CVPR 2026 论文列表

Recent advancements adopt online reinforcement learning (RL) from LLMs to text-to-image rectified flow diffusion models for reward alignment. The use of group-level rewards successfully aligns the model with the targeted reward. However, it faces challenges including low efficiency, dependency on stochastic samplers, and reward hacking. The problem is that rectified flow models are fundamentally different from LLMs: 1) For efficiency, online image sampling takes much more time and dominates the time of training. 2) For stochasticity, rectified flow is deterministic once the initial noise is fixed. Aiming at these problems and inspired by the effects of group-level rewards from LLMs, we design Group-level Direct Reward Optimization (GDRO). GDRO is a new post-training paradigm for group-level reward alignment that combines the characteristics of rectified flow models. Through rigorous theoretical analysis, we point out that GDRO supports full offline training that saves the large time cost for image rollout sampling. Also, it is diffusion-sampler-independent, which eliminates the need for the ODE-to-SDE approximation to obtain stochasticity. We also empirically study the reward hacking trap that may mislead the evaluation, and involve this factor in the evaluation using a corrected score that not only considers the original evaluation reward but also the trend of reward hacking. Extensive experiments demonstrate that GDRO effectively and efficiently improves the reward score of the diffusion model through group-wise offline optimization across the OCR and GenEval tasks, while demonstrating strong stability and robustness in mitigating reward hacking.

Co-speech gesture generation has significantly advanced human-computer interaction, yet speaker movements remain constrained due to the omission of text-driven non-spontaneous gestures (e.g., bowing while talking). Existing methods face two key challenges: 1) the semantic prior gap due to the lack of descriptive text annotations in gesture datasets, and 2) the difficulty in achieving coordinated multimodal control over gesture generation. To address these challenges, this paper introduces CoordSpeaker, a comprehensive framework that enables coordinated caption-empowered co-speech gesture synthesis. Our approach first bridges the semantic prior gap through a novel gesture captioning framework, leveraging a motion-language model to generate descriptive captions at multiple granularities. Building upon this, we propose a conditional latent diffusion model with unified cross-dataset motion representation and a hierarchically controlled denoiser to achieve highly controlled, coordinated gesture generation. CoordSpeaker pioneers the first exploration of gesture understanding to tackle the semantic gap in gesture generation while offering a novel perspective of bidirectional gesture-text mapping. Extensive experiments demonstrate that our method produces high-quality gestures that are both rhythmically synchronized with speeches and semantically coherent with arbitrary captions, achieving superior performance with higher efficiency compared to existing approaches.

We present Lighting in Motion (LiMo), a diffusion-based approach to spatiotemporal lighting estimation. LiMo targets both realistic high-frequency detail prediction and accurate illuminance estimation. To account for both, we propose generating a set of mirrored and diffuse spheres at different exposures, based on their 3D positions in the input. Making use of diffusion priors, we fine-tune powerful existing diffusion models on a large-scale customized dataset of indoor and outdoor scenes, paired with spatiotemporal light probes. For accurate spatial conditioning, we demonstrate that depth alone is insufficient and we introduce a new geometric condition to provide the relative position of the scene to the target 3D position. Finally, we combine diffuse and mirror predictions at different exposures into a single HDRI map leveraging differentiable rendering.We thoroughly evaluate our method and design choices to establish LiMo as state-of-the-art for both spatial control and prediction accuracy.

Early prediction of Mild Cognitive Impairment (MCI) conversion is hampered by a trade-off between immediacy--making fast predictions from a single baseline sMRI--and accuracy--leveraging longitudinal scans to capture disease progression. We propose MCI-Diff, a diffusion-based framework that synthesizes clinically plausible future sMRI representations directly from baseline data, achieving both real-time risk assessment and high predictive performance. First, a multi-task sequence reconstruction strategy trains a shared denoising network on interpolation and extrapolation tasks to handle irregular follow-up sampling and learn robust latent trajectories. Second, an LLM-driven "linguistic compass" is introduced for clinical plausibility sampling: generated feature candidates are quantized, tokenized, and scored by a fine-tuned language model conditioned on expected structural biomarkers, guiding autoregressive generation toward realistic disease patterns. Experiments on ADNI and AIBL cohorts show that MCI-Diff outperforms state-of-the-art baselines, improving early conversion accuracy by 5-12%.

