用数据统计指导修复，无需训练就能高质量补全图像缺失区域。
Image Inpainting via Stochastic Dynamics
- 基于参考数据集的统计信息，直接引导像素演化修复
- 在多个数据集上优于传统方法，尤其在结构敏感区域表现佳
- 无需网络训练，适合对模型依赖低的场景
图像修复旨在恢复缺失区域的同时保持结构一致性。我们提出一种非参数化方法，无需网络训练，基于数据引导的随机动力学。从掩码图像出发，通过反向时间随机微分方程演化缺失像素，利用参考数据集直接估计核加权修正项。该经验修正项引导重建趋向数据分布的高密度区域，无需训练神经网络或拟合参数化密度模型。在MNIST、Fashion-MNIST和MVTec上的实验表明，该方法在PSNR、SSIM和视觉质量上均优于均值填充、Telea和泊松修复；在CelebA上仍具竞争力，能生成结构敏感遮挡的合理补全结果。这些结果证明了经验参考统计作为图像修复非参数先验的有效性。
原文摘要 · Abstract (English)
Image inpainting aims to recover missing regions while preserving structural consistency. We propose a non-parametric method without network training based on data-guided stochastic dynamics. Starting from a masked image, the missing pixels are evolved through a reverse-time stochastic differential equation with a kernel-weighted correction estimated directly from a reference dataset. This empirical correction guides the reconstruction toward high-density regions of the data distribution without training a neural network or fitting a parametric density model. Experiments on MNIST, Fashion-MNIST, and MVTec show that the proposed method outperforms Mean Fill, Telea, and Navier-Stokes inpainting in PSNR, SSIM, and visual quality. On CelebA, it remains competitive and produces plausible completions for structure-sensitive occlusions. These results demonstrate the effectiveness of empirical reference statistics as a non-parametric prior for image inpainting.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。