Face super-resolution via robust online dictionary learning

Abstract

In order to solve the problems of face image super-resolution, a robust online dictionary learning method based on sparse representation is proposed in this paper. The online dictionary learning algorithms which can be used to train big sample datasets is introduced in the dictionary learning phase to generate better overcomplete dictionaries. Additionally, the classic L2-regularization is replaced by the robust L1-regularization in the spare coding procedure. The simulation comparisons and verifications in the experiments prove that the PSNR and SSIM of the proposed method are much higher than some state-of-the-art super-resolution algorithms. The PSNR is 0.72dB higher than [12] Mairal's online dictionary learning application in super-resolution, while the SSIM is 0.0187 higher. The performance of the proposed algorithm is promising with few artifacts along the edges. Meanwhile, the denoising effect is much better than some classic algorithms while processing noisy face images.