Recent diffusion and flow matching models have demonstrated strong capabilities in image generation and editing by progressively removing noise through iterative sampling. While this enables flexible inversion for semantic-preserving edits, few-step sampling regimes suffer from poor forward process approximation, leading to degraded editing quality. Existing few-step inversion methods often rely on pretrained generators and auxiliary modules, limiting scalability and generalization across different architectures.To address these limitations, we propose BiFM (Bidirectional Flow Matching), a unified framework that jointly learns generation and inversion within a single model. BiFM directly estimates average velocity fields in both "image \to noise" and "noise \to image" directions, constrained by a shared instantaneous velocity field derived from either predefined schedules or pretrained multi-step diffusion models. Additionally, BiFM introduces a novel training strategy using continuous time-interval supervision, stabilized by a bidirectional consistency objective and a lightweight time-interval embedding. This bidirectional formulation also enables one-step inversion and can integrate seamlessly into popular diffusion and flow matching backbones.Across diverse image editing and generation tasks, BiFM consistently outperforms existing few-step approaches, achieving superior performance and editability. Our code and models will be released upon acceptance.

We demonstrate that in knowledge distillation for diffusion models, the teacher network's highly complex denoising process--stemming from its substantially larger capacity--poses a significant challenge for the student model to faithfully mimic. To address this problem, we propose a coarse-to-fine distillation framework with LInear FiTting-based distillation (LIFT) and Piecewise Local Adaptive Coefficient Estimation (PLACE). First, LIFT decomposes the objective into a "coarse" alignment and a "fine" refinement. The student is then trained on coarse alignment before proceeding to hard refinement. Second, PLACE extends LIFT to address spatially non-uniform errors by partitioning outputs into error-based groups, providing locally adaptive guidance. Our experiments show that LIFT and PLACE is effective across diffusion spaces (image/latent), backbones (U-Net/DiT), tasks (unconditional/conditional), datasets, and even extends to flow-based models such as MMDiT (SD3). Furthermore, under extreme compression with a 1.3M-parameter student (only 1.6% of the teacher), conventional KD fails to provide sufficient guidance for stable training, with FID scores often degrading to 50-200+, but our method remains stably convergent and achieves an FID of 15.73.

Flow matching models have emerged as a powerful framework for realistic image generation by learning to reverse a corruption process that progressively adds Gaussian noise. However, because noise is injected in the latent domain, its impact on different frequency components is non-uniform. As a result, during inference, flow matching models tend to generate low-frequency components (global structure) in the early stages, while high-frequency components (fine details) emerge only later in the reverse process. Building on this insight, we propose Frequency-Aware Flow Matching (FreqFlow), a novel approach that explicitly incorporates frequency-aware conditioning into the flow matching framework via time-dependent adaptive weighting. We introduce a two-branch architecture: (1) a frequency branch that separately processes low- and high-frequency components to capture global structure and refine textures and edges, and (2) a spatial branch that synthesizes images in the latent domain, guided by the frequency branch's output. By explicitly integrating frequency information into the generation process, FreqFlow ensures that both large-scale coherence and fine-grained details are effectively modeled--low-frequency conditioning reinforces global structure, while high-frequency conditioning enhances texture fidelity and detail sharpness. On the class-conditional ImageNet-256 generation benchmark, our method achieves state-of-the-art performance with an FID of 1.38, surpassing the prior diffusion model DiT and flow matching model SiT by 0.79 and 0.58 FID, respectively. Code is available at https://github.com/OliverRensu/FreqFlow.

We propose PatchScene, a novel diffusion-based framework for large-scale LiDAR scene completion. Unlike existing methods that rely on global latent representations or dense voxel grids, PatchScene adopts a patch-based voxel diffusion paradigm that explicitly generates fine-grained geometry within localized 3D regions. To ensure coherent reconstruction at both spatial and temporal scales, we introduce a confidence-guided spatio-temporal fusion mechanism that integrates overlapping patches and adjacent frames in a unified generative process. Furthermore, we design an Annular-Flow diffusion strategy that leverages the radial density pattern of LiDAR scans to progressively propagate high-fidelity information from near-range to far-range regions, enabling spatially unbounded scene completion. Extensive experiments on the SemanticKITTI benchmark demonstrate that PatchScene achieves state-of-the-art performance across all standard metrics, surpassing previous approaches in both geometric accuracy and temporal consistency. Remarkably, the model trained on 20 m LiDAR ranges generalizes effectively to 50 m scenes without retraining, highlighting its strong scalability and generalization capability for real-world autonomous driving applications.

