FlowPainter is introduced, a diffusion-based optical flow framework that reformulates dense-flow generation as confidence-guided soft inpainting and offers a practical way to integrate reliable discriminative priors with diffusion-based refinement for optical flow estimation.
Abstract
Existing optical flow methods broadly follow two paradigms: iterative optimization and diffusion-based estimation. Iterative methods, exemplified by RAFT, achieve high accuracy through recurrent refinement, but remain challenged by large displacements and complex motion. Diffusion-based methods introduce generative modeling and show promise in such ambiguous regions. However, existing diffusion models usually denoise the entire dense flow field from Gaussian noise, including simple regions where reliable motion can already be estimated by a lightweight network. This increases the denoising burden and may cause slow convergence and unstable training. To address this issue, we introduce FlowPainter, a diffusion-based optical flow framework that reformulates dense-flow generation as confidence-guided soft inpainting. FlowPainter employs a lightweight confidence-aware network to predict a rough flow and a pixel-wise confidence mask, distinguishing reliable simple regions from uncertain hard regions. The resulting simple-flow prior is used for confidence-based initialization and further injected into iterative denoising through confidence-gated residual guidance. With dynamically decaying guidance strength, FlowPainter stabilizes early denoising while preserving the flexibility of the diffusion model for late-stage detail refinement. Extensive experiments on public benchmarks, including Sintel, KITTI, and Spring, show that FlowPainter achieves strong accuracy under comparable training settings and converges more efficiently than existing diffusion-based optical flow methods, with notable gains on challenging benchmark splits. Our approach offers a practical way to integrate reliable discriminative priors with diffusion-based refinement for optical flow estimation. Our code is publicly available at https://github.com/mya012/FlowPainter.
RadioVIL is proposed, an efficient two-stage framework that reformulates joint radio map inpainting and zero-shot vehicle localization as a prior-guided physical inverse problem and unlocks accurate zero-shot vehicle localization directly from sparse radio maps, paving a robust way for ISAC at the 6G edge.
Ruixin Zhao, Xiucheng Wang, Qiming Zhang et al.· 0 citations
Results demonstrate the effectiveness of empirical reference statistics as a non-parametric prior for image inpainting by showing that the proposed method outperforms Mean Fill, Telea, and Navier-Stokes inpainting in PSNR, SSIM, and visual quality.
Jiaqi Kuang, Zihao Guo, Z. Qian· arXiv.org· 0 citations
ScaleResfusion is presented, a scalable diffusion framework for real-world image restoration built on pre-trained text-to-image rectified-flow models that achieves state-of-the-art performance with much higher efficiency and a knowledge-distillation pipeline to reduce sampling cost while maintaining restoration quality.
Image inpainting is a fundamental task in computer vision and multimedia processing. With the rapid development of denoising diffusion models, text-guided image inpainting has gradually become a mainstream research direction, enabling flexible content creation and localized semantic editing. Although existing text-guided diffusion inpainting approaches have greatly improved the generation quality and prompt alignment, they still face a core challenge: effectively balancing the fidelity preservation of unmasked regions and the semantic consistency of generated content in masked regions, especially suppressing spectral discontinuities and boundary artifacts caused by unreasonable frequency control. To address these problems, we propose Adaptive Frequency-Aware Diffusion(AFID), an adaptive framework for high-quality text-guided image inpainting. First, we design an Adaptive Frequency Threshold Network (AFTN) to dynamically predict multi-band frequency cutoffs according to mask information, denoising timesteps and text conditions, replacing manually designed fixed rules. Second, we propose a two-stage smooth spectral blending strategy to alleviate spectral abruptness and enhance the natural coherence between masked and unmasked areas. Third, we introduce a boundary ring smoothing module to further eliminate stitching artifacts near mask edges without excessive blurring. Experimental results show that our method achieves a leading performance on standard public benchmarks. We conduct comprehensive comparisons against prevailing state-of-the-art inpainting approaches, which solidly confirm our superiority in visual fidelity, intact-area preservation and text-prompt consistency. Extensive ablation studies verify the necessity of all three core modules and further analyze the impact of frequency thresholds and blending configurations on final restoration outcomes.
Text-guided image editing using rectified flow models such as Multimodal Diffusion Transformer (DiT) has demonstrated impressive generation quality. However, existing methods apply edits globally, inevitably modifying background regions unrelated to the intended semantic change. Our key observation is that the Text-to-Image (T2I) sub-block within the MM-DiT joint self-attention naturally encodes highly discriminative spatial localization signals. In this paper, we propose MaskFlow, a training-free framework for spatially localized image editing in rectified flow models. By strategically extracting and aggregating these attention maps from edit-relevant tokens during the standard ODE forward passes, MaskFlow automatically constructs a soft spatial mask. This mask confines semantic edits to the target region while perfectly preserving the original background. MaskFlow operates as a lightweight, plug-and-play extension without retraining the baseline model. Experiments on the FlowEdit benchmark show that MaskFlow reduces LPIPS by 28.1% relative to the baseline while maintaining competitive semantic alignment.
Trong-Tai Dam Vu, Vinh-Tiep Nguyen· International Conference on...· 0 citations
Observation Operator Diffusion is proposed, a unified framework that aligns both the supervision trajectory and feature refinement with the intrinsic recovery order of image structures and introduces GL-CoDA, a decoder that injects scale-specific Gaussian-Lanczos observations across decoding stages for coarse-to-fine feature refinement.
Shaojie Guo, Li-Chen Ma, Haoyang Tong et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
