Lune

CVPR2023顶会

DR2: Diffusion-Based Robust Degradation Remover for Blind Face Restoration

Zhixin Wang, Ziying Zhang, Xiaoyun Zhang, Huangjie Zheng, Mingyuan Zhou, Ya Zhang, Yanfeng Wang

2023年份
42顶会引用

摘要

Ours Figure 1. DR2 uses Denoising Diffusion Probabilistic Models to remove degradation. The generative process is conditioned on the low-quality input after being diffused into a noisy status. As a result, DR2 predicts coarse faces x0 regardless of the degradation type. On severely degraded images, our final restoration results achieve high quality with fewer artifacts than previous arts [14, 48] .

问问这篇 Paper

智能体会读完全文。

Lune 把这篇 Paper 索引到了每一个公式,引用它的顶会 Paper 也一样。你提问,回答直接引用原文。

可以从这些问题问起

智能体调用

Luneget_paper_fulltext

在 Lune 里问

免费开始,无需绑卡

引用它的顶会 Paper42

问问它们各自怎么用它

它引用的顶会 Paper15

相关 Paper

黄昏的海面,两侧是细线勾勒的悬崖