Despite the inherent advantages of thermal infrared(TIR) imaging, large-scale data collection and annotation remain a major bottleneck for TIR-based perception. A practical alternative is to synthesize pseudo TIR data via image translation; however, most RGB-to-TIR approaches heavily rely on RGB-centric priors that overlook thermal physics, yielding implausible heat distributions. In this paper, we introduce TherA, a controllable RGB-to-TIR translation framework that produces diverse and thermally plausible images at both scene and object level. TherA couples TherA-VLM with a latent-diffusion-based translator. Given a single RGB image and a user-prompted condition pair, TherA-VLM yields a thermal-aware embedding that encodes scene, object, material, and heat-emission context reflecting the input scene-condition pair. Conditioning the diffusion model on this embedding enables realistic TIR synthesis and fine-grained control across time of day, weather, and object state. Compared to other baselines, TherA achieves state-of-the-art translation performance, demonstrating improved zero-shot translation performance up to 33% increase averaged across all metrics.

This paper studies the training-testing discrepancy (a.k.a. exposure bias) problem for improving the diffusion models. During training, the input of a prediction network at the training timestep is the corresponding ground-truth noisy data that is an interpolation of the noise and the data, and during testing, the input is the generated noisy data. We present a novel training approach, named MixFlow, for improving the training performance. Our approach is motivated by the Slow Flow phenomenon: the ground-truth interpolation that is the nearest to the generated noisy data at a given sampling timestep is observed to correspond to a higher-noise timestep (termed slowed timestep), i.e., the corresponding ground-truth timestep is slower than the sampling timestep. MixFlow leverages the interpolations at the slowed timesteps, named slowed interpolation mixture, for post-training the prediction network at each training timestep. Experiments over class-conditional image generation (including SiT, REPA, and RAE) and text-to-image generation, validate the effectiveness of our approach. Our approach MixFlow over the RAE models achieve strong generation results on ImageNet: 1.43 FID (without guidance) and 1.10 (with guidance) at 256 x256, and 1.55 FID (without guidance) and 1.10 (with guidance) at 512 x512.

Learning transferable knowledge from unlabeled video data and applying it in new environments is a fundamental capability of intelligent agents. This work presents VideoWorld 2, which extends VideoWorld and provides the first investigation of learning transferable knowledge for complex, long-horizon tasks directly from raw real-world videos. At its core, VideoWorld 2 introduces a dynamics-enhanced Latent Dynamics Model (dLDM) that decouples action dynamics from visual appearance: a pretrained video diffusion model handles visual appearance modeling, enabling the dLDM to learn latent codes that focus on compact and meaningful task-related dynamics. These latent codes are then modeled autoregressively to learn task policies and support long-horizon reasoning. We evaluate VideoWorld 2 on challenging real-world handcraft making tasks, where prior video generation and latent-dynamics models struggle to operate reliably. Remarkably, VideoWorld 2 achieves up to 70% improvement in task success rate and produces coherent long execution videos. In robotics, we show that VideoWorld 2 can acquire transferable manipulation knowledge from the Open-X dataset, which substantially improves task performance on CALVIN, demonstrating strong cross-domain generalization. This study reveals the potential of learning transferable world knowledge directly from raw videos, with all code, data, and models open-sourced for further research.

Diffusion models have demonstrated exceptional generative capabilities but are computationally intensive, posing significant challenges for deployment in resource-constrained or latency-sensitive environments.Quantization offers an effective means to reduce model size and computational cost, with post-training quantization (PTQ) being particularly appealing due to its compatibility with pre-trained models without requiring retraining or training data.However, existing PTQ methods for diffusion models often rely on manual, architecture-specific heuristics that limit their generalizability and hinder integration with industrial deployment pipelines.To address these limitations, we propose SegQuant, a deployment-aware quantization framework that adaptively combines complementary techniques to enhance cross-model versatility.SegQuant consists of a segment-aware, graph-based quantization strategy (SegLinear) that captures structural semantics and spatial heterogeneity, along with a dual-scale quantization scheme (DualScale) that preserves polarity-asymmetric activations using a hardware-native dual-path computation, avoiding performance penalties from custom implementations, which is crucial for maintaining visual fidelity in generated outputs.SegQuant is broadly applicable beyond Transformer-based diffusion models, achieving strong performance while ensuring seamless compatibility with mainstream deployment tools.

