CAJ | 학술논문

提出一种改进的基于相似性约束的人脸超分辨率重建算法，采用迭代计算的方式将训练过程和学习过程整合在一起。首先从训练集中遴选出与待重建人脸最相似的训练库人脸参与迭代过程，随着迭代次数的增加，重建得到的高分辨率人脸越来越接近于原始高分辨率人脸；其中每次迭代分别统计待重建低分辨率人脸和训练集本次迭代参与的低分辨率人脸的相似性以及与训练集本次迭代参与的高分辨率人脸在局部结构上的相似性，以减少流形学习中低维空间到高维空间的一对多映射的限制。实验结果表明，与其他算法相比，文中所提的人脸重建算法不仅具有较低的空间复杂度，并且具有更好的主观和客观效果。
제출일충개진적기우상사성약속적인검초분변솔중건산법，채용질대계산적방식장훈련과정화학습과정정합재일기。수선종훈련집중린선출여대중건인검최상사적훈련고인검삼여질대과정，수착질대차수적증가，중건득도적고분변솔인검월래월접근우원시고분변솔인검；기중매차질대분별통계대중건저분변솔인검화훈련집본차질대삼여적저분변솔인검적상사성이급여훈련집본차질대삼여적고분변솔인검재국부결구상적상사성，이감소류형학습중저유공간도고유공간적일대다영사적한제。실험결과표명，여기타산법상비，문중소제적인검중건산법불부구유교저적공간복잡도，병차구유경호적주관화객관효과。
An improved face Super-Resolution ( SR) reconstruction algorithm based on similarity constraints is proposed. The proposed algorithm incorporates training stage and learning stage together. Select the most similar face sets ( low resolution faces and corresponding high resolution faces) from the whole training face sets with the input Low Resolution ( LR) face. With the increasing of iterative num-bers,the reconstruction result gets more and more close to the original High Resolution (HR) face. During each iterative learing,the simi-larity between the input LR face image and the training LR face image is computed as well as the local structure similarity between the in-put LR face and the training HR face. The experimental results demonstrate that the proposed algorithm not only occupies less space com-plexity but also produces better subjective and objective results compared with other leading super-resolution reconstruction algorithms.