arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2303.02584cs.CV
收藏

超分辨率神经算子

Super-Resolution Neural Operator

Beijing University of Posts and Telecommunications(北京邮电大学)

Min Wei, Xuesong Zhang

摘要:

我们提出了超分辨率神经算子(SRNO),这是一种深度算子学习框架,可以从低分辨率(LR)对应物中解析任意尺度的高分辨率(HR)图像。SRNO将LR-HR图像对视为用不同网格大小近似的连续函数,学习相应函数空间之间的映射。从近似理论的角度来看,SRNO首先将LR输入嵌入到更高维的潜在表示空间中,试图捕获足够的基函数,然后使用核积分机制迭代近似隐式图像函数,然后进行最后的降维步骤以生成目标坐标处的RGB表示。SRNO区别于以往连续SR工作的主要特点是:1)每一层的核积分通过Galerkin型注意力有效地实现,它在空间域具有非局部性,因此有利于无网格连续体;以及2)多层注意力体系结构允许动态潜在基础更新,这对于SR问题来说是至关重要的,以从LR图像中“产生幻觉”高频信息。实验表明,SRNO优于现有的连续SR方法的准确性和运行时间。我们的代码在https://github.com/2y7c3/Super-Resolution-Neural-Operator

英文摘要:

We propose Super-resolution Neural Operator (SRNO), a deep operator learning framework that can resolve high-resolution (HR) images at arbitrary scales from the low-resolution (LR) counterparts. Treating the LR-HR image pairs as continuous functions approximated with different grid sizes, SRNO learns the mapping between the corresponding function spaces. From the perspective of approximation theory, SRNO first embeds the LR input into a higher-dimensional latent representation space, trying to capture sufficient basis functions, and then iteratively approximates the implicit image function with a kernel integral mechanism, followed by a final dimensionality reduction step to generate the RGB representation at the target coordinates. The key characteristics distinguishing SRNO from prior continuous SR works are: 1) the kernel integral in each layer is efficiently implemented via the Galerkin-type attention, which possesses non-local properties in the spatial domain and therefore benefits the grid-free continuum; and 2) the multilayer attention architecture allows for the dynamic latent basis update, which is crucial for SR problems to "hallucinate" high-frequency information from the LR image. Experiments show that SRNO outperforms existing continuous SR methods in terms of both accuracy and running time. Our code is at https://github.com/2y7c3/Super-Resolution-Neural-Operator

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。

↑