Cross-modal ship re-identification (ReID) between optical and synthetic aperture radar (SAR) imagery has recently emerged as a critical yet underexplored task in maritime intelligence and surveillance. However, the substantial modality gap between optical and SAR images poses a major challenge for robust identification. To address this issue, we propose MOS, a novel framework designed to Mitigate the Optical-SAR modality gap and achieve modality-consistent feature learning for optical-SAR cross-modal ship ReID. MOS consists of two core components: (1) Modality-Consistent Representation Learning (MCRL) applies denoise SAR image procession and a class-wise modality alignment loss to align intra-identity feature distributions across modalities. (2) Cross-modal Data Generation and Feature fusion (CDGF) leverages a brownian bridge diffusion model to synthesize cross-modal samples, which are subsequently fused with original features during inference to enhance alignment and discriminability. Extensive experiments on the HOSS ReID dataset demonstrate that MOS significantly surpasses state-of-the-art methods across all evaluation protocols, achieving notable improvements of +3.0%, +6.2%, and +16.4% in R1 accuracy under the ALL to ALL,Optical to SAR, and SAR to Optical settings, respectively. The source codes are available at https://github.com/yjzhao1019/MOS.

We present a method for harmonizing the lighting of a foreground video to match a target background scene, adjusting shadows, color tone, and illumination intensity (relightful harmonization). Unlike images, acquiring labeled data for videos, where identical motions are recorded under different lighting conditions, is practically infeasible and non-scalable. While one way to create such paired data is to apply existing image-based harmonization models frame by frame to a video, the resulting outputs often suffer from significant temporal jitters. We overcome this problem by introducing a novel lighting deflickering model that can stabilize the global and local lighting flickering artifacts. Our video diffusion model learns from these upgraded deflickered data with a volume of real and synthetic videos to generate high-quality video harmonization results. We further propose an asymmetric alpha mask conditioning technique to learn the clean boundaries from real videos. Experiments demonstrate that our model achieves strong temporal coherence, naturalness, cleaner boundaries, and physically meaningful lighting behavior, while maintaining strong relighting expressiveness compared to prior image-based and video-based harmonization methods.

Despite recent progress, diffusion-based video frame interpolation methods still struggle with large, complex motions, resulting in discontinuous motions and inconsistent object appearances across frames. We observe that these limitations arise from both the current full-sequence interpolation strategy and the pixel reconstruction training objective. To solve these challenges, we propose ARVFI, a novel video diffusion-based interpolation method for large complex motion interpolation. Instead of generating all intermediate frames simultaneously, ARVFI interpolates in an autoregressive manner from two input frames to the middle ones. Thus, ARVFI interpolates a frame that is further away from the inputs based on all previous interpolation results, resulting in smoother motion transitions and better temporal consistency. Additionally, ARVFI further utilizes DINOv3 features as motion representations, which provide high-level semantics for accurate motion estimation, compared with a simple pixel-level loss. With all these designs, ARVFI generates the intermediate DINOv3 features first and then the frames with an effective conditional generation method for frames. Our ARVFI consistently outperforms existing methods with superior interpolation accuracy and visual quality.

Recent progress in video diffusion models has enabled remarkable generative fidelity, yet leveraging these priors for restoration remains limited by the strong coupling between conditional and unconditional branches in standard classifier-free guidance. We introduce a training-free framework that enhances distorted and low-resolution videos by decoupling these signals in time. Our proposed Decoupled Time Guidance (DTG) evaluates the unconditional branch at a cleaner diffusion timestep, providing a lookahead prior that preserves geometry while suppressing replication of warped content. This temporal bias is annealed throughout sampling, allowing the model to transition from structure correction to detail refinement without retraining. Combined with any off-the-shelf restoration module in a plug-and-play manner, our approach improves perceptual coherence and restores plausible structure in AI-generated and real-world videos alike. To facilitate evaluation, we curate GenWarp480, a benchmark of 4,400 distorted 480p videos synthesized from diverse text-to-video models. GenWarp480 focuses on characteristic generative degradations such as warped faces, body misalignments, and spatial artifacts, providing a purpose-built testbed for assessing robustness to generative errors. Extensive experiments demonstrate that our method achieves significant improvements in structural fidelity and temporal stability without any model training.

