face image super-resolution 人脸图像超分辨率
A novel region adaptive learning-based super resolution algorithm for human face images is proposed, which divides a face image into flat regions and detailed regions.
提出一种基于区域自适应学习的人脸图像超分辨率复原算法。
The method considers that a high-resolution face image is composed of the global face image and local high-frequency details.
该方法将一幅高分辨率人脸图像视为由全局脸和局部高频细节两者组成的。
Finally, the global face image and local details are combined to the final face super-resolution results.
最后，将全局脸图像和局部细节信息相结合，得到最终的人脸超分辨率结果。
应用推荐