precision. However, the scarcity of temporally dense ground-truth annotations limits the effectiveness of supervised learning, while contrast maximization (CM) frameworks, focused on sharpening the Image of Warped Events (IWE), often neglect temporal continuity and structural coherence, leading to distorted trajectories under complex motion.To overcome these challenges, we propose a hybrid-supervised framework for continuous-time optical flow estimation, grounded in the principle of Spatio-temporal Structural Consistency (STSC). This paradigm jointly enforces local structural stability and trajectory continuity, ensuring physically coherent motion across time. To further enhance representation and robustness, we design a bidirectionally complementary multi-scale architecture and employ a curriculum-guided hybrid training strategy, enabling a smooth transition from supervised point constraints to self-supervised manifold regularization.Comprehensive experiments across multiple benchmarks show that our method achieves state-of-the-art performance in both continuous-time and standard optical flow estimation, demonstrating the effectiveness of the proposed learning paradigm.

Recently, the introduction of Chain-of-Thought (CoT) has largely improved generation ability of unified models. However, it is observed that the current thinking process during generation mainly focuses on the text consistency with the text prompt, ignoring the visual context consistency with the visual reference images during the multi-modal generation, e.g., multi-reference generation. The lack of such consistency results in the failure in maintaining key visual features (like human ID, object attribute, style). To this end, we integrate the visual context consistency into the reasoning of unified models, explicitly motivating the model to sustain such consistency by 1) Adaptive Visual Planning: generating structured visual check list to figure out the visual element of needed consistency keeping, and 2) Iterative Visual Correction: performing self-reflection with the guidance of check lists and refining the generated result in an iterative manner. To achieve this, we use supervised finetuning to teach the model how to plan the visual checking, conduct self-reflection and self-refinement, and use flow-GRPO to further enhance the visual consistency through a customized visual checking reward. The experiments show that our method outperforms both zero-shot unified models and those with text CoTs in multi-modal generation, demonstrating higher visual context consistency.

Diffusion large language models (dLLMs) are emerging as promising alternatives to autoregressive (AR) LLMs. Recently, this paradigm has been extended to multimodal tasks, leading to the development of diffusion multimodal large language models (dMLLMs). These models are expected to retain the reasoning capabilities of LLMs while enabling faster inference through parallel generation. However, when combined with Chain-of-Thought (CoT) reasoning, dMLLMs exhibit two critical issues.First, we observe that dMLLMs often generate the final answer token at a very early timestep. This trend indicates that the model determines the answer before sufficient reasoning, leading to degraded reasoning performance. Second, during the initial timesteps, dMLLMs show minimal dependency on visual prompts, exhibiting a fundamentally different pattern of visual information utilization compared to AR vision-language models. In summary, these findings indicate that dMLLMs tend to generate premature final answers without sufficiently grounding on visual inputs.To address these limitations, we propose Position & Step Penalty (PSP) and Visual Reasoning Guidance (VRG). PSP penalizes tokens in later positions during early timesteps, delaying premature answer generation and encouraging progressive reasoning across timesteps. VRG, inspired by the classifier-free guidance, amplifies visual grounding signals to enhance the model's alignment with visual evidence.Extensive experiments across various dMLLMs demonstrate that our method achieves up to 7.5% higher accuracy while delivering more than 3x speedup compared to reasoning with four times more diffusion steps. Our code will be released after publication.

We consider the problem of imaging a dynamic scene when scene appearance variations can outpace photon arrivals. Under such conditions, a pixel is effectively "blind" to changes in appearance that occur within the timespan separating the photons it detects, and so the inter-photon interval presents a significant speed barrier to video acquisition systems. To analyze and advance imaging capabilities at the inter-photon limit, we introduce a novel reparameterization of time-varying flux that reveals the intrinsic difficulty of signal reconstruction by relating the Fourier decomposition of a flux function to the number of photons arriving within each oscillation period. We find that inter-photon-limited videography of general scenes is underexplored and beyond the reach of existing reconstruction techniques. To this end, we introduce Neural Flux Fields---a technique that combines statistical modeling of photon arrival with intrinsic priors of a neural network to achieve robust videography at the inter-photon limit. Using this approach, we demonstrate never-before-seen capabilities in video reconstruction across a range of captured single-photon video datasets spanning the inter-photon-limited regime.

