Variational AutoEncoder for Reference based Image Super-Resolution
This addresses the limitation of single image super-resolution methods for high upsampling factors, enabling diverse outputs from arbitrary references, though it is incremental in combining VAE with reference-based techniques.
The paper tackles the problem of image super-resolution at large upsampling factors (e.g., 8×) by proposing a reference-based approach using a Variational AutoEncoder (RefVAE), which can generate diverse super-resolved images from a hidden space and achieves higher diverse scores compared to state-of-the-art methods.
In this paper, we propose a novel reference based image super-resolution approach via Variational AutoEncoder (RefVAE). Existing state-of-the-art methods mainly focus on single image super-resolution which cannot perform well on large upsampling factors, e.g., 8$\times$. We propose a reference based image super-resolution, for which any arbitrary image can act as a reference for super-resolution. Even using random map or low-resolution image itself, the proposed RefVAE can transfer the knowledge from the reference to the super-resolved images. Depending upon different references, the proposed method can generate different versions of super-resolved images from a hidden super-resolution space. Besides using different datasets for some standard evaluations with PSNR and SSIM, we also took part in the NTIRE2021 SR Space challenge and have provided results of the randomness evaluation of our approach. Compared to other state-of-the-art methods, our approach achieves higher diverse scores.