arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2501.09054eess.IVcs.GR
收藏

NeurOp-Diff:基于神经算子扩散的连续遥感图像超分辨率

NeurOp-Diff:Continuous Remote Sensing Image Super-Resolution via Neural Operator Diffusion

Guangdong Laboratory of Artificial Intelligence and Digital Economy (SZ)(广东省人工智能与数字经济实验室);Shenzhen University(深圳大学)

Zihao Xu, Yuzhi Tang, Bowen Xu, Qingquan Li

AI总结:

针对遥感数据分辨率低的问题,提出NeurOp-Diff,利用神经算子学习任意尺度表示并作为先验条件引导扩散模型去噪,有效解决了伪影与过度平滑问题,实现了高质量的连续超分辨率。

摘要:

大多数公开的遥感数据分辨率低,限制了其实际应用。为此,我们提出了一种由神经算子引导的扩散模型,用于连续遥感图像超分辨率(NeurOp-Diff)。神经算子用于学习任意尺度的分辨率表示,将低分辨率(LR)图像编码为高维特征,随后将其作为先验条件引导扩散模型进行去噪。这有效解决了现有超分辨率(SR)方法中存在的伪影和过度平滑问题,能够生成高质量的连续超分辨率图像。具体而言,我们通过缩放因子s调整超分辨率尺度,使模型能够适应不同的超分辨率倍率。此外,在多个数据集上的实验证明了NeurOp-Diff的有效性。代码可在https://github.com/zerono000/NeurOp-Diff获取。

英文摘要:

Most publicly accessible remote sensing data suffer from low resolution, limiting their practical applications. To address this, we propose a diffusion model guided by neural operators for continuous remote sensing image super-resolution (NeurOp-Diff). Neural operators are used to learn resolution representations at arbitrary scales, encoding low-resolution (LR) images into high-dimensional features, which are then used as prior conditions to guide the diffusion model for denoising. This effectively addresses the artifacts and excessive smoothing issues present in existing super-resolution (SR) methods, enabling the generation of high-quality, continuous super-resolution images. Specifically, we adjust the super-resolution scale by a scaling factor s, allowing the model to adapt to different super-resolution magnifications. Furthermore, experiments on multiple datasets demonstrate the effectiveness of NeurOp-Diff. Our code is available at https://github.com/zerono000/NeurOp-Diff.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