Test-Time Adaptation (TTA) enhances model robustness to out-of-distribution (OOD) data by updating the model online during inference, yet existing methods lack theoretical insights into the fundamental causes of performance degradation under domain shifts. Recently, Neural Collapse (NC) has been proposed as an emergent geometric property of deep neural networks (DNNs), providing valuable insights for TTA. In this work, we extend NC to the sample-wise level and discover a novel phenomenon termed Sample-wise Alignment Collapse (NC3+), demonstrating that a sample's feature embedding, obtained by a trained model, aligns closely with the corresponding classifier weight. Building on NC3+, we identify that the performance degradation stems from sample-wise misalignment in adaptation which exacerbates under larger distribution shifts. This indicates the necessity of realigning the feature embeddings with their corresponding classifier weights. However, the misalignment makes pseudo-labels unreliable under domain shifts. To address this challenge, we propose NCTTA, a novel feature-classifier alignment method with hybrid targets to mitigate the impact of unreliable pseudo-labels, which blends geometric proximity with predictive confidence. Extensive experiments demonstrate the effectiveness of NCTTA in enhancing robustness to domain shifts. For example, NCTTA outperforms Tent by 14.52% on ImageNet-C. Project page is publicly available at https://github.com/Cevaaa/NCTTA.

Vision-Language Models (VLMs) typically rely on static initial frames for video reasoning, restricting their ability to incorporate essential dynamic information as the reasoning process evolves. Existing methods that augment Chain-of-Thought (CoT) with additional frame information often exhibit suboptimal CoT quality and lack the crucial ability to synthesize visual information for hypothetical or counterfactual scenarios. We introduce Act-to-See (Act2See), a novel framework that enables active visual perception by empowering VLMs to actively interleave video frames within text CoTs. Act2See is developed via Supervised Fine-Tuning (SFT) on a high-quality dataset of reasoning traces generated by a frontier VLM. These traces integrate active calls to either retrieve existing frames or generate new ones, and are rigorously verified against human-annotated CoTs to ensure quality. This approach cultivates an emergent capability: at inference time, the model actively determines when to search for or synthesize the necessary visual evidence. Act2See establishes new state-of-the-art results on challenging benchmarks, including VideoEspresso and ViTIB, and outperforms comparable or larger models on Video-MME, EgoNormia, and VCR-Bench, demonstrating an advancement in enabling VLMs with active visual perception for video reasoning. Code: https://github.com/martinmamql/act2see.

Reliable long-horizon trajectory prediction requires both high positional accuracy and physically plausible temporal motion consistency. However, existing methods suffer from two fundamental limitations. First, they overlook the inherent difference in prediction logic: near-future trajectories are primarily governed by historical dynamics, whereas distant-future behaviors are driven by high-level semantic context. Yet, most methods employ a unified decoding pathway that blurs the temporal distinction.Second, although the near future is relatively easier to predict, existing methods lack mechanisms for coherent trajectory propagation across time horizons, often resulting in kinematically implausible predictions with inconsistent heading evolution and degraded long-horizon performance. To address these challenges, we propose NDPNet, a dual-stage architecture that decouples near- and distant-horizon modeling into specialized pathways, with a dedicated transition module ensuring smooth temporal bridging. Furthermore, we introduce a novel motion-aware coherence loss that explicitly embeds kinematic priors to enforce trajectory consistency. Extensive experiments show that NDPNet achieves SOTA performance on Argoverse 2 and WOMD. Notably, on WOMD, it ranks 1^ \text st in both minFDE _6 and minADE _6 across all standard horizons (3s, 5s, 8s) without ensemble learning or NMS post-processing, and is the first to achieve sub-1.75 minFDE _6 for 8s prediction, surpassing prior methods by a large margin. The code will be released subsequently.

Recent multimodal reasoning models, inspired by DeepSeek-R1, have significantly advanced vision-language systems. However, in remote sensing (RS) tasks, we observe widespread pseudo reasoning: models narrate the process of reasoning rather than genuinely reason toward the correct answer based on visual evidence. We attribute this to the Glance Effect, where a single, coarse perception of large-scale RS imagery results in incomplete understanding and reasoning based on linguistic self-consistency instead of visual evidence. To address this, we propose RS-EoT (Remote Sensing Evidence-of-Thought), a language-driven, iterative visual evidence-seeking paradigm. To instill this paradigm, we propose SocraticAgent, a self-play multi-agent system that synthesizes reasoning traces via alternating cycles of reasoning and visual inspection. To enhance and generalize these patterns, we propose a two-stage progressive RL strategy: first, RL on fine-grained Grounding tasks to enhance RS-EoT capabilities, followed by RL on RS VQA to generalize to broader understanding scenarios. Experiments show RS-EoT achieves state-of-the-art performance on multiple RS VQA and grounding benchmarks. Analyses reveal clear iterative cycles of reasoning and evidence seeking, confirming RS-EoT mitigates the Glance Effect and enables genuine evidence-grounded reasoning. Our code, data, and models are available at https://geox-lab.github.io/Asking_like_Socrates/.

