Blind Image Super-Resolution with Spatial Context Hallucination
This addresses the challenge of real-world image restoration for applications like photography or surveillance, where degradation kernels are unknown, but it is incremental as it builds on existing super-resolution techniques.
The paper tackles the problem of blind image super-resolution, where high-resolution images must be reconstructed from low-resolution inputs corrupted by unknown blur and noise, by proposing a Spatial Context Hallucination Network that integrates denoising, deblurring, and super-resolution into one framework, achieving better performance than state-of-the-art methods on datasets like DIV2K and Flickr2K.
Deep convolution neural networks (CNNs) play a critical role in single image super-resolution (SISR) since the amazing improvement of high performance computing. However, most of the super-resolution (SR) methods only focus on recovering bicubic degradation. Reconstructing high-resolution (HR) images from randomly blurred and noisy low-resolution (LR) images is still a challenging problem. In this paper, we propose a novel Spatial Context Hallucination Network (SCHN) for blind super-resolution without knowing the degradation kernel. We find that when the blur kernel is unknown, separate deblurring and super-resolution could limit the performance because of the accumulation of error. Thus, we integrate denoising, deblurring and super-resolution within one framework to avoid such a problem. We train our model on two high quality datasets, DIV2K and Flickr2K. Our method performs better than state-of-the-art methods when input images are corrupted with random blur and noise.