Abstract
The recent increase in the extensive use of digital imaging technologies hasbrought with it a simultaneous demand for higher-resolution images. We developa novel edge-informed approach to single image super-resolution (SISR). TheSISR problem is reformulated as an image inpainting task. We use a two-stageinpainting model as a baseline for super-resolution and show its effectivenessfor different scale factors (x2, x4, x8) compared to basic interpolationschemes. This model is trained using a joint optimization of image contents(texture and color) and structures (edges). Quantitative and qualitativecomparisons are included and the proposed model is compared with currentstate-of-the-art techniques. We show that our method of decoupling structureand texture reconstruction improves the quality of the final reconstructedhigh-resolution image. Code and models available at:https://github.com/knazeri/edge-informed-sisr