Implicit Diffusion Models for Continuous Super-Resolution
Sicheng Gao1, Xuhui Liu, Bohan Zeng, Sheng Xu1, Yanjing Li1, Xiaoyan Luo1 Jianzhuang Liu2, Xiantong Zhen3, Baochang Zhang1,4 1Beihang University 2Shenzhen Institute of Advanced Technology, Shenzhen, China 3United Imaging 4Zhongguancun Laboratory, Beijing, China These authors contributed equally.Corresponding Author:
Abstract
Image super-resolution (SR) has attracted increasing attention due to its widespread applications. However, current SR methods generally suffer from over-smoothing and artifacts, and most work only with fixed magnifications. This paper introduces an Implicit Diffusion Model (IDM) for high-fidelity continuous image super-resolution. IDM integrates an implicit neural representation and a denoising diffusion model in a unified end-to-end framework, where the implicit neural representation is adopted in the decoding process to learn continuous-resolution representation. Furthermore, we design a scale-adaptive conditioning mechanism that consists of a low-resolution (LR) conditioning network and a scaling factor. The scaling factor regulates the resolution and accordingly modulates the proportion of the LR information and generated features in the final output, which enables the model to accommodate the continuous-resolution requirement. Extensive experiments validate the effectiveness of our IDM and demonstrate its superior performance over prior arts. The source code will be available at https://github.com/Ree1s/IDM.
中文速览
图像超分辨率（Super-Resolution, SR）长期面临两大痛点：生成结果过于平滑、有伪影，以及只能输出固定放大倍数的图像。本文提出了一种隐式扩散模型（Implicit Diffusion Model, IDM），将隐式神经表示（Implicit Neural Representation）嵌入去噪扩散模型的解码过程，让模型能够在连续分辨率范围内生成高保真超分图像，无需为每个放大倍数单独训练或设计多阶段流程。为了让模型灵活应对任意放大比例，研究者还设计了一套自适应缩放条件机制，通过一个缩放因子动态调节低分辨率原始信息与扩散模型生成细节之间的比例，整个框架端到端训练，无需额外先验知识。在自然图像和人脸图像的多个基准测试上，IDM 在感知质量和定量指标上均超越了此前的方法，证明了将扩散模型的细节生成能力与隐式表示的连续分辨率优势结合起来，是解决超分辨率实际应用限制的一条有效路径。
原文 arXiv:2303.16491;中英对照 + 大白话阅读 https://aha.fim.ai/paper/2303.16491v2