通過擴散反演的任意步驟圖像超分辨率

Arbitrary-steps Image Super-resolution via Diffusion Inversion

December 12, 2024

作者: Zongsheng Yue, Kang Liao, Chen Change Loy

cs.AI

摘要

本研究提出了一種基於擴散反演的新圖像超分辨率（SR）技術，旨在利用大型預訓練擴散模型中所包含的豐富圖像先驗信息來提高SR性能。我們設計了一種部分噪聲預測策略，用於構建擴散模型的中間狀態，作為起始採樣點。我們方法的核心是一個深度噪聲預測器，用於估計前向擴散過程的最佳噪聲映射。一旦訓練完成，該噪聲預測器可用於部分初始化沿著擴散軌跡的採樣過程，生成理想的高分辨率結果。與現有方法相比，我們的方法提供了一種靈活且高效的採樣機制，支持從一到五個任意數量的採樣步驟。即使僅進行一次採樣步驟，我們的方法表現優越或與最近的最先進方法相當。代碼和模型可在https://github.com/zsyOAOA/InvSR 公開獲得。

English

This study presents a new image super-resolution (SR) technique based on diffusion inversion, aiming at harnessing the rich image priors encapsulated in large pre-trained diffusion models to improve SR performance. We design a Partial noise Prediction strategy to construct an intermediate state of the diffusion model, which serves as the starting sampling point. Central to our approach is a deep noise predictor to estimate the optimal noise maps for the forward diffusion process. Once trained, this noise predictor can be used to initialize the sampling process partially along the diffusion trajectory, generating the desirable high-resolution result. Compared to existing approaches, our method offers a flexible and efficient sampling mechanism that supports an arbitrary number of sampling steps, ranging from one to five. Even with a single sampling step, our method demonstrates superior or comparable performance to recent state-of-the-art approaches. The code and model are publicly available at https://github.com/zsyOAOA/InvSR.