Diffusion models achieve remarkable generation quality, yet face a fundamental challenge known as memorization, where generated samples can replicate training samples exactly. We develop a theoretical framework to explain this phenomenon by showing that the empirical score function (the score function corresponding to the empirical distribution) is a weighted sum of the score functions of Gaussian distributions, in which the weights are sharp softmax functions. This structure causes individual training samples to dominate the score function, resulting in sampling collapse. In practice, approximating the empirical score function with a neural network can partially alleviate this issue and improve generalization. Our theoretical framework explains why: In training, the neural network learns a smoother approximation of the weighted sum, allowing the sampling process to be influenced by local manifolds rather than single points. Leveraging this insight, we propose two novel methods to further enhance generalization: (1) Noise Unconditioning enables each training sample to adaptively determine its score function weight to increase the effect of more training samples, thereby preventing single-point dominance and mitigating collapse. (2) Temperature Smoothing introduces an explicit parameter to control the smoothness. By increasing the temperature in the softmax weights, we naturally reduce the dominance of any single training sample and mitigate memorization. Experiments across multiple datasets validate our theoretical analysis and demonstrate the effectiveness of the proposed methods in improving generalization while maintaining high generation quality.

Modern optical flow estimation, though empowered by recent deep neural architectures, remains rooted in the discrete correspondence paradigm inherited from classical vision. Most networks infer frame-to-frame displacements, capturing where pixels move but not how motion evolves continuously through time. Yet physical motion in the real world follows smooth dynamics governed by underlying velocity fields, as long established in fluid mechanics and transport theory. To bridge this gap, we introduce Optical Flow Matching (OFM), a continuous formulation that learns a time-dependent velocity field to transport pixel coordinates along motion distribution coherent trajectories. A key component of our OFM is Triangle Velocities Synergy (TVS), a neat yet pragmatic geometric mechanism that provides a stable and physically meaningful velocity construction, ensuring that continuous transport remains well-defined. Combined with an Euler-based ODE solver, OFM yields flow fields that are temporally smooth, geometrically consistent, and process-interpretable. Experiments on Sintel, KITTI, and Spring demonstrate that OFM achieves state-of-the-art accuracy, enhanced temporal stability, and notably stronger cross-dataset generalization, advancing optical flow estimation from correspondence inference to continuous dynamical reasoning.

Non-line-of-sight (NLOS) imaging seeks to recover hidden-scene information from indirect light transport beyond the direct line of sight. Existing NLOS methods can be broadly categorized into active and passive approaches. Active methods rely on controlled illumination and time-resolved sensors, but their dependence on specialized and expensive hardware limits practical deployment. Passive methods instead use steady-state or uncontrolled indirect reflections with conventional sensors, but often suffer from weak signals, unstable observations, and insufficient measurement constraints. In this paper, we present the Similarity-Likelihood Diffusion Network (SLD-Net) for corner-camera hidden-person imaging from multi-exposure wall observations. SLD-Net consists of two stages: DeLi-Inversion estimates an initial reconstruction and a pixel-wise precision map to construct a heteroscedastic pseudo-likelihood, and SiCo-Diffusion incorporates this likelihood into deterministic DDIM sampling and fuses it with the diffusion prior via annealed Bayesian precision updating, producing reproducible and measurement-consistent reconstructions. Experiments on the Reflect-Corridor and Reflect-Room datasets demonstrate consistent improvements over generic, physics-inspired, and NLOS-specific baselines. SLD-Net improves PSNR from 13.84 to 15.58 dB and reduces FID from 264.91 to 73.54 on Reflect-Corridor, and improves PSNR from 11.58 to 12.49 dB and reduces FID from 177.05 to 26.89 on Reflect-Room.