用生成压缩提升去噪感知真实感，避免过度平滑。
Perception-based Image Denoising via Generative Compression
- 从熵编码潜空间重建，强制低复杂度结构
- 结合LPIPS和Wasserstein距离，恢复真实纹理
- 支持率-失真-感知权衡，适合高噪声场景
图像去噪旨在去除噪声的同时保留结构细节和感知真实感，但基于失真的方法在强噪声和分布偏移下常导致过度平滑。本文提出一种基于生成压缩的感知去噪框架：通过熵编码的潜表示重建，强制低复杂度结构；生成解码器则利用学习的感知图像块相似性（LPIPS）损失和Wasserstein距离恢复真实纹理。提出两种互补方案：(i) 条件Wasserstein GAN（WGAN）压缩去噪器，显式控制率-失真-感知（RDP）权衡；(ii) 基于条件扩散的迭代去噪策略，由压缩潜变量引导。进一步在加性高斯噪声下建立了压缩最大似然去噪器的非渐近保证，包括重构误差和解码错误概率的上界。在合成与真实噪声基准测试中，该方法持续提升感知质量，同时保持竞争力的失真性能。
原文摘要 · Abstract (English)
Image denoising aims to remove noise while preserving structural details and perceptual realism, yet distortion-driven methods often produce over-smoothed reconstructions, especially under strong noise and distribution shift. This paper proposes a generative compression framework for perception-based denoising, where restoration is achieved by reconstructing from entropy-coded latent representations that enforce low-complexity structure, while generative decoders recover realistic textures via perceptual measures such as learned perceptual image patch similarity (LPIPS) loss and Wasserstein distance. Two complementary instantiations are introduced: (i) a conditional Wasserstein GAN (WGAN)-based compression denoiser that explicitly controls the rate-distortion-perception (RDP) trade-off, and (ii) a conditional diffusion-based reconstruction strategy that performs iterative denoising guided by compressed latents. We further establish non-asymptotic guarantees for the compression-based maximum-likelihood denoiser under additive Gaussian noise, including bounds on reconstruction error and decoding error probability. Experiments on synthetic and real-noise benchmarks demonstrate consistent perceptual improvements while maintaining competitive distortion performance.
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。