Embodied agents face a critical dilemma that end-to-end models lack interpretability and explicit 3D reasoning, while modular systems ignore cross-component interdependencies and synergies. To bridge this gap, we propose the Dynamic 3D Vision-Language-Planning Model (D3D-VLP). Our model introduces two key innovations: 1) A Dynamic 3D Chain-of-Thought (3D CoT) that unifies planning, grounding, navigation, and question answering within a single 3D-VLM and CoT pipeline; 2) A Synergistic Learning from Fragmented Supervision (SLFS) strategy, which uses a masked autoregressive loss to learn from massive and partially-annotated hybrid data. This allows different CoT components to mutually reinforce and implicitly supervise each other. To this end, we construct a large-scale dataset with 10M hybrid samples from 5K real scans and 20K synthetic scenes that are compatible with online learning methods such as RL and DAgger. Our D3D-VLP achieves state-of-the-art results on multiple benchmarks, including Vision-and-Language Navigation (R2R-CE, REVERIE-CE, NavRAG-CE), Object-goal Navigation (HM3D-OVON), and Task-oriented Sequential Grounding and Navigation (SG3D). Real-world mobile manipulation experiments further validate the effectiveness.

Agent-based editing models have substantially advanced interactive experiences, processing quality, and creative flexibility. However, two critical challenges persist: (1) instruction hallucination--text-only chain-of-thought (CoT) reasoning cannot fully prevent factual errors due to inherent information bottlenecks; (2) reward hacking--dynamic policy optimization against static reward models allows agents to exploit flaws in reward functions. To address these issues, we propose JarvisEvo, a unified image editing agent that emulates an expert human designer by iteratively editing, selecting appropriate tools, evaluating results, and reflecting on its own decisions to refine outcomes. JarvisEvo offers three key advantages: (1) an interleaved multimodal chain-of-thought (iMCoT) reasoning mechanism that enhances instruction following and editing quality; (2) a synergistic editor-evaluator policy optimization (SEPO) framework that enables self-improvement without external rewards, effectively mitigating reward hacking; and (3) support for both preservative and generative editing through seamless integration of Adobe Lightroom and Qwen-Image-Edit tools. On ArtEdit-Bench, JarvisEvo outperforms Nano-Banana by an average of 18.95% on preservative editing metrics, including a substantial 44.96% improvement in pixel-level content fidelity, while maintaining competitive performance in generative editing tasks.

Vocabulary-free fine-grained image recognition aims to distinguish visually similar categories within a meta-class without a fixed, human-defined label set. Existing solutions for this problem remain limited by either the usage of a large and rigid list of vocabularies or by the dependency on complex pipelines with fragile heuristics where errors propagate across stages. Meanwhile, the ability of recent large multi-modal models (LMMs) equipped with explicit or implicit reasoning to comprehend visual-language data, decompose problems, retrieve latent knowledge, and self-correct suggests a more principled and effective alternative. Building on these capabilities, we propose FiNDR (Fine-grained Name Discovery via Reasoning), the first reasoning-augmented LMM-based framework for vocabulary-free fine-grained recognition. The system oper- ates in three automated steps: (i) a reasoning-enabled LMM generates descriptive candidate labels for each image; (ii) a vision-language model filters and ranks these candidates to form a coherent class set; and (iii) the verified names instantiate a lightweight multi-modal classifier used at inference time. Extensive experiments on popular fine-grained classification benchmarks demonstrate state-of-the-art per- formance under the vocabulary-free setting, with a significant relative margin of up to 18.8% over previous ap- proaches. Remarkably, the proposed method surpasses zero-shot baselines that exploit pre-defined ground-truth names, challenging the assumption that human-curated vo- cabularies define an upper bound. Ablations further confirm that advanced prompting techniques and built-in rea- soning mechanisms significantly enhance naming quality. Additionally, we show that carefully engineered prompts enable open-source LMMs to match proprietary counter- parts. These findings establish reasoning-augmented LMMs as an effective foundation for scalable, fully automated, open-world fine-grained visual recognition. The source code and relevant prompting guidelines will be released.

LiDAR semantic segmentation must remain robust under various sensor and environmental corruptions to be reliable in safety-critical applications. Existing test-time adaptation methods, including approaches based on pseudo-labels and normalization statistics, have shown promising results but can still struggle under severe distribution shifts. To complement these approaches, we propose a geometry-aware test-time training framework that leverages an auxiliary self-supervised objective. Our method is based on geometric inlier discrimination (GeoID), which injects synthetic off-manifold points into the input and trains the model to distinguish geometry-consistent inliers from synthetically displaced outliers, enabling adaptation on unlabeled test data. To further stabilize this process under real corruptions, we introduce bidirectional unreliable point filtering (BiUPF), which uses inlier scores from the source-trained model to filter out unreliable regions on both original and synthetic points, focusing updates on high-confidence samples. Experiments on two large-scale corruption benchmarks, SemanticKITTI-C and nuScenes-C, show that our method consistently outperforms strong test-time adaptation baselines and improves robustness across diverse LiDAR corruptions. Our code is available at https://github.com/hskim617/GeoID.

