用球形编码器加速图像修复，首迭代即出清晰结果
SP$^3$: Spherical Priors for Plug-and-Play Restoration
- 用球形编码器替代传统去噪器，构建紧致潜在空间投影
- 首轮迭代即生成清晰图像，速度比顶尖方法快3到630倍
- 无需梯度计算，适合实时修复场景，支持任意时间输出
本文提出SP³，一种新型即插即用图像修复算法，通过用球形编码器（SE）作为生成先验，取代传统去噪器，加速最大后验估计。SP³利用SE紧致的潜在空间，近似难以计算的近端先验步骤，实现对自然图像流形的鲁棒投影。结合半二次分裂法的闭式数据一致性步骤，交替进行优化，无需推理时计算梯度，确保稳定收敛。该独特设计实现了“任意时间”修复能力，从第一轮迭代即可生成清晰、合理图像。在多种图像修复任务上的评估表明，SP³在感知质量上媲美最先进的零样本扩散与流模型，同时速度提升3至630倍。
原文摘要 · Abstract (English)
In this paper, we introduce SP$^3$, a novel Plug-and-Play algorithm that accelerates maximum a posteriori image restoration by replacing denoisers with Spherical Encoders (SE) as generative priors. SP$^3$ approximates the intractable proximal prior step by utilizing the SE tightly structured latent space as a robust projection onto the natural image manifold. Alternating this projection with a closed-form data-consistency step, via Half-Quadratic Splitting, achieves stable convergence without requiring gradient computation during inference. This unique formulation unlocks "anytime" restoration capabilities, producing sharp, plausible images from the first iteration. Evaluations across a variety of image restoration tasks demonstrate that SP$^3$ achieves perceptual quality comparable to state-of-the-art zero-shot diffusion and flow methods while being $3$-$630\times$ faster.
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。