Learning local Gaussian process regression for image super-resolution
Learning local Gaussian process regression for image super-resolution
复制标题
学习图像超分辨率的局部高斯过程回归
DOI：
10.1016/j.neucom.2014.11.064
复制
发表时间：
2015-04
期刊：
影响因子：
6
通讯作者：
Fan Jianping
中科院分区：
文献类型：
--
作者：
Li Jianmin;Qu Yanyun;Li Cuihua;Xie Yuan;Wu Yang;Fan Jianping
基于学习的超分辨率（SR）方法预测高分辨率像素值但不直接提供不确定性的估计，通常是非概率的并且具有有限的泛化能力。高斯过程可以为导出具有显式不确定性模型的回归技术提供框架，但高斯过程回归（GPR）具有耗时的显著缺点。GPR的计算复杂度是训练样本数量的立方，这对于大规模训练集来说是非常昂贵的。在这篇文章中，我们提出了学习局部GPR的图像SR。两个算法的开发，以支持局部GPR的超分辨率。首先提出了一种基于数据驱动的超分辨率算法，该算法在中等规模的输入训练数据集上为每个LR块学习局部GPR模型。为了进一步提高算法的运行速度，提出了一种基于原型的超分辨探地雷达算法。该算法比数据驱动的GPR算法快一个数量级，因为它是为图像块的原型而不是为每个图像块建模。因此，局部回归工作被大大减少，仅为每个LR图像块找到最近的原型，并将其对应的预先计算的投影矩阵应用于超分辨率预测。我们的算法具有更强的鲁棒性和可用性，因为它们提供了一种公式化的方法来自动学习用于优化协方差函数的超参数，而大多数最先进的超分辨率方法只能以交叉验证的方式利用这些参数。此外，我们的算法提供了在测试点，有利于像素的后处理的置信度值。我们的算法进行了评估，在超分辨率文献中广泛使用的流行数据集，实验结果表明，我们提出的算法的效率和有效性是比较几个国家的最先进的超分辨率方法。
Learning based super-resolution (SR) methods, which predict the high-resolution pixel values but not directly provide an estimation of uncertainty, are typically non-probabilistic and have limited generalization ability. Gaussian processes can provide a framework for deriving regression techniques with explicit uncertainty models, but Gaussian Process Regression (GPR) has a significant drawback in being time consuming. The computational complexity of GPR is cubic in the number of training examples, which is prohibitively expensive for a large-scale training set. In this article, we have proposed learning local GPR for image SR. Two algorithms are developed to support local GPR for super resolution. A data-driven GPR based super-resolution algorithm is first developed to learn a local GPR model for every LR patch on an input oriented training dataset with moderate size. In order to further improve the running speed, a prototype based GPR algorithm is developed for super resolution. The proposed algorithm is about one-order faster than the data-driven GPR solution because it makes models for the prototypes of image patches rather than for each image patch. Thus, the local regression efforts are greatly reduced to just finding the nearest prototype for each LR image patch and applying its corresponding pre-computed projective matrix for super-resolution prediction. Our algorithms have greater robustness and usability as they provide a formularized way to automatically learn the hyper-parameters introduced for optimizing the covariance function, while most of the state-of-the-art super-resolution methods could only utilize these parameters in a cross-verification way. Moreover, our algorithms offer confidence values at the test points which benefit the pixels’ post-processing. Our algorithms are evaluated on popular datasets that are widely used in the super-resolution literature, and the experimental results have demonstrated that the efficiency and effectiveness of our proposed algorithms are comparative with several state-of-the-arts super-resolution methods.
登录
查看更多内容
DOI：
10.1109/tassp.1978.1163154
发表时间：
1978-12
期刊：
IEEE Transactions on Acoustics, Speech, and Signal Processing
影响因子：
--
作者：
H. Hou;H. Andrews
通讯作者：
H. Hou;H. Andrews
DOI：
10.1109/cvpr.2006.15
发表时间：
2006-06
期刊：
2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06)
影响因子：
--
作者：
R. Urtasun;David J. Fleet;P. Fua
通讯作者：
R. Urtasun;David J. Fleet;P. Fua
DOI：
10.1109/cvpr.2000.854852
发表时间：
2000-06
期刊：
Proceedings IEEE Conference on Computer Vision and Pattern Recognition. CVPR 2000 (Cat. No.PR00662)
影响因子：
--
作者：
Simon Baker;Takeo Kanade
通讯作者：
Simon Baker;Takeo Kanade
DOI：
10.1016/1049-9652(91)90045-l
发表时间：
1991-05-01
期刊：
CVGIP-GRAPHICAL MODELS AND IMAGE PROCESSING
影响因子：
--
作者：
IRANI, M;PELEG, S
通讯作者：
PELEG, S
DOI：
10.1109/iccv.2005.193
发表时间：
2005-10
期刊：
Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1
影响因子：
--
作者：
R. Urtasun;David J. Fleet;Aaron Hertzmann;P. Fua
通讯作者：
R. Urtasun;David J. Fleet;Aaron Hertzmann;P. Fua