Autonomous driving requires generating safe and reliable trajectories from complex multimodal inputs. Traditional modular pipelines separate perception, prediction, and planning, while recent end-to-end (E2E) systems learn them jointly. Vision-language models (VLMs) further enrich this paradigm by introducing cross-modal priors and commonsense reasoning, yet current VLM-based planners face three key challenges: (i) a mismatch between discrete text reasoning and continuous control, (ii) high latency from autoregressive chain-of-thought decoding, and (iii) inefficient or non-causal planners that limit real-time deployment. We propose ColaVLA, a unified vision-language-action framework that transfers reasoning from text to a unified latent space and couples it with a hierarchical, parallel trajectory decoder. The Cognitive Latent Reasoner compresses scene understanding into compact, decision-oriented meta-action embeddings through ego-adaptive selection and only two VLM forward passes. The Hierarchical Parallel Planner then generates multi-scale, causality-consistent trajectories in a single forward pass. Together, these components preserve the generalization and interpretability of VLMs while enabling efficient, accurate and safe trajectory generation. Experiments on the nuScenes benchmark show that ColaVLA achieves state-of-the-art performance in both open-loop and closed-loop settings with favorable efficiency and robustness.

Large multimodal models (LMMs) have achieved impressive performance on multimodal reasoning, becoming crucial technology for the advancement of intelligent question-answering systems. In real-world educational scenarios, effective teaching extends far beyond providing answers. Experienced teachers analyze students' incorrect answers to trace underlying errors and provide corrective feedback, termed educational diagnostic reasoning, a capability that remains under-explored in existing LMMs. To bridge this research gap, we introduce Edudiag benchmark, requiring LMMs to reconstruct erroneous reasoning chains from incorrect answers and generate corrective feedback. Through an AI-assisted annotation pipeline with rigorous human verification, we create 8K erroneous reasoning chains and corresponding feedback, spanning three representative educational domains: commonsense, science, and mathematics. Extensive evaluation across 28 leading LMMs highlights Edudiag as a challenging testbed, where even leading proprietary LMMs struggle on it and supervised fine-tuning (SFT) on open-source LMMs achieves marginal performance gains. Moreover, we conduct analysis experiments and identify three critical insights for educational diagnostic reasoning: (i) Effective error tracing remains the primary bottleneck, while SFT models still fail to reversely identify errors that commonly occur. (ii) Group relative policy optimization (GRPO) mitigates this bottleneck and boosts performance. (iii) LMMs optimized with GRPO can generate plausible yet challenging distractors for multiple-choice questions based on their self-constructed erroneous reasoning chains. We believe Edudiag provides a new direction for evaluating the advanced LMMs.

Current multimodal image retrieval benchmarks focus on relatively simple queries where target images are either described directly or by simple composition with an input image. When retrieval requires complex reasoning to determine the target image, the task becomes significantly more challenging, yet standardized benchmarks for this setting do not exist. To fill this gap, we introduce RMIR, a benchmark dataset of 1634 queries requiring reasoning across three categories: functional (object affordances), temporal (time-based relationships), and causal (cause-effect reasoning). Each query combines visual and textual inputs that demand robust visual understanding together with logical inference, beyond surface-level matching, to identify correct target images. In addition to the dataset itself, we present a pipeline to generate the dataset, which can be used to generate additional reasoning-intensive retrieval data at scale. Evaluation of state-of-the-art models on RMIR reveals significant performance gaps, with the best model achieving only 46.53% recall@20 averaged across reasoning categories. Our evaluation also shows that generative embedding models with explicit reasoning substantially outperform discriminative approaches, with reasoning-aware training proving more impactful than model scale. Our systematic analysis exposes fundamental limitations in current multimodal retrieval systems and establishes RMIR as a challenging testbed for developing multimodal, reasoning-capable retrieval models. Our dataset and code are available at https://github.com/amazon-science/rmir

Chain-of-Thought (CoT) has recently shown encouraging progress in the vision language model. However, the pure-vision CoT (i.e., chain-of-vision) has been underexplored in visual in-context learning. In this paper, we introduce Diffusion Guided Chain-of-Vision, which integrates an explicit chain-of-thought process into autoregressive vision models through vision prior from pre-trained diffusion models. Concretely, we find that pre-trained diffusion models induce a reliable probability flow in image space, where intermediate images sampled along this flow exhibit visual coherence and serve as task-free, chain-of-vision supervision for pure-vision autoregressive models. Extensive experiments on diverse vision tasks and multi-scale models validate the effectiveness of our proposed method for visual in-context learning. Code and dataset will